Evaluation of long-term performance of sodium silicate grouted in embankment dams

(1)

OCH HUVUDOMRÅDET MILJÖTEKNIK,

AVANCERAD NIVÅ, 30 HP STOCKHOLM SVERIGE 2019,

Evaluation of long-term

performance of sodium silicate grouted in embankment dams

JENNY FU

KTH

SKOLAN FÖR ARKITEKTUR OCH SAMHÄLLSBYGGNAD

(2)

(3)

Abstract

Embankment dams is the most common type of dams in operation in Sweden today. Due to the nature of embankment dams, seepage through them will always occur. If the seepage velocity exceeds a critical velocity, internal erosion is initiated, which could lead to damage in form of piping and sinkholes. To treat this problem, remedial grouting has been performed involving a combination of conventional grouts, i.e. cement and cement-bentonite as well as sodium silicate, which is a chemical grout that also known as water glass. Regarding the sodium silicate grout, there is concern about the long-term permanence.

The aim of this thesis has been to study the potential performance of sodium silicate grouted in embankment dams. The first part of this thesis is a literature review of the general behavior of sodium silicate as a grout, its degradation processes and the factors that could induce degradation.

The second part suggests monitoring methods to control and evaluate the performance of the treated dam and the grout if degradation has occurred.

Findings from literature generally indicates a high risk of instability and low permanence of sodium silicate when grouted in an embankment dam.

This type of grout will undergo degradation mainly in two forms: syneresis induced shrinkage and leaching due to grout erosion or dissolution. As the degradation has developed, an increase in permeability of the repaired dam core is a potential consequence.

How the potential degradation of sodium silicate will affect the treated dams is suggested to be observed by monitoring the permeability of the grouted core. Applicable monitoring methods for this purpose are measurements of pore pressure and temperature using piezometers. The second direct method of monitoring a changed dam behavior is suggested to be leakage analysis, in order to detect potentially increased leakage because of the grout degradation. An indirect way to investigate the dam performance is suggested to be monitoring of the grout state. Measurement

(4)

of ion concentration of sodium and silicon respectively in leakage water using selective-ion electrodes will reveal any increase in ion concentration due to the potential grout dissolution or leaching.

Keywords

Sodium silicate, water glass, embankment dams, remedial grouting, dam monitoring

(5)

Sammanfattning

Fyllningsdammar är den vanligaste typen av dammar som används i Sverige idag. Eftersom denna typ av dammar består av packad jord- och bergmaterial kommer vattengenomströmning genom dammen alltid att ske. När genomströmningshastigheten överstiger dess kritiska hastighet kan inre erosion initieras. Piping och sjunkhål i dammen kan uppstå om fortskridande inre erosion utvecklats. Reparationsåtgärder i form av injektering har tillämpats för att behandla denna typ av problem kopplat till inre erosion i svenska fyllningsdammar. En kombination av cement, cement-bentonit och natriumsilikat som också kallas för vattenglas hade tillämpats som injekteringsbruk vid reparation. Beständigheten av natriumsilikat-bruk har emellertid varit ifrågasatt.

Detta examensarbete har utförts med syfte att analysera beständigheten av natriumsilikat-bruk i fyllningsdammar. Första delen av rapporten är en litteraturstudie som beskriver det generella beteendet av natriumsilikat-baserat injekteringsmedel, dess nedbrytningsprocesser samt att identifiera faktorer som kan ge upphov till nedbrytning av natriumsilikat. Den andra delen är förslag till övervakningsmetoder för att bevaka funktionen av den behandlade dammen vid en eventuell nedbrytning av natriumsilikat-bruket.

Resultatet från litteraturstudien indikerar att det finns hög risk för nedbrytning av natriumsilikat injekterat i fyllningsdammar. Urlakning och krypning av injekteringsmassan är två olika nedbrytningsmekanismer. Ett möjligt resultat av nedbrytningen av natriumsilikat är att den reparerade tätkärnan får högre permeabilitet.

För att övervaka en eventuell förändrad funktion av dammen p.g.a.

nedbrytningen av injekteringsbruket har tre metoder föreslagits.

Övervakningen kan genomföras med en direkt metod för att bevaka förändringar och inkluderar mätning av portryck och temperatur i vattenståndsrör till följd av en förändrad permeabilitet av den reparerade

(6)

dammen och dess tätkärna. Övervakning av läckageflöde är en annan direkt metod som föreslagits för att detektera förändring av injektering. I den indirekta metoden mäts jonkoncentrationen av natrium- och kiseljoner i läckagevatten från dammen för att erhålla indikationer på om en pågående nedbrytning av natriumsilikatbruket äger rum.

Nyckelord

Natriumsilikat, vattenglas, fyllningsdammar, reparationsinjektering, övervakning av dammar

(7)

Acknowledgements

This thesis project is funded by Energiforsk and made in collaboration with Sweco Power Generation and Dams. I would like to thank Energiforsk that made it possible to conduct this project.

I would like to thank my examiner Assoc. Prof. Fredrik Johansson for his advice and guidance throughout this project. I would like to thank my supervisors M.Sc. Ingvar Ekström, engineer and expert within embankment dams and dam safety at Sweco, and my co-supervisor Dr Marie Westberg Wilde, researcher within dams and dam safety at ÅF and KTH for their invaluable help and encouragement of my learning. Also, I would like to thank Peter Viklander, Adj. Prof. at LTU and Vattenfall Vattenkraft AB for the valuable review comments on the performed work.

Finally, I would like to thank Mattias Jender, group manager for Vattenkraft och Dammar, and everyone else at Sweco Power Generation and Dams for being so welcoming and kind.

(8)

(9)

1. Introduction

Embankment dams are the most common type of dams in operation in Sweden today. This type of dam is constructed with compacted earth material, therefore seepage through the dam will always occur even when operating as designed. If the seepage velocity becomes high enough to detach fines in the core and transport them, internal erosion may be initiated.

ICOLD (2017) defined that internal erosion is the phenomena when the fine soil particles in the core are carried away by seepage. Continuous internal erosion can occur when filter criteria are not met, leading to insufficient protection against mass movement and material loss. As erosion progresses it may result in enlarged seepage paths and excessive flow. As a result, damage in form of piping and sinkholes may occur.

Several dams in Sweden have experienced damage caused by progressive internal erosion because the downstream filters were inadequate to prevent movement of finer grains from the core (Nilsson et al., 1999).

Remedial actions such as grouting have been performed to treat dam cores damaged by internal erosion, using cement-based grout or combined with chemical grouts, such as sodium silicate (Sjöström, 1999). According to published reports, at least five embankment dams in Sweden including Bastusel, Hällby, Suorva, Näs and Stenkullafors have been grouted with sodium silicate, mainly during the 1980s (Ekström et al., 2016; Göthlin, 2004). Sodium silica is characterized by its relatively high penetrability compared to other grouts available at the time. This property is suitable to treat finer soils, e.g. the impervious core.

Although improved penetration was achieved using sodium silicate in remedial grouting, there have been concerns regarding the internal stability and permanence of its end-product. Hansson (1999) stated in a report that sodium silicate would only be stable for one month. Its long-

(12)

term durability is thus questionable, indicating potential degradation of the grout and thereby decreased remedial effect with time. In turn, it implies a risk of reoccurred damage in the treated dam.

However, a systematic study of sodium silicate and its performance in embankment dams has not been performed. This thesis project is therefore carried out to compile information regarding the behavior of sodium silicate based grout and its potential degradation process under typical dam conditions; as well as to suggest monitoring and instrumentation to control the state of sodium silicate in a dam and the performance of a dam treated by sodium silicate grout.

1.1. Aims and Objectives

There are two aims with this thesis. The first is to investigate the possible long-term performance of sodium silicate grout in embankment dams by identifying potential degradation processes and identifying factors inducing degradation. The second aim is to suggest monitoring instrumentations to detect potentially changed dam functionality and changed grout condition as a consequence of the degradation.

The following objectives have been established to achieve the aims by answering the following questions:

• Under what conditions can sodium silicate become degraded and how?

• What changes in a dam treated with sodium silicate are expected to occur when the remedial effect is decreasing?

• How can potential negative changes of a dam related to decreased remedial effect be detected by monitoring?

1.2. Limitations

This thesis project aims to describe the general behavior of sodium silicate grout and the potential degradation of it based on available literature. In addition, suggested monitoring and instrumentations to control the performance of the dam with respect to the state of the remedial grouting

(13)

are based on already established and commercially available methods.

Hence, development of new instrumentation is not performed in this thesis.

1.3. Disposition

The disposition of the thesis is as follows:

Chapter 2: Contains a description of typical embankment dams in Sweden. The chapter also contains a description of the process of internal erosion, the common type of remedial grouting to treat dam cores, and general dam monitoring for controlling its performance related to internal erosion.

Chapter 3: In this chapter, the methodology conducted to achieve the stated aims is presented.

Chapter 4: A description of sodium silicate grout in general and its potential degradation processes are presented in this chapter.

Chapter 5: This chapter presents suggestions of monitoring and instrumentations to detect potential grout degradation.

Chapter 6: Contains a discussion of the susceptibility of sodium silica grout to degradation under typical dam conditions. Also, the overall representativeness as well as sensitivity of the suggested monitoring and instrumentations are discussed.

Chapter 7: Conclusions.

(14)

(15)

2. Embankment dams

This chapter first describes the main parts of a zoned embankment dam.

After that, the internal erosion phenomena and its development in an embankment dam is described. This is followed by a description of the general principle of remedial grouting performed to treat dam cores that is experiencing internal erosion. Lastly, existing monitoring related to internal erosion are presented.

2.1. Zoned embankment dams in general

The following description is mainly based on Vattenfall (1988) and RIDAS (2011) if no other sources are mentioned.

Embankment dams are constructed from compacted earth and/or rock material. The history of embankment dams can be dated to as early as 504 B.C. in Ceylon, Sri Lanka for irrigation purposes (U.S. Bureau of Reclamation, 2012). Sadd-El-Kafara in Egypt constructed around 2650 B.C. is an example of zoned embankment dam with long history (Mays, 2009). Until today, embankment dams are one of the most common types of dams in operation. Since the construction of embankment dams often uses locally available material close to the dam site, and the earth work can be carried out by a local contractor without advanced equipment, the construction can be both economically and constructively advantageous.

Normally, a zoned embankment dam contains four main zones as illustrated in Figure 2.1. There is an impervious dam core impounding the reservoir (1). The core is surrounded by one or more filters to prevent migration of fine material from the core (2-4). The core and filter(s) are surrounded by supporting fill to ensure the stability of the dam (5). Finally, there is a slope protection to protect the supporting fill from erosion due to

(16)

wave action, rainfall and potential flood debris etc. (10). This zone can be extended over the entire dam surface.

Figure 2.1: A typical zoned embankment dam founded on rock. Source:

Vattenfall (1988).

The impervious core

The impervious core in a zoned embankment dam is designed to prevent excessive seepage. The core material must be selected with regard to both permeability and workability. For a typical Swedish zoned dam, the impervious core usually consists of broadly and well-graded moraine.

Moraine with a high content of silt and sand is a suitable material to meet the required properties.

Before construction, the core material is subjected to modified Proctor tests to investigate the water content in relation to the optimal degree of compaction. During the construction, material segregation must be prevented. Generally, a wide core is advantageous to ensure the resistance to internal erosion. A thinner core is more sensitive to deficiencies during construction. However, the design of the core must also consider the material availability within a limited distance from the site. This affects the design, since for example less competent core material can be acceptable, if it is combined with a wider core and filters.

Filter

Seepage through an embankment dam will always occur due to the nature of it. However, there is a risk that fines from the core is transported in the downstream direction when seepage through the dam exceeds a certain velocity. If the internal erosion process is not controlled or treated, the

(17)

development can damage the water retaining function and eventually become a threat to the dam function. Filters and supporting fill at the downstream side are meant to prevent such material transport from the core.

Filters are placed between the core and the supporting fill, and both zones consist of material with various properties. The core consists of well- graded moraine, whereas the supporting fill commonly consists of rockfill.

Filter criteria must be met in transition between these two soil materials bordering it. Therefore, there are often at least two filters, a fine filter against the core and a coarse filter bordering the supporting fill.

Three main filter criteria are filtration ability, drainage ability and the ability to prevent material segregation. The purpose of filtration ability is to prevent the mitigation of fines in the core. To achieve this criterion, the voids between the particles in the filter must be small enough to prevent continuous transport of the core soil. At the same time, the filter should be sufficiently pervious to drain seepage and prevent pore pressure to build up. Thirdly, controlled placing of the filter is important in order to avoid material segregation, which otherwise could influence the filter capability and create layers of material with higher permeability. Fine filters are usually designed from sand and coarse filters from gravel. Due to environmental concerns, these materials are generally processed from crushed rock.

Supporting fill

The purpose of the supporting fill is to sustain the stability of the dam and to drain the seepage flow without erosion. To ensure this, coarse grained material or rock are used.

The safety factor against sliding and for various loading conditions is verified by stability calculations. For example, the loads to be considered are the weight from the dam itself before its first impounding, pore-water pressure due to seepage flow and due to extreme seepage flow respectively, and eventual rapid draw-down of the impounded water.

The drainage capacity must be dealt with as it influences the dam stability in several ways, e.g. a sudden increase in pore pressure may lead

(18)

to dam instability. Therefore, the risk of sudden excessive flow is also considered when designing the drainage capacity of the supporting fill.

Slope protection

Slope protection are generally placed on the surface of a dam. The material should consist mainly of rock material that are prone against surface erosion. The protection requirement is highest on the upstream side, where the erosion risk is high due to destructive wave action and ice load. The purpose of the downstream protection concerns erosion caused by rainfall and snow melting, and to increase the erosion resistance in case of overtopping of the dam.

2.2. Internal erosion

Internal erosion is the phenomenon when seepage through an embankment dam has become high enough to detach finer particles of the core or its foundation to cause mass movement downstream (ICOLD, 2017). Well-developed internal erosion without treatment will result in excessive flow. Turbid leakage containing eroded soil and sinkholes formed on the surface of the upstream crest are often signs of material transport due to internal erosion. Mechanisms leading to initiation of internal erosion are concentrated leak, backward erosion, suffusion and contact erosion. Once erosion has occurred, it will continue to detach and transport fines if the hydraulic force is not reduced and material migration is not limited by adequate downstream filters. Figure 2.2 illustrates the continuous material loss due to internal erosion.

(19)

Figure 2.2: Material loss of the core and backward erosion piping due to internal erosion. Source: Rönnqvist (2002).

As the internal erosion progresses, the erosive action will continue towards the source of seepage at the upstream side. Furthermore, erosion will continue throughout the dam core and a seepage tunnel within the dam is formed and enlarged if the dam soil is able to hold such a seepage tunnel. This type of damage is called piping, also illustrated in Figure 2.2.

As erosion and piping reaches the upstream filter, the filter material may gradually move down to fill the pipe formed tunnel and lead to a sinkhole at the upstream slope of the crest surface. Erosion caused by most of the initiation modes can give rise to piping (ICOLD, 2017).

There are three potential dam failure modes related to excessive flow caused by internal erosion. The first failure mode is due to extremely increased leakage to such an extent that the dam is no longer able to pass it through safely. As a result, rock materials at the toe of the dam can become unstable and continuous backward erosion could be initiated. In the second failure mode, pore pressure increases due to excessive leakage flow and it leads to decreased shear strength at the downstream slope.

Potentially, it can cause slope instability. The third failure mode is dam breach due to pipe formation and sinkholes, which can lead to loss of freeboard. Consequently, overtopping of the dam can develop. (RIDAS, 2011)

(20)

2.3. Remedial measures to repair the impervious core

Diaphragm wall, slurry wall, pile wall using sheet piles or secant piles and grouting are examples of remedial measures to increase or restore the impervious property of an embankment dam. Remedial grouting has mostly been performed during the 1980s and 1990s in Sweden to treat several dams experiencing internal erosion. The aim was both to seal the excessive leak and repair the core.

Grouting of dam cores is performed by injecting the grout(s) into boreholes drilled around the damaged area. Once the grout is injected and has penetrated the voids and/or seepage channels, it will harden to seal them. A combination of cement-based grout and chemical grout was relatively common at the time, aimed to seal the main leak with cement grout and seal the remaining part of the leak through the core with sodium silicate. This was because penetration of cement or cement-bentonite grout is insufficient for sealing the impervious core. Furthermore, the finish- product of cement grout is much stiffer than the soil to be treated, thus there is a potential risk of new seepage path between the grout and the soil.

Sodium silicate in the other hand is of higher penetrability and its finish- product can become a soft gel, suitable for repairing the core.

As an example, Lagerlund (2007) reported that one dam has been grouted after a sinkhole and sudden increase in leakage were observed.

Cement-based grout was injected to the outer rows of boreholes first in order to seal the leak, as well as to prevent potential leak of sodium silicate.

Sodium silicate was injected into the middle rows of the boreholes to repair the core after injection of cement. Totally, 42 m³cement-bentonite and 164 m³sodium silicate was injected respectively.

2.4. General surveillance and monitoring related to internal erosion

Dam surveillance is a vital part within dam safety work. The purpose of the surveillance is to control dam performance and evaluate dam safety by detecting and identifying signs that indicate changes in a dam, before a

(21)

damage occurs. Visual inspection and monitoring by instrumentations are two main parts of dam surveillance.

Common visual detections related to internal erosion are sinkholes in the dam crest or in the upstream slope and wet spots found in the downstream slope etc. Increased leakage and turbid leakage are two parameters of abnormality of the dam, since turbidity of leakage that contains eroded material is a result of material transport. Sinkhole is a clear indication that damage has occurred due to extensive internal erosion and piping. Such visual inspection provides qualitative information of the dam performance (DSIG, 2012).

Quantitative information of a damage, e.g. quantity and velocity of seepage in a dam can be achieved by monitoring with instruments.

Common instrumentations include measurements of pore-water pressure by piezometer, surface settlement by levelling, leakage monitoring by weirs and settlement inside a dam by settlement gauges (ICOLD, 1988).

Piezometers are often installed in the supporting fill, but sometimes they can even be installed in the core, as illustrated in Figure 2.3.

Figure 2.3: The potential location to install piezometers for dam monitoring.

The brown part illustrates the dam core.

Continuous measurements are in some cases required to achieve relatively accurate interpretation of monitoring data. For instance, quantity of leakage downstream of the dam must be measured regularly to minimize influence of e.g. seasonal changes such as rainfall and snow melt.

(22)

Detection of a damage

Visual inspection and monitoring are combined to assess performance of a dam and the quantity of a potential damage, e.g. seepage velocity through temperature measurements respective measurements of the amount of leakage. Once a damage has occurred and become detectable by visual inspection, for instance through a sinkhole, it means that the damage has already become extensive. Therefore, monitoring needs to be performed over time to detect early signs of a change or a potential damage.

Furthermore, measurement data needs to be interpreted accurately. Also, location and cause of a change or an abnormality needs to be addressed.

Once internal erosion has occurred, water level measurement and temperature measurement carried out with piezometers installed alongside the filters and/or the supporting fill downstream can help to determine the location of this damage and the amount of leakage.

Sometimes, several piezometers are required to be installed to achieve accurate result. Also, turbidity test is a way to determine whether a sudden increased leakage is actually caused by internal erosion and material loss of the core.

(23)

3. Methodology

This thesis project is divided into two parts. The first part is an evaluation of the performance of sodium silicate grout by a literature review. The second part aims to suggest suitable monitoring methods to follow up the performance of a dam grouted with sodium silicate. The evaluation of the performance of sodium silicate grout aims to identify potential grout degradation process and factors causing respective types of degradation.

Suggestions of dam monitoring are based on the potential influence on a dam when grout degrades. The overall workflow is illustrated in Figure 3.1.

Here, the treated dam does not refer to any real dam, but a theoretical embankment dam grouted with sodium silicate.

Figure 3.1: Workflow to conduct the evaluation and monitoring of grouted dam.

(24)

3.1. Evaluation of grout performance

The first step of the evaluation is a description of the background of sodium silicate grout including the common composition, the hardening process and chemical reactions of sodium silicate. The second step is to identify grout degradation processes and explain why it occurs. Definition of grout degradation in this context is changed grout behavior that would affect remediation of the treated dam negatively. Identification of factors that induce potential degradation is the third step. These factors are those that could be existing under conditions typical for an embankment dam.

3.2. Suggest monitoring and instrumentations

The second part of this project is to suggest how performance of dams repaired with sodium silicate can be monitored and evaluated. If sodium silicate grout will undergo degradation, it would potentially cause changes and even reoccurred damage in the treated dam when grout degradation reaches a certain degree.

Changes of a dam behavior due to decreased grouting efficiency can develop to a damage in several stages, from a gradually decreased remedial effect to a reoccurred damage in form of e.g. excessive flow. The development from initiation to continuation and finally progression leading to a damage is identified. Parameters indicating changed dam behavior at each of the damage development stage are also identified.

Monitoring and instrumentations to evaluate the dam function are suggested based on these identified parameters controlling dam behavior at respective damage development stage. The suggestion includes both direct methods to monitor the dam behavior itself and indirect methods to monitor potential grout degradation which could lead to changes in the grouted dam.

(25)

4. Sodium Silicate grout

Soil solidification and soil stabilization are two common applications of grouting with cement-based grouts and/or chemical grouts. Injection with sodium silicate has been common due to its high potential of good penetrability in fine soil and the possibility of grout setting control.

Nevertheless, its long-term performance has been uncertain which gives rise to concerns, especially to grouting work aimed for permanent support and remediation.

This chapter will describe grouting in general, properties of sodium silicate grout, observed uncertain performance and factors that could cause degradation of sodium silicate grout.

4.1. Grouting

The definition of grouting is the injection of a fluid material into a soil or rock formations that will harden to change their original physical characteristics (Karol, 2003). In practice, the change of the physical characteristics of a certain geological formation by grouting often refers to improvements of this formation, e.g. decrease of its permeability to gain improved waterproofing ability, solidify loose soil particles and enhance the soil strength.

Different aims of grouting

The following description is based on Karol (2003) if no other sources are mentioned.

Preventing or reducing water flow through a geological formation or soil can be achieved by decreasing the permeability. Tunnels under the groundwater table, as well as embankment and concrete dams are examples of structures that may experience different types of water inflow through both the structures themselves, their foundation and the

(26)

impervious dam cores. Water intrusion and leakage of an unacceptable extent can lead to problems affecting both stability and functionality of a structure negatively. Grouting is a measure to treat such a problem, by sealing the formation and the fissures where leakage is taking place. For sealing purpose, the grout should be selected with consideration of void size in relation to the penetrability of the grout. Furthermore, grouting is a relatively beneficial method if the fissures where leakage occurs is known (Bell, 2007).

For seepage control purpose, the grout is expected to be in constant contact with groundwater and/or reservoir water. Therefore, this grout must not be susceptible nor sensitive to hydraulic pressure. It should also be chemically stable against influence from the water and groundwater.

Typical grouting materials

The following description is based on Palmström and Stille (2010) if no other sources are mentioned.

In general, grouting materials are classified in two main groups, according to their physical properties: cement-based grouts and chemical grouts. The cement grout mix is assumed to behave as a Bingham fluid of suspension type. It means that the fluid contains particles, and penetrability of this fluid is partly affected by the particle size. Due to this fluid property, the penetration of cement-based grout is limited. This means that cement is suitable for soil fractions ranging from coarse sand to gravel. Bell (2007) stated that cement cannot enter fissures smaller than approximately 0.1 mm. The finished-product of cement grout is another limiting factor, since it is hard and behaves as a rigid body. Consequently, there is a risk of new seepage developing along the surface of the grouted area, if the grout is more rigid than the treated medium. Because of this concern, bentonite is sometimes added to the cement grout mixture in order to achieve a more elastic behavior. These two described phenomena may indicate that cement is less suitable for repair of finer soils, e.g. an impervious core of an embankment dam, which usually have a high content of silt.

(27)

Chemical grouts are often assumed to behave as Newtonian fluids and are characterized by solution type. Sodium silicate which often goes under the name “water glass” is a chemical grout. In Sweden, sodium silicate has been applied for remedial treatment on cores of several embankment dams due to internal erosion of the core caused by inadequate filter design. This remediation with sodium silicate was performed mainly during the 1980s and 1990s (Ekström et al., 2016).

Since the sodium silicate grout is a solution, it has better penetrability than the grouts of suspension type. The penetration of sodium silicate is mainly controlled by the viscosity, which can be controlled by adjusting the grout composition. The penetrability of four common grouts in relation to grain size is illustrated in Figure 4.1. The silicate-based grout can penetrate fine sand, but not into silt. It has however a notable better penetration that can successfully be used in fractions down to coarse sand.

Figure 4.1: Soil particle size limitations on grout penetration. Source: Bell (2007).

Usually, the sodium silicate solution and a reagent are mixed and injected into the soil, because hardening or gelation of sodium silicate requires reaction with the reagent to decrease the grout’s pH-value. A decrease in pH results in that silicate ions polymerize, and the solution turns into a gel. A chemical grout such as sodium silicate is associated with time-dependent flow properties. The time required for sodium silicate to harden can also be controlled through the choice of reagent and adjusting the grout composition.

(28)

In addition, grouting by a combination of cement and sodium silicate are not unusual. For example, remedial grouting performed to treat several dams in Sweden was based on a combination of these two grouts. The principle behind this type of grouting is that the cement-bentonite grout should seal the larger voids and at the same time prevent sodium silicate solution to leak away.

Grout requirements

The following description is based on Karol (2003) and Palmström and Stille (2010) if no other sources are stated.

Selection of a grout should consider both its workability and its grouting efficiency. Workability concerns the application aspects, i.e. the penetration of the grout as well as the setting time which the grout needs to harden (Göthlin, 2004). Grouting efficiency refers to the strength and durability of the grout, meaning that a grout should obtain enough internal strength and is durable enough to withstand influence from its surroundings (Krizek and Madden, 1985). Criteria for grouting can thus be categorized into either mechanical or chemical properties.

Mechanical properties

1. The penetrability of the grout should be high enough to penetrate the treated zone thoroughly. This criterion is directly related to the grouting workability. Viscosity is the determining factor affecting penetrability in case a grout in solution is injected. For sodium silicate grout, viscosity is controlled by the silicate concentration in the grout as well as the weight ratio between sodium and silicate in the grout.

2. Strength provided by the grout should be sufficiently high. When a grout has completely filled the soil voids and pores, it forms a continuous latticework which binds the grains together. By doing so, the grout will increase the shear strength and resistance against deformation of the soil grains. If the grout provides only low shear strength to the geological formation, there is risk of internal erosion at high water pressure and a high hydraulic gradient.

(29)

3. Permanence of a chemical grout should be sufficiently high against mechanical influence. For example, the grout can be influenced mechanically due to freezing-thawing cycles, and/or wetting- drying cycles that leads to mechanical deterioration. The grout is most sensitive when water which once contained in the grout no longer is bound to it chemically.

Chemical properties

1. The grout must be chemically stable to withstand chemical influence. An appropriate grout should be durable against the influence of the environment which a grout is injected into, e.g.

groundwater. If not, there will be risks that the grout becomes unstable and leaches due to reaction with groundwater or water.

2. The grout setting time should be under control. Controllable gelation process of sodium silicate has the impact on the final outcome and completion of grouting. Without control, the grouted formation could undergo uneven grouting and the result might differ from the desired outcome.

4.2. The sodium silicate solution

This section describes in general the composition of sodium silicate-based grout, the chemical reactions behind the hardening and gelation of it, as well as injection procedures of sodium silicate.

Composition and properties of sodium silicate

As its name indicates, sodium silicate is a mixture of sodium and silicate in a water solution. Commercially, this product is known as water glass and it is available in form of powder or solution. The chemical formula is 𝑛𝑆𝑖𝑂₂· 𝑁𝑎₂𝑂, where n is the ratio that can refer to both molar and weight ratio between silicate and sodium, since these two values are almost identical.

For grouting purpose, the ratio often varies between 3 to 4, but never exceeds 4 (Karol, 2003; Tallard and Caron, 1977; Hamouda and Amiri, 2014). Depending on the applied reagent, the finished-product becomes either a soft gel or a hard gel. In the case of grouting the impervious core of embankment dams, a soft finished-product is more advantageous.

(30)

Sodium silicate is alkaline, because silicate is only weakly acidic, while sodium is strongly alkaline. When the grout reacts with a reagent which is often an acid or a metal salt, the solution will harden. Adding of such a reagent will decrease the pH by neutralizing the sodium ion in the sodium silicate solution, thereby releasing free silicate ions. At lower pH, these free silicate ions are polymerized into a longer chain of polysilicic ions, which turns into a gel by hardening (Yonekura and Miwa, 1993).

Sodium silicate is a flexible material and it is possible to adjust the composition in order to optimize its grouting ability such as viscosity and strength. The silicate concentration controls both density and viscosity of the solution. The higher the content, the higher its viscosity becomes. At higher concentrations, molecules are denser, resulting in higher viscosity (Weldes and Lange, 1969).

Compared to cement, chemical grouts have limited strength. The strength of the end-product of sodium silicate is positively related to the concentration of both silicate and reagent in the solution. The more sodium that is neutralized by the reagent, the higher the strength becomes.

Consequently, a higher silicate content can lead to higher strength of the end-product. However, there is a risk that the strength will be low by lowering the silicate content in order to achieve low viscosity and thereby a high penetrability (Zheng, 2000).

Setting time and neutralization of sodium silicate

Setting time is the time required for sodium silicate to harden and stabilize.

This property is influenced by the grout composition in several ways. Also, permanence of the finished grout product can be influenced by the reactivity, i.e. the degree of sodium neutralization when sodium silicate has undergone gelation.

The most accepted theory of grout gelation in the references is that sodium silicate hardens through pH decrease. First, free silicic ions are discharged from the grout solution by neutralization of sodium ions. When acids or salts based reagent is applied, the pH of the grout in solution decreases. At low or neutral pH, silicate ions will polymerize to form silica.

(31)

(Tallard and Caron, 1977; Lagerblad et al., 1995; Yonekura and Miwa, 1993)

Since the gelification of silicate sodium is achieved through neutralizing the sodium ions, the degree of neutralization is an important parameter affecting performance of the grout after injection. It is desirable to have as low sodium content as possible, thereby a high silicate/sodium molar ratio.

However, a ratio above 4 will lead to an unstable end-product. With respect to this, the molecular upper limit of sodium silicate should not exceed 4 (Tallard and Caron, 1977). The degree of neutralization decides the strength and durability. The higher the neutralization, the stronger the gel and the treated soil becomes. A relatively high degree of neutralization can be achieved with a high amount of reagent.

The reagent, which is also called reactant, activator or coagulant in the literatures has an impact on the gel setting time as well as on the degree of sodium neutralization. Generally, the higher reagent content, the faster is the setting time. For a given reagent, higher silicate concentration will also lead to faster setting time. Further, Yonekura and Miwa (1993) categorized the reagents into inorganic reactant, inorganic gas reactant and organic reactant with regard to both the setting time and the degree of sodium ion neutralization.

1. Inorganic reagents: the gelation process is fastest with inorganic reagent, such as calcium chloride. However, after the gelation, the remaining of sodium ions is high within the gel structure, meaning a risk of instability.

2. Organic gas reagents: the gelation process is fast. Some sodium ions remain in the gel network after gelation.

3. Organic reagents: the setting time is longest with organic reagents.

However, little sodium ions are remained afterwards.

Injection of sodium silicate

Injection of sodium silicate is performed with either the one-step or the two-step method. One-step injection means that the grout and reagent are mixed first and then injected together. This method is more frequently performed today than the two-step method, and the mixture is usually prepared just before injection on the site. The grout will harden and form

(32)

a gel with relative slow rate because the gelation initiation is delayed (US Army Corps of Engineers, 1995). The slower rate of gel formation would allow more control of grout penetration, leading to more even and thorough penetration.

Two-step process, also referred as the Joosten-method is performed by first injecting the sodium silicate, and at the second step injecting the reagent into the same zone to form a gel and stabilize the grout. A salt reactant, often calcium chloride in solution form is used. This process enhances the soil strength most but is the most expensive approach (US Army Corps of Engineers, 1995; Tallard and Caron, 1977). The chemical reaction with calcium chloride is shown in Eq. 1.

𝑁𝑎2𝑂 ∙ 𝑛𝑆𝑖𝑂2+ 𝐶𝑎𝐶𝑙2+ 𝐻2𝑂 → 𝐶𝑎(𝑂𝐻)2+ 𝑛𝑆𝑖𝑂2+ 2𝑁𝑎𝐶𝑙 [Eq. 1]

The two-step process using calcium chloride results in an instantaneous harden reaction. This is advantageous when aiming to seal a sudden leakage and stop the flow. But too rapid gelation can lead to less thorough penetration of the grout, thus it can result in uneven gelation which limit the remedial effect.

4.3. Degradation of sodium silicate

Permanence of sodium silicate is a concern since the grout gel can undergo different types of degradation. Potentially, it can lead to worsened remediation. Göthlin (2004) reviewed monitoring data of measured sodium concentration in a dam where both the core and its foundation had been grouted with sodium silicate. The leakage from the foundation showed much higher sodium content compared to sodium concentration measured at the downstream weir. Göthlin concluded that this was an indication that the grout had leached from the dam foundation. Göthlin’s analysis based on field data has thus indicated an uncertain behavior and questionable long-term permanence of sodium silicate in an embankment dam.

(33)

Other reviewed literatures that described results from laboratory tests have shown that the soil grouted with sodium silicate generally experiences increased permeability with time. Avcı (2017) measured permeability of soil specimens grouted with sodium silicate under 120 days in the laboratory. All specimens were observed to experience an increase in permeability. Krizek and Madden (1985) also measured permeability of several sand specimens treated with sodium silicate for 600 days in the lab.

Continuous increased permeability was also observed in this experiment.

Both studies concluded that the increased permeability was related to grout degradation.

The gradually increased permeability after grouting with sodium silicate is mainly caused by syneresis or leaching of the grout, or a combination of both. Syneresis can induce gel shrinkage, and sodium silicate leaches when it has dissolved or partly dissolved. Table 1 has summarized the identified factors that could give rise to these two types of grout degradation according to the references (Avci, 2017; Einstein and Schnitter, 1970; Karol, 2003; Krizek and Madden, 1985; Lagerblad et al., 1995; Littlejohn et al., 1997; Yonekura and Kaga, 1992).

A decrease in strength is another consequence of grout degradation, mainly due to grout leaching. Strength of several soil specimens grouted with sodium silicate were decreasing continuously under 1000 days, observed by Yonekura and Kaga (1992) in the laboratory. This was because sodium silicate leached, therefore it was no longer able to support the soil.

This decrease was significant, because in this experiment, the strength of the soil measured in the end of this experiment was only half of its initial value.

Syneresis induced shrinkage

Syneresis is the phenomenon related to silicate gel when water once contained in the grout gel leaks, leading to decreased gel volume and shrinkage of the gel. Syneresis of silicate gel will occur because the electric charge of the silicate molecules decreases, resulting in repulsion between different parts in the molecular structure. This means either shrinkage, or that the pressure of the grout decrease (Sjöblom, 1995). Furthermore,

(34)

Table 1: Factors that can induce syneresis or leaching of the sodium silicate.

Degradation Factors causing grout degradation

Syneresis induced gel shrinkage

A certain range of silicate content can cause higher syneresis.

The type of applied reagent to harden the grout. Certain reagent can lead to much higher syneresis than others.

Cement or concrete that contains calcium ion may induce high syneresis.

Leaching of grout when flowing water erodes the grout, or when the gel dissolves.

High pH can cause grout dissolution.

Insufficient curing of the grout can lead to lower grout strength, thus it is more susceptible to erosion caused by flowing water.

Insufficient neutralization of sodium can cause dissolution.

Being in contact with water or groundwater can cause dissolution.

syneresis can also lead to higher susceptibility of the grouted soil to erosion caused by flowing water (Krizek and Madden, 1985).

Avcı (2017) and Littlejohn et al. (1997) found that syneresis of the pure grout gel can be as high as 80% and 60% respectively. Both studies suggest that syneresis is related to the silicate content in the grout. Theoretically, sodium silicate can also experience shrinkage when it is in contact with water and in contact with concrete.

Influence of syneresis or gel shrinkage on the grouted soil is also dependent on grain size of the treated soil. Generally, syneresis has larger influence on coarser soil but less effect on fine soils. It means that sometimes syneresis would not develop sufficiently to induce higher soil permeability. This is because the particles of fine soil are placed closely,

(35)

therefore they can provide more support to prevent shrinkage or loss of the grouting material (Einstein and Schnitter, 1970; Karol, 2003).

Syneresis of a silicate gel will always occur, but the degree of it can be controlled by several factors. The factors that may lead to a high syneresis rate are identified to be the silicate concentration of the grout, the type of reagent used to harden the grout solution and the influence of calcium ion from e.g. cement grout or concrete.

Grout composition and silicate content

Avcı (2017) studied the permeability of soils grouted with sodium silicate under 720 days. These studies showed that the syneresis would increase with increased silicate concentration. But when syneresis reached a peak, the rate decreased. Figure 4.2 shows the relationship between syneresis and silicate concentration in Avcı’s study, where the highest syneresis rate was at 80%.

Figure 4.2: The relationship between syneresis and the silicate content in the grout. Source: Avcı (2017).

Littlejohn et al. (1997) provided the same information, that there was a relationship between the syneresis following the same trend as Avcı’s

(36)

study, as shown in Figure 4.3. The highest syneresis observed in this study was at 60%. The degree of syneresis is defined as the ratio between the volume water leaking from the grout gel and the initial grout volume in both studies.

b)

Figure 4.3: Degree of syneresis due to silicate concentration. Source:

Littlejohn et al. (1997).

A study by Yonekura and Kaga (1992) led to the same conclusion. A15 and A20 shown in Figure 4.4 are sodium silicate grouts, but A15 contained less silicate than A20. A20 showed to experience higher syneresis than A15 for all 1000 days during the observation. After 1000 days, syneresis was measured to be around 4% and 6% respectively as shown in Figure 4.3.

This value of syneresis seems to be much lower compared to Avcı (2017) and Littlejohn et al. (1997), but the degree of syneresis was defined differently here. The definition of syneresis was the ratio between weight of leaked water and the weight of the initial grout gel. Yonekura and Kaga’s study also showed that shrinkage developed with increasing rate under the first 100 days of the observation. After the first 100 days, syneresis rate ceased and became relatively constant.

(37)

Figure 4.4: Syneresis of four silicate based grouts. A15: sodium silicate of lower silicate content. A20: sodium silicate of higher silicate content. CH: silica

sol. CSN: colloidal silica. Source: Yonekura and Kaga (1992).

These three independent studies have suggested that syneresis has a wide range. In different labs, syneresis was observed to range from 4% up to 80%. Thereby, shrinkage may not always be significant. Since syneresis rate is also shown to be related to the silicate content, it is theoretically possible to regulate syneresis by adjusting the grout composition.

Type of reagent

The second factor leading to high syneresis is the reagent used to harden sodium silicate. Krizek and Madden (1985) measured the syneresis of five sodium silicate grouts under 21 days, where different reagents were used for respective grout sample. The sample using Terraset as the reagent was observed to experience much higher syneresis rate than the other four samples. Terraset had also led to continuously increased syneresis rate, whereas syneresis of the other four samples already stabilized under these 21 days. See Figure 4.5, where Terraset is expressed as a line marked with the black triangles. At the 21^st day of this experiment, syneresis of the

(38)

Terraset sample was observed to be above 30%, whereas the other four samples experienced syneresis around or below 10%. One conclusion is that certain types of reagent can lead to much higher syneresis.

This finding suggests that when sodium silicate will be used but syneresis might be a concern, the influence of the reagent to be applied on syneresis should be examined. In Sweden, the commercial reagent Dynagrout was commonly used in relation to remedial grouting of dams.

How this reagent would affect the syneresis and the grout can be important to examine in future studies.

Figure 4.5: Syneresis in relation to different reagents. Source: Krizek and Madden (1985).

(39)

Influence of calcium ion

Lagerblad et al. (1995) observed in the lab that when sodium silicate gel and concrete were placed in water together, the grout gel became enriched with calcium ions and sodium ions had leached. Shrinkage of the gel had also occurred. One reason was because sodium ions in the grout gel was replaced by calcium ions from the concrete. It resulted in a calcium silicate gel. However, calcium silicate gel is more polymerized than sodium silicate gel. This could have resulted that the calcium gel is not able to hold as much water as the sodium gel, thereby the grout gel can shrink once the sodium gel becomes a calcium gel due to the concrete (Lagerblad et al., 1995).

Replacement of sodium ions by calcium ions occurs due to ion exchange. Calcium silicate is thermodynamically more stable than sodium silicate since the electrostatic energy between calcium and oxygen atoms in a silicate silanol group is higher than that of sodium and oxygen.

Therefore, the sodium silicate group is more susceptible to dissociation than the calcium silicate group (Wang and Gillott, 1993; Lagerblad et al., 1995).

Grout leaching caused by gel dissolution and erosion by flowing water

Lagerblad et al. (1995) observed that the volume of the grout decreases not only because of syneresis, but also dissolution of the grout gel, especially when the gel is in constant contact with water. Furthermore, the grout gel can be eroded when it is subjected to a high hydraulic gradient. The degree of erosion is controlled by the gel strength, and the stronger the gel is, the less it is affected by the seepage. Stronger grout gel can be achieved with longer curing time.

Influence of the pH

The hardening of sodium silicate is achieved by decreasing its pH by applying a reagent. This means a potential risk that instability and even dissolution can occur when the grout gel experiences high pH. The increased pH can lead to depolymerization of the grout, potentially resulting in an unstable gel behavior.

(40)

Lagerblad et al. (1995) found that two sodium silicate grout samples lost continuously weight and became partially dissolved under 49 days during the experiment, when the grout sample was placed in water together with concrete. pH of the water increased gradually from 10 to 11.5 during the experiment. Lagerblad et al. (1995) explained that the dissolution was a result of increased pH due to the concrete. This experiment also showed that sodium ions had leached from the grout gel. The gel had in the other hand gained more calcium. It indicated that calcium from the concrete would form a calcium silicate gel. Potentially, syneresis and shrinkage might have occurred.

The grout gel will gain higher strength at lower pH for a certain silicate content. Hamouda and Amiri (2014) found in an experiment that sodium silicate gel gained a maximum strength around 3000 Pa when it was experiencing a pH at 10.10. The maximum strength was only 700 Pa for another grout sample when the pH increased to 10.70.

Consequently, the long-term permanence of the silicate grout can be questionable when the grouted area has been treated with both the cement and the sodium silicate grout. Since cement is an alkaline product which also contains calcium, there is a risk of both dissolution and shrinkage of the grout gel. At higher pH, the sodium silicate can become more susceptible to e.g. a hydraulic gradient and seepage due to the potentially decreased grout strength.

Gel strength and the curing time

Strength of the grout and the treated soil against a hydraulic gradient is related to the curing time, which is the time the sodium silicate being grouted in the soil. In the short-term, the longer a soil is grouted, the higher strength is achieved to withstand a certain hydraulic pressure. Krizek and Madden (1985) performed an experiment where five sodium silicate grouted soil specimens were all experiencing a hydraulic gradient of 100.

Curing time of each specimen was ranged from 5 minutes to 6 hours. The specimen allowed to cure for 6 hours could resist this hydraulic gradient, whereas the specimens cured only for 5 and 10 minutes respectively were eroded totally by this hydraulic gradient. The remaining two grout

(41)

specimens that had been cured for 30 minutes and 1 hour respectively, both experienced partial erosion at this hydraulic gradient.

However, the sodium silicate grout in an embankment dam can be expected to undergo relatively short curing time because of the constant seepage. Thus, in the beginning of the grouting, the strength can be relatively low and erosion of the grout may occur due to seepage through the treated area.

There is another risk regarding the grout efficiency related to the flowing water. The grout solution used to treat several dams in Sweden is fluent and requires relatively long setting time to harden (Najder, 2012).

This suggests a risk that grout could have leaked before it could harden to seal the damaged area.

Insufficient degree of sodium neutralization

There is often some sodium remaining in the grout gel structure, despite the neutralization with the reagent. The residue of sodium within the gel structure can lead to dissolution of the grout gel by depolymerization.

Consequently, it will result in a weaker gel, or even a breakdown of it.

(Yonekura and Miwa, 1993)

Yonekura and Kaga (1992) observed the strength of four sand specimens grouted with both sodium silicate and two other chemical grouts under 1000 days. Here, A15 and A20 were the two sodium silicate grouts, but A20 contained higher amount of silicate and the reagent. The two other studied grouts were silica sol (CH) and colloidal silica (CSN). The measurement under 1000 days showed that A20 and A15 experiencing continuously decreased strength. In the end of this observation, the strength had decreased to 1 MPa, whereas initial strength for both A15 and A20 was around 2 MPa. On the other hand, soil specimens grouted with silica sol respective colloidal gained higher strength with time.

One difference between sodium silicate and the two other chemical grouts in Yonekura and Kaga’s study was that only the former one contained sodium. This was considered to be the main reason leading to the decrease in strength for soils grouted with sodium silicate. The remaining sodium due to insufficient sodium neutralization gave rise to the

(42)

depolymerization of the grout gel, in turn resulting in leaching of the grout gel.

The degree of grout leaching was recorded by measuring the silicate content in the water, which the grouts were being exposed to. Figure 4.6 shows leaching of all four studied soil specimens. In the figure, it can be seen that sodium silicate had a leaching above 40%, whereas silica sol and colloidal silica had 2% and 1% respectively at the 1000^th day of this experiment.

Figure 4.6: Silicate leaching from four types of grouts under 1000 days. A15:

sodium silicate of lower concentration. A20: sodium silicate of higher concentration. CH: silica sol. CSN: colloidal silica. Source: Yonekura and Kaga

(1992).

To conclude, sodium silicate will be leached with time from the grouted soil due to residue of sodium after the gelation. Since neutralization of all the sodium is hard to reach, there will always be some sodium remaining in the grout gel. However, according to Yonekura and Miwa (1993), the organic reagents would neutralize 80%-90% of sodium in the grout (see section 4.2.2), which is the highest neutralization rate compared to the

(43)

other types of reagents. This suggests that the risk of sodium residue and leaching of the grout gel might be less by applying an organic reagent.

Instability due to flowing water and groundwater

A seepage through a soil grouted with sodium silicate can cause erosion if the hydraulic gradient is sufficiently high. Lagerblad et al. (1995) found that when a specimen of pure grout gel was exposed to flowing water, the chemical test showed that the salts in the grout had leached. After 7 days of exposure to the flowing water, the weight of this grout sample decreased from 73.4 g to 17.0 g, indicating that a large part of the grout sample had been eroded.

Since glyoxal was the reagent used in this experiment, the instability of glyoxal is another potential factor leading to grout dissolution because glyoxal is soluble in contact with groundwater. Glycolic acid is formed from glyoxal to harden the grout when the glyoxal reacts with sodium silicate.

However, glycolic acid is soluble in water due to e.g. diffusion, meaning that the silicate gel containing glycolic acid also becomes soluble (Sjöblom, 1995)

To conclude, dissolution of the sodium silicate gel due to water is complex. The grout gel might be eroded by flowing water, and the reagent can be unstable when exposed to water. However, this reviewed study by Lagerblad et al. (1995) and Sjöblom (1995) mainly provided information for a case when using glyoxal as the reagent, but in Sweden Dynagrout that consisted of sodium aluminate was the reagent used to treat several dams.

One reason was that this type of reagent would result in a soft gel which was considered to be suitable to repair the impervious core. However, how other types of reagents e.g. Dynagrout may be affected by groundwater remains to be examined.

Influence of grout degradation and the soil grain size

Pure grout gel can undergo degradation in form of syneresis and leaching due to dissolution or erosion of the grout gel. However, fine soils would be less affected from syneresis or grout leaching compared to the coarse soils (Einstein and Schnitter 1970; Littlejohn et al. 1997; Karol, 2003).

(44)

Einstein and Schnitter (1970) observed three sands specimens grouted with sodium silicate under 5 months, whereas these specimens were ranged from fine to coarse soils. The result showed that the specimen of the finest soil experienced least grout leaching compared to the other two specimens. Also, the permeability of this soil sample increased from 10^-7 cm/s to 5·10^-6cm/s, which was not considered to be a significant increase.

It was concluded that influence of leaching to some extent is a function of the grain size and the size of the pores, since finer and tighter placed soil particles could give more support to the grout to prevent it from leaching.

(Einstein and Schnitter, 1970)

Figure 4.7 shows syneresis presented by Littlejohn et al. (1997) of both a pure grout gel as well as three treated soils ranging from gravel to sand.

The pure grout gel experience highest degree of syneresis at 70%, whereas the sand sample experienced lowest degree of syneresis at 3%. The difference in syneresis between the pure grout gel and a grouted soil sample is thus large.

Figure 4.7: Syneresis in relation to soil grain size. Source: Littlejohn et al.

(1997), derived from Caron (1975).

Avci (2017) and Littlejohn et al. (1997) have observed and suggested that high syneresis of a pure grout gel could occur, as described in the

(45)

earlier section. But when analyzing the effect of syneresis or leaching on a grouted soil, the soil grain size should also be considered to examine the potential influence or decreased efficiency on this treated soil.

(46)

(47)

5. Monitoring to control the performance of the dam

The main purpose of remedial grouting in a damaged dam is usually to repair the core, ensure and restore its intended permeability in order to stop excessive leakage flow. If sufficient grouting efficiency is not achieved or the grouting efficiency decreases when sodium silicate grouted in the dam has degraded, it could lead to a decreased remedial effect, meaning a risk of reoccurred damage.

Damage caused by decreased grouting efficiency occurs and develops mainly in three stages as shown in Figure 5.1. The first stage is instability of the sodium silicate grout occurring in form of dissolution, leaching and syneresis of the sodium silicate grout. This will gradually result in larger voids and cavities in the grouted area. At the second stage, the larger voids and cavities will lead to higher permeability in the treated core as the grout continues to degrade, which allows more seepage passing through the remediated area. At the final stage, a damage in form of excessive flow and sudden increased in leakage will occur, caused by the loss of the remedial effect.

Monitoring can be used to identify the phenomena described above, and to conclude whether an unwanted change is taking place in a dam and determine the stage of the change. This chapter describes the methods suitable to monitor grout efficiency and its long-term performance.

(48)

Figure 5.1: Different stages of damage development related to degradation of sodium silicate and possible indicators to detect degradation.

5.1. Analysis of ion content in leakage water

The first stage of a decreasing grout efficiency is leaching of sodium silicate caused by erosion or dissolution. Excessive leaching of grout will lead to an increased permeability, and therefore more water is able to seep through or adjacent to the treated part of the core. Potential leaching and dissolution of sodium silicate can be detected by chemical analysis, e.g.

measuring the ion content in the leakage water, since leaching of sodium silicate will lead to a higher concentration of, for instance, sodium ions that are being released from the grout as it dissolves. Where leakage monitoring is intended to specifically identify potential changes in a grouted area, the measurement of the ion content should be a long-term commitment to obtain reference values.

The ion content can be measured in leakage water collected in the downstream drainage system and weir. The concentration of specific ions, in this case sodium respective silicon ions can be quantified by using ion-

Evaluation of long-term performance of sodium silicate grouted in embankment dams