Fibre optic measurements
Network Dike Monitoring
workshop Fibre optic monitoring for flood defence management
On 27 June, the fibre-optic monitoring workshop for flood defence management took place in Tiel. This workshop covered the technical principles of fibre optic monitoring. In addition, the results of measurements with fibre optics were presented to the attendees. The State of the Art regarding fibre optic monitoring was also presented. This report looks back on the workshop and reports the results in outline.
Welcome and introduction
Everyone is welcomed by Wouter Zomer. He briefly explains what the Dike Monitoring Network does. The network works as a facilitator and agenda-setting body, among other things by organising meetings like this one. The aim is to hold about eight such meetings a year, each with a different topic. The next workshop will take place in the autumn. The programme for the afternoon will then be presented. It was decided to swap the State of the Art presentation with the case. This is also reflected in the programme below.
13:00 Reception at Waterschap Rivierenland (De Blomboogerd 1, Tiel)
13:10 - 13:25 Welcome and presentation programme by Wouter Zomer
13:25 - 13:40 Presentation on the application of fibre optics at WSRL by Gerjan Westerhof
13:40 - 14:20 hrs Presentation technology and results achieved by Axel Fabritius
14:20 - 15:20 Presentation State of the Art by André Koelewijn and Martin van der Meer
15:20 - 15:30 Pause
15:30 - 16:15 hrs Working out the case
16:15 - 16:45 hrs Discussion by Wouter Zomer
16:45 Closure and drinks
Monitoring at WSRL
When monitoring, it should always be kept in mind that it is not only done for implementation, but that monitoring is applied over the entire life cycle of the dike.
There was room for innovation in the dike reinforcement project Kinderdijk - Schoonhovense Veer (KIS). The dike there lies on 15 metres of peat, making subsidence common. Using fibre-optic cables in the dike, this subsidence can be monitored. The cables and equipment used there only show that subsidence is taking place, but not in which direction. However, because of the thick peat pack under the dike, it seems pretty clear that the displacement is downwards. However, it never managed to link the monitoring to the calamity programme. No automatic signal goes off when x cm of subsidence occurs. The available data has also not yet been processed. Consideration is now being given to having this processed and analysed by means of a graduation project.
Besides the application of glass fibre at KIS, glass fibre strips have also been applied to anchors at Streefkerk to monitor what the anchors are doing underground. This did show that a lot of different parties have to be involved, someone has to glue the fibre-optic cable to the anchors, the contractor has to know what to do, and so on.
Furthermore, a vertical sand-tight geotextile has also been put into the ground. This is also monitored through fibre-optic cables. When a temperature difference is observed, it can be said that groundwater flow is taking place at that location. The tests that have been done are nearing completion. It must then be determined whether to continue with the measurements or not. This does raise the question of who will take over the management of such measurement systems.
Another disadvantage of fibre optics is that it is a very sensitive measurement method. That is why nothing can be done in the surrounding area in terms of buildings and the like. Gerjan emphasises again that the life cycle of a dike should be more integrated, with monitoring also having a role in management and maintenance, testing and strengthening.
Technology and results achieved GTC Kappelmeyer
Axel begins his presentation with a description of what can be measured with fibre-optic cables. By GTC Kappelmeyer, this technique is mainly applied in detecting leaks in dams and dykes. In doing so, temperature (difference) is used as a natural tracer. The temperature increase or decrease of the water is shifted over time and flattened as one goes deeper to measure. Around a depth of 10 to 15 metres, the temperature in the ground becomes constant (about 10 degrees Celsius).
As an example, a dike is outlined where a slurry trench had been constructed to a depth of 12 metres. However, after about 10 years, water resurfaced at the toe of the dike. Therefore, temperature readings up to a depth of 20 metres were taken using fibre optics. This showed that water was flowing under and through the trench.
Operation
The operation of temperature measurements with glass fibre are as follows. A light signal is sent through the cable. This signal is reflected back, its frequency changing with temperature. The part of the signal that matters is the Anti-Stokes part, as this is temperature-sensitive. This also means that the cable itself is the sensor, allowing measurement along the entire length of the cable.
There are roughly two methods of measuring temperature with fibre-optic cables. The first is the gradient or passive method. With this, a difference is measured between the temperature of the ground in which the cable lies and the temperature of the water. If water flows past the cable, the measured temperature will change.
The second way is the 'heat-up' method. This involves heating the cable and measuring the degree of cooling. This degree of cooling says something about the water flowing along the cable. This method is more accurate than the passive method and particularly suitable for situations where the water and the ground are (almost) the same temperature.
The system can also work automatically by means of a traffic light system. Green then means that nothing is wrong, yellow deserves attention and red means that a closer look should be taken as soon as possible at what is going on. This uses the difference between the reference situation, the results of the heat-up measurements and furthermore the gradient method and the effective thermal conductivity.
In addition, there is also the possibility of installing glass fibre sensors in the ground every 10 metres, for example. This was done in Avignon, for example, where fibre-optic cables were inserted into the ground every 10 metres to a depth of 16 metres. This was done to check the quality of the slurry trench constructed by the contractor. If the trench was not watertight, the contractor would not get paid. Measurements were taken before, during and after the construction of the slurry trench and in the end it was found that the slurry trench was indeed watertight.
In dykes, horizontal fibre-optic cables have generally not been found to be very effective, as they are often shallow and therefore suffer too much from outside influence. If they are laid deeper, the effectiveness is greater, but the greatest effectiveness is achieved when the fibre-optic cables are laid vertically in the dike. At least the first metre below ground level suffers from the daily influence, below that this becomes less.
By default, the fibre-optic sensor only gives the location where something is different, but further analysis of the data can also find out the order of magnitude of the leakage rate. By sophisticated calibrations, the estimate of seepage can be close to the actual value.
The use of glass fibre sensors generally focuses only on temperature or only on deformation. The combination can also be measured, but this does not improve accuracy. A single-strand (thin) cable is used for deformation measurements, while a multiple-strand (thicker) cable is used for temperature measurements. In practice, the accuracy of temperature measurements with fibre-optic cables is 0.1 to 0.2 degrees Kelvin (or degrees Celsius). Only under laboratory conditions are higher accuracies achievable, up to 0.01 to 0.02 K.
The heat-up method has the disadvantage that only shorter sections can be measured at once compared to the passive method. This is caused by the energy required to heat up the cable. If the cable becomes too long, a standard mains connection no longer suffices. This is not very practical. The energy consumption for heating up the cable is about 10 W/meter for one to one and a half hours of heating up the cable.
The prices of fibre optic monitoring obviously depend on the size of the installation. Equipment prices range roughly between 35k€ and 70k€. In addition, cables and installation are then required. In total, the costs are around 80k€ to 100k€ for a permanent installation with alarms, etc.
State of the Art
The report summarising the State of the Art relating to fibre will be made available soon. A link to this report, once published, will be sent after to participants.
Lifecycle monitoring is very important. As Gerjan also pointed out, monitoring is not something that is only applied in one phase of the life cycle of the dike, but something that should be applied over the entire life cycle of the dike. Connecting monitoring systems to management is often very difficult. This connection often works for each project, but it is not yet successful for all projects as a whole. Measuring and monitoring are not only important during floods, but also outside them, for instance for human and animal digging activities.
Only one device or instrument can never be used in monitoring. A combination is always needed. This only increases accuracy and reliability. For monitoring to be successful, a good monitoring plan is essential. In addition, good quality is important. This can be achieved through a good monitoring plan. The quality of the monitoring plan can be checked with the Handreiking Life Cycle Monitoring. This does not describe the best method or technique, but it does describe what to pay attention to when drawing up or assessing a monitoring plan, which also provides concrete insight into how this can be improved in your own specific situation.
You will never have too much data. However, good data management is very important in this respect. If data management is not in order, little will be done with the information gathered. However, in case of a flood, for example, it is very useful to have historical data and to have as much information as possible about the area.
The relationship between the parameter(s) used in the formulas and the cause of failure are sometimes hard, but sometimes also indirect. It is therefore important to measure the right parameters at the time and location that it is needed. In this regard, it is also very important to have a good baseline measurement. After all, without a good baseline measurement, the data collected does not say much.
There are three techniques that can be used to measure. The first is Fibre Bragg Grating (FBG). This involves making unique notches in the fibre-optic cable. These notches each create a slightly different reflection, allowing the exact location to be determined. However, such a cable is more vulnerable (because of the notches) and more expensive than a standard fibre-optic cable. It is possible to apply up to about 500 unique notches to a cable. There is also the distributed technique, which relies on changing the reflection of the signal due to changed conditions in and around the cable. This is also the method explained by Axel. It allows measurements to be made in close succession, but the equipment needed for this is very expensive. The cable, on the other hand, is cheaper. Finally, there is also the possibility of measuring only at the end of the cable. However, this means using a separate cable for each measurement. With this technique, water stresses, for example, can be measured. This has also been applied in the DMC system, among others. It is mentioned that the reading equipment for Rayleigh scattering is very expensive (between 20k€ and 100k€).
The application of these techniques is already possible, but the question is which question of the managers can be solved/answered with it. The problem here is that dykes are not continuously loaded and it could therefore be that nothing is measured for 10 years because no high water has passed. In response, regional flood defences are more vulnerable and have lower standards. These regional flood defences will therefore be loaded more often and could possibly serve as a testing ground. It is also noted that a good baseline measurement can take years, because you want to measure over all seasons.
When using monitoring, there are not always windfalls (the dike is stronger than expected). There can also be setbacks (the dike is weaker than expected). However, as manager, you should also want to know about these setbacks. Nowadays, that is part of the duty of care.
The biggest gain to be made from fibre-optic monitoring is in the reduction of the schematic uncertainty. This can now also include the behaviour of the dike, instead of just the measured parameters.
During the IJkdijk trials in 2008, fibre-optic cables were also used in the crest, halfway up the slope, in the toe and in the field in front of the dike. The measuring equipment stopped measuring 24 hours before embankment collapse. However, severe stresses could already be observed in the fibre-optic measurements at that time. So fibre-optic monitoring can have a warning and predictive effect when a dike is about to collapse. At that point, intervention can then be taken. In 2012, fibre-optic sensors were also used at another IJkdijk to observe piping. There too, the performance of the fibre-optic sensors was promising.
In addition, another example is shown of the application of fibre-optic sensors in a dike in Italy. This is the first (and only) place where fibre-optic sensors have been applied in Italy. More about this project can be found in the attached paper ('Effectiveness of distributed temperature measurements for early detection of piping in river embankments') by Bersan et al.
It is again stressed that context is very important. "1 measurement system is not a measurement system", a combination of systems will always have to be applied.
Case
The case worked out relates to fibre optic monitoring in the Willemspolder. A trial was conducted there with vertical sand-tight geotextile to counter piping. The trial tried to find out whether the geotextile would silt up with sand and therefore act as a short piping screen. In 2012, the Vertical Sand Dense Geotextile was also tested and then it proved effective against piping. In that test, seepage was seen, but no sand-carrying seepage (after an initial familiarisation period).
A test with this geotextile was conducted at a summer dike in the Willemspolder. The audience is asked what measuring systems would be needed to determine whether the geotextile still allows water to pass through or whether it is silted up with sand.
After some elaboration time, similar solutions emerge from the different groups. Some solutions mention the installation of a glass fibre cable in front of and behind the geotextile at the level of one or more sand layers. When water still flows through this, no difference in temperature will be observed. If the geotextile is sealed, a difference in temperature will be observed.
Another option put forward is installing a heatable fibre-optic cable in the fabric at the level of one or more of the sand layers to determine cooling. When the cable does not cool, no water will flow through it, when it does cool, water will flow through it.
The final solution applied is as follows. A heatable cable was woven into the fabric and a heatable cable was placed 1.30 metres behind the geotextile for control. In addition, water pressure sensors were installed at the same height as the fibre-optic cables, just like at the bottom of the geotextile.
It should be noted, however, that the trial did not go entirely according to plan. The geotextile was eventually installed over only 30 metres, instead of the full 100 metres. In addition, the geotextile was only laid 3.5 metres deep, instead of the planned 5 metres. However, the results showed that the cable cooled down and thus water continued to flow through the geotextile. In order to use fibre-optic cables in this way, the depth of the sand layers will have to be known in advance. Only in the sand layers will the fibre-optic cable deliver usable results.
Closing
After the discussion of the case, this afternoon's speakers will be thanked by means of a Deventer biscuit package. Afterwards, everyone is invited to discuss the proceedings while enjoying a snack and a drink over drinks.
Description
On Wednesday 27 June 2018, from 13:00 to 17:00, a workshop on fibre optic monitoring for dykes will be organised by the Dike Monitoring Network. This workshop will present how fibre optic measurements can be used in the practice of flood defence management. In addition, the State of the Art of these techniques will be presented.
During the workshop, a case will be discussed. With the experts present, you can elaborate on your case to determine whether and, if so, how fibre optic monitoring can be applied to you. Would you like to present your case during the workshop? When registering for the workshop, please mention that you would like to submit a case, along with a brief explanation of the case. We will then contact you.
Speakers during the workshop include Martin van der Meer of Fugro, André Koelewijn of Deltares and Axel Fabritius of GTC Kappelmeijer from Germany. Besides the substantive presentations, there will again be plenty of room to discuss usefulness and necessity for you as a dike professional and which applications of fibre-optic systems are suitable and applicable to you, for management and assessment.
The meeting will take place at Waterschap Rivierenland. You can register to participate in this meeting at netwerk@dijkmonitoring.nl. When registering, please mention that you are registering for the workshop on fibre optic measurements. After registering, you will be informed about your participation and kept up to date about this meeting by e-mail. There is no charge for participating in this workshop
Location
De Blomboogerd 1, 4003 BX Tiel
Organiser
Land Use and Water Management
Network Dike Monitoring
Name and contact details for information
Wouter Zomer
Apply via
Network Dike Monitoring
netwerk@dijkmonitoring.nl