Monitoring foundations

Abstract

It is estimated that about 400.000 houses in the Netherlands – both shallow and wooden pile foundations – could be facing problems with their foundations. Although, house owners are becoming increasingly aware of the fact that problems (and thereby costs) can be prevented if remedial measures are taken in time. Monitoring of the foundation can possibly give insights in the remaining capacity of the foundation. In this thesis, research is performed into monitoring of foundations. The main research question is: Which methods can be used to monitor a foundation and to what extent are these useful and adequate? To answer this question, five different monitoring methods have been investigated: InSAR (Interferometric Synthetic-Aperture Radar), measurement bolts, a subsidence sensor, a groundwater level sensor and a tilt sensor. Other methods have not been considered and are beyond the scope of this research.
First, InSAR data has been compared to subsidence data gathered by measurement bolts in two different cases, one in Zaandam (Netherlands) and Amsterdam (Netherlands). Measurement bolts are used as golden standard in this comparison. The comparison of measurement bolts with INSAR data resulted in a maximum error margin of ± 0,86 mm/year for Case Study Block A and ± 1,31 mm/year for all measurements for the testcase in the Case Study Block B within a 95% confidence interval. Secondly, the measurement bolts method has been qualitatively compared to InSAR data. Also, the pros and cons have been discussed. Measurement bolts have a better precision (0,3-0,5 mm) than InSAR data (2-3 mm) for individual measurements. Possible errors that can influence the accuracy of this method are: personal errors, instrumental errors and natural errors. InSAR and measurement bolts both have their strengths and weaknesses. Prescreening is an easier application for InSAR but highly accurate measurements in a specific location is more reliable if measurement bolts are used. In a way, these methods complement each other rather than measuring the same way. The datasets collected by the subsidence, groundwater level and tilt sensors is made available for this research project by KennisCentrum Aanpak Funderingsproblematiek (KCAF) which is an independent knowledge and network organization funded by the government of the Netherlands. Generally, several factors can add noise to the signal. It is unknown if the signal of the sensors is calibrated to measure the right quantity. This research project has shown that the data showed a similar trend of the groundwater level as local monitoring wells. Therefore, it can be concluded that the sensors are adequate to measure the groundwater level if they are calibrated and referenced to NAP. Moreover, the groundwater level sensor can possibly be used to measure how long dry periods are and how much wood is exposed to air, in order apply measures well in time.
Tilt sensors are currently not used for monitoring of foundations. The factors that can be considered when a tilt sensor is used to monitor a foundation are investigated with a literature study and with expert opinion. As the results show the measured rotation can be related to the stiffness of the wall where the sensor is mounted on. As for all sensors, it has to be made sure if they are calibrated to actually measure what they observe. Further research could usefully explore how viable a system of sensors is. In general, classification or a prediction of the remaining service life of a foundation can be challenging. Further research might explore how useful it is when big dataset are combined to be able to give better judgement.