Proceedings of the First International Conference on Self Healing Materials 18-20 April 2007, Noordwijk aan Zee, The Netherlands V.M. van Beek et al. BIOSEALING, A NATURAL SEALING MECHANISM THAT LOCATES AND REPAIRS LEAKS V.M. van Beek*, D. den Hamer, J.W.M. Lambert, M.N. Latil and W.H. van der Zon * GeoDelft Stieltjesweg 2, 2628 CK Delft The Netherlands Tel: 0031 (0) 15 2693528 e-mail: V.M.vanBeek@GeoDelft.nl In civil-engineering projects, leakage problems can origin from deficiencies in water-retaining constructions or in natural layers. These problems result in higher drainage costs, damage to the surrounding areas as a result of settlements, inflow of undesired salty or polluted ground water or loss of water from reservoirs. Even if the leaks can be located with traditional techniques, it is often difficult to seal them because of the poor accessibility of the leak or the high costs involved. In addition, it is also desirable to solve these problems in an environmentallyfriendly and sustainable way. In order to self-heal leaks in civil engineering applications, a novel approach based on stimulating of naturally present bacteria in the subsurface has been investigated. This stimulation involves injection of a nutrient solution in the surrounding area of the leak. As the nutrients mix with the ground water flow they are automatically transported to the leak resulting in an increase in bacterial activity at the leak location. This increased biological activity in the vicinity of the leak results in a sustainable mechanical blockage within several weeks. The clogging is caused by a combination of bacterial activity and converging streamlines towards the leak. Laboratory studies were initiated to demonstrate the efficiency of this process and successful clogging was obtained within several weeks after the injection of nutrients. The tests were conducted under a lot of different conditions in sand-filled PVC columns which contained simulated leak in a water retaining layer (a disc with a hole in its center) and a realistical water flow to simulate a leaking construction pit. Several weeks after the injection of nutrients, a significant clogging at the location of the disk was observed. In addition to laboratory tests, field-scale trials were also conducted at the Maasvlakte in the port area of Rotterdam, and a practical scale test was performed along a newly constructed aqueduct beneath the Haarlemmermeer ring canal in the Netherlands. Both tests were successful. Biosealing is a promising selfhealing technology to reduce leakage in civil engineering applications. The technique is simple, makes use of naturally present microorganisms in the soil and is self-regulating for leak detection and reduction. 1 Introduction In civil-engineering projects, damage to water retaining constructions or natural non-porous layers can result in leakage problems within a construction site. These problems result in higher drainage costs, damage to the surrounding areas as a result of settlements, inflow of undesired salty ground water or loss of water from reservoirs. The leak is often difficult to locate, which makes treatment difficult, costly and delaying. Once the location is known, the treatment may consist of chemical injections close to the leak. All these factors imply additional costs in the project as well as delay in time. 1 © Springer 2007 Proceedings of the First International Conference on Self Healing Materials 18-20 April 2007, Noordwijk aan Zee, The Netherlands V.M. van Beek et al. In a preliminary feasibility study solutions for the locating as well as sealing of the leak were investigated and clogging by means of feeding bacteria that are naturally present in the subsurface appeared to be most promising. In this research it is attempted to clog a leak in the subsurface by means of feeding bacteria that are initially present in the subsurface with a substrate. A laboratory study was executed to prove the mechanism at small scale, followed by a field experiment to test the method at a realistic scale. The purpose of the research was to test whether the injection of substrate leads to significant clogging of a leak. Furthermore it is attempted to gain insight into the processes which contribute to the clogging effect. 2 Clogging mechanisms Clogging is a well-known and extensively studied subject in literature. Clogging is often encountered as an unwanted process in injection or extraction wells in ground water. Research in this area has lead to the statement of various clogging mechanisms that may occur depending on the situation. Although in this research this proces is not an unwanted sideeffect, it is believed that the same mechanisms that are important in clogging wells may play a role in the clogging of leaks, as performed in this research. Several causes of decrease of permeability in soils are named by Baveye et al. (1998). These involve the presence of suspended particles, swelling and dispersion of clay particles in the soil and several biological causes, like the accumulation of metabolic byproducts, the entrapment of gases and the precipitation of iron sulfides. In this research the presence of suspended particles is also initiated by the injection of the substrate, as the substrate itself contains small particles. The swelling and dispersion of clay particles may be caused by a change in electrolyte concentration. Although this is considered to be a chemical process, this change may be induced by the degradation of organic material by bacteria. Due to the fermentation process acid is formed and the availability of H+-ions causes instability of clay particles, as binding ions in clay (Mg2+, Fe2+ and Al3+) are exchanged by the abundant H+-ions. A change of environment may result in renewed stabilization of the particles. Dispersion of clay particles may also be caused by the addition of the substrate itself. The substrate is a highly conducting fluid. As salt flushes trough the soil pore volume it binds to clay particles exchanging with binding ions in clay. The sodium causes instability of the clay particles resulting in mobilization of clay colloids. When stimulated, most bacteria excrete slimy or gummy materials on their surfaces, so called exopolymers. The production of this material may temporarely block the pores in the soil. Several processes exist, related to bacterial activity, by which gas is formed. Gas which is trapped in the pores results in a reduction of the permeability. 2 © Springer 2007 Proceedings of the First International Conference on Self Healing Materials 18-20 April 2007, Noordwijk aan Zee, The Netherlands V.M. van Beek et al. Proceeding bacterial activity causes a sulphate reducing environment. By this time ferri (Fe3+) is reduced into Ferro (Fe2+), which can react with sulfide into iron sulfide (FeS). The above-named processes may interact with eachother. A combination of several processes may lead to permanent clogging whereas separate processes do not. 3 Laboratory experiments Laboratory experiments were initiated to test the hypothesis that feeding of bacteria may result in clogging of a leak. Many experiments have been performed from which one of them has been chosen as an example (GeoDelft, 2006). Perspex columns (Ø 7.85 cm x 50.56 cm length) were filled with sand and a disk with a centered hole (Ø 1.2 cm) was placed in the column to simulate a leak. A constant hydraulic head difference over the column has been applied, causing a constant upward flow of water in the column. The columns were operated under environmental conditions comparable to subsurface conditions, concerning exposure to sunlight, ambient temperature and complete saturated conditions. The set up is located in a controlled temperature room, set at 12 ± 1°C. The columns were rapped in dark plastic bags to prevent income of sunlight during reading of measurements. A natural sand species is chosen to resemble both chemical and biological properties in a field situation. Monitoring consists of measuring the flow through the columns as well as the pressure heads at various locations in the columns. Using these parameters the dimensionless clogging factor can be calculated: the ratio of the current and initial hydraulic resistance between two piezometers. The environmental parameters that have been monitored are conductivity, pH, temperature, oxygen concentration and redox potential. The used substrate is Nutrolase, which is a protamylasse, a wasteproduct of processed potatoes, which is supplied by the AVEBE concern. Nutrolase is injected into the column after dilution. The column clearly shows a severe increase in clogging factor within two and a half weeks after the injection of Nutrolase solution. During the same time, a Blanco column has been monitored without nutrition. The clogging factor has been calculated over the disk en shows a maximum of 460 after 60 days. The clogging appears to concentrate at the location of the hole, as clogging factors calculated for other parts of the column are considerably smaller. The clogging in the Blanco column is insignificant. 3 © Springer 2007 Proceedings of the First International Conference on Self Healing Materials 18-20 April 2007, Noordwijk aan Zee, The Netherlands V.M. van Beek et al. Clogging factor 500 450 400 Clogging factor [-] 350 300 250 Feeding period of 8.5 h 1600 ml diluted (1:80) Nutrolase with particles 200 150 100 50 0 0 20 40 60 80 100 120 140 160 180 200 Time [days] Column nutrition Blanco Column Clogging factor in the column with nutrition and the blanco column Observations during demobilization do not show any evidence of gas formation or iron sulfide. Redox measurements during the test confirm that the formation of iron sulfides has not been possible, as conditions have not been sulfate-reducing. 4 Field experiments To upscale the laboratory experiments to a realistic situation, a field experiment has been perfomed [GeoDelft, 2004]. The purpose of the field experiment was to reduce the flow in a known leak with at least a factor 5, to prove that BioSealing is applicable at field scale, and to gain insight in the clogging processes. A leakage has been simulated by burying three 20ft sea containers vertically in the subsurface at a location on the Maasvlakte nearby Rotterdam. All three of the containers had holes in different shapes. The shapes varied from an elongated crack to several small circular holes (0.08m) or one big hole (0.25m). A constant head difference caused flow through the leak. To simulate a building pit situation in and outside the container, the hydraulic head difference was maintained at 3 m. 4 © Springer 2007 Proceedings of the First International Conference on Self Healing Materials 18-20 April 2007, Noordwijk aan Zee, The Netherlands V.M. van Beek et al. Burying of the sea containers The nutrients were injected directly into the subsurface by using a packer which is placed in a tube à manchette (a vertically placed perforated pipe). The space around the tube à manchette was filled with gravel to prevent clogging directly around the tube. The tubes were placed at 2.5 to 5 m distance from the leak. Each leak was surrounded by three tubes à manchette to ensure the substrate to reach the leak. During the injection of Nutrolase towards the first container (container 2) it was found that the optimal concentration of Nutrolase was a dilution of 1:40 and continous injection of 15 l/day. Monitoring consisted of measuring flow, hydraulic heads in the subsurface around the sea container and various environmental parameters. From the observations the clogging factor was calculated in time. A large circular hole of 0.25 m in diameter has been created in the bottom of container 2. Container 2 was the first container to be injected and no clogging has been achieved after the single shot which was initially planned. After this period the schedule has been changed to continuous injection of 15 l/day. The clogging factor increased then to a maximum of 30. No clogging was observed in the nearby buried container 3, which was kept as a blanco until that time. Clogging factor container 2 C [-] 100 10 1 16 February 2004 1 March 2004 15 March 2004 29 March 2004 12 April 2004 26 April 2004 10 May 2004 date Clogging factor container 2 5 © Springer 2007 Proceedings of the First International Conference on Self Healing Materials 18-20 April 2007, Noordwijk aan Zee, The Netherlands V.M. van Beek et al. 6 smaller circular holes were made in container 3. Container 3 was the second container to be injected with nutrition. The injection was continuous at a rate of 15l/day. Monitoring has taken place during 66 days from the first injection after which a clogging factor was reached of 3.6. At that time the test had to be interrupted, although the clogging factor was still increasing. blanco Clogging factor container 3 injection C [-] 10 1 16 February 2004 1 March 15 March 29 March 2004 2004 2004 12 April 2004 26 April 2004 10 May 2004 24 May 2004 7 June 2004 21 June 2004 5 July 2004 19 July 2004 date Clogging factor container 3 An elongated crack, 1m long and a decreasing width of 0.1 to 0m, has been made in the bottom of container 1. After 57 days of injection the clogging factor was 7.3. The factor was still increasing at the time the test was interrupted. Clogging factor container 1 C [-] 10 1 17 May 2004 31 May 2004 14 June 2004 28 June 2004 12 July 2004 26 July 2004 date Clogging factor container 1 After the experiment visual observations were executed to obtain information about the clogging materials. After removal of the containers, a black colouring of the sand was observed at locations between injection and leak, with increasing black colour in the direction of leakage. Acidification of the black precipitate results in the formation of H2S, indicating that the observed black colouring is iron sulfide. The presence of iron sulfides is confirmed with ESEM/EDAX techniques. The iron sulfide seems not to be totally blocking the pores, but is rather present as a layer around the sand grains. 6 © Springer 2007 Proceedings of the First International Conference on Self Healing Materials 18-20 April 2007, Noordwijk aan Zee, The Netherlands 5 V.M. van Beek et al. Discussion Both the field experiment and the laboratory experiment have shown that the injection of diluted Nutrolase results in sealing of the leak. The process of clogging can not be completely retrieved from the performed experiments. It is unknown what the influence of the precipitation of iron sulfide on the clogging process. The circumstances in the laboratory were aerobic, preventing the formation of iron sulfide. The laboratory conditions were therefore not equal to subsurface conditions. Gas formation is unlikely, as this has not been observed in the experiment. The other processes, dispersion of clay particles and formation of metabolic by-products may be combined in the process of BioSealing. Based on all the experiments performed on BioSealing processes the hypothesis on the clogging mechanism for BioSealing is as follows: 1. hydrocarbons (such as sugars and/or peptides) are decomposed under anaerobic conditions. During this fermentation organic acids are formed. 2. The formed acids initiate an erosion of the natural particles in the soil, such as feldspar. The particles formed, are so small that they can migrate through the pores of the sand. 3. Further decomposition yields into the formation of biomass on that specific spot where the most nutrients are available (there where the flow is high, thus in the surrounding and inside the leak). 4. This biomass captures the eroded particles. 5. As a result of the presence of multi valent ions, such as calcium, iron and/or magnesium a flocculation of the captured particles will occur. This results in a “lump” of clay-like particles inside the pores of the sandgrains. 6 Conclusions The process of BioSealing is successful in the field and in the laboratory. BioSealing proves to be a clean and sustainable technique, whereby bacteria are stimulated which are intially present in the subsurface, which is most active at the location of the leak. BioSealing can therefore be considered suitable for locating and repairing leaks in practice. Further research needs to be performed to gain insight in the exact process of clogging. 7 © Springer 2007 Proceedings of the First International Conference on Self Healing Materials 18-20 April 2007, Noordwijk aan Zee, The Netherlands V.M. van Beek et al. REFERENCES 1. 2. 3. 4. 5. [Baveye et al., 1998] Environmental impact and Mechanisms of the Biological Clogging of saturated soils and aquifer materials, Crit.Rev. in Environmental Science and Technology, 28(2); 123-191 (1998) [GeoDelft, 2003] Extensieve herstelmethoden, Factual report invloed omgevingsfactoren, GeoDelft rap.no. CO-406290.0032 (2003) [GeoDelft, 2004] BioSealing Veldexperiment, GeoDelft rap.nr. CO-406290.0051 Extensieve herstelmethoden van waterremmende constructies, field experiment, graduation report by V. Veenbergen, Januari 2004 [GeoDelft, 2006] BioSealing laboratory experiments, GeoDelft rap. no. CO-421180.0004 (2006) (draft) 8 © Springer 2007