The hydrojacking and hydrofracturing behaviour of discontinuities in rock is analysed by evaluating water pressure (Lugeon) tests carried out for investigation of grouting at 23 dam sites in Spain. Comparing the results with those obtained in a further 30 dam projects allowed Friedrich-Karl Ewert to assess the deformation which occurs in most rock types
DURING the injection of water or grouting materials into rock, open paths are dilated (hydrojacking) or closed discontinuities are pressed open (hydrofracturing) once the grouting pressure exceeds the resistance of the rock surrounding the borehole. Both occurrences are momentous, particularly in the sphere of dam construction: a larger permeability of the rock mass is expected often causing too extensive grouting measures.
Cases of hydrofracturing occur frequently, largely due to the practice of increasing the grouting pressure with depth. This causes hydrofracturing and backfilling of latent discontinuities particularly in deeper sections. Such grouting wastes time and money, which could be avoided if it were considered that latent discontinuities in many rock types are hydrofractured at rather low pressures already, even in greater depths.
Approximately 25 years ago the author emphasised the negative consequences of hydrofracturing for the technical and economical success of grouting programmes. The examples available at that time only comprised a few rock types. Given the fact that this complexity is not duly considered even today, it is time to continue the discussion. Since then a multitude of further projects have became available providing important knowledge on the hydrofracturing behaviour of almost all rock types. It confirms the former conclusions and proves their universal validity.
Data sets are now available from permeability tests and grouting carried out worldwide in approximately 80 projects. Some of the sets are fragmentary, others are complete. The treasure of data is extensive and the scope required for a completely comprehensive evaluation of all these projects reaches far beyond the available space. Hence, the number of projects considered had to be restricted, nevertheless the selection comprises all rock types. The evaluation focuses on the dilation of fissures and, above all, on the fracturing behaviour of la-tent discontinuities; all other aspects necessarily remain disregarded.
The data considered were obtained during testing and grouting programmes executed at 23 dam sites in Spain. They were collected during a three-month stay at the Escuela Técnica Superior de Ingenieros de Caminos, Canales y Puertos de la Universidad de Cantabria in Santander in 1994 financed by the Spanish government.
The presentation of all examples, the discussion of all complexities involved and the deduction of the conclusions needed much more space than available here. As a result, the unabridged paper has been published in IWP&DC’s sister publication, Dam Engineering [A]. This abridged version includes only Figures 1 and 25 out of the 42 figures and five tables of the original. The designation and numbering of the figures and chapters used in the original is kept. If further figures are referred to they are quoted in Italic. The abridgment was not done linearly: while only three out of the 22 projects are kept, more space is given to the discussions and conclusions.
For the past 70 years, the permeability of rock masses has usually been determined by means of ‘water pressure tests’ (WPT) introduced by Maurice Lugeon. The amount of water pressed into isolated sections of boreholes serves as a measure describing the absorption capacity of the rock and, under certain suppositions, the permeability [4, 7/2, 8, 9]. Lugeon-values (or WPT-values) are given in Lugeon-units (LU); 1 LU is defined as ‘1 liter per meter and minute at a reference pressure of 10 bar’. That quantity became the worldwide accepted Lugeon-criterion. Lugeon concluded that rock would be practically tight if at this pressure less water is absorbed than 1 LU. It is still largely in use although many cases proved it by far too stringent. Individually defined reference pressures are applied to meet particular project conditions. WPT-units based on unequal reference pressures are not comparable, of course.
Hydrojacking, i.e. the dilation (or widening) of existing paths begins in rock types of little strength at pressures smaller than 10 bar. It causes an over-proportionate increase of the water take before the reference pressure is reached. Under such conditions the Lugeon-value referring to 10 bar pretends a larger permeability compared to the original one. This discrepancy impairs our assessment. In rock types of great strength the dilation of existing paths begins at pressures above 10 bar, thus the absorption rate at the reference pressure reflects still the original permeability.
Hydrofracturing splits latent discontinuities producing a fissure as soon as the testing or grouting pressure reaches the ‘critical pressure’, different from case to case. It causes a much larger effect of pretence: the latent planes absorb no water during the low-pressure steps but large amounts after fracturing. Figure 1 exemplifies both cases. It is obvious that the wrong assessment of the original permeability has considerable consequences.
Hydrojacking and hydrofracturing are particularly effective in grouting work where even higher pressures are usually applied. Hydrojacking and backfilling of dilated joints by grouting yields a positive effect unless too much grout is needed to reach the maximum grouting pressure: the grout travels farther, more excess water is pressed out and a better adhesion along the contact between groutstone and rock is achieved. By contrast, hydrofracturing produces too large grout takes without yielding a better sealing: latent planes are pressured open while smaller paths remain ungrouted. Fracturing is particularly momentous if the grouting pressure increases with depth since it exceeds the critical pressure even in rather strong rock types. Therefore a close interrelation exists between the fracturing behaviour and the optimal grouting pressure.
Basis of examination
Fracturing is recognised by comparing the results of testing and grouting. The process can be explored particularly well if a section of a core drilling is tested and grouted: the WPT reveals the critical pressure, the groutstone layer in connection with the unweathered rock bordering that layer confirms the intercalation along a fractured plane. The examples available, including that shown in Figure 1, show that dilation and fracturing processes are expressed by P/Q-diagrams.
Classification of P/Q-diagrams
WPT’s yield various types of P/Q-diagrams. They reflect different courses of the tests caused by individual geotechnical conditions. P/Q-diagrams have been repeatedly classified. The author presented such a system at first in 1979; during the following years it has been adapted to the growing knowledge. For this analysis the classification system established in 1992 and successfully applied since then is used [4, 7/2, 8, 9] (Figure 5).
Factors ruling hydrofracturing and hydrojacking
The rock bond, the state of the discontinuities and the strength properties control whether hydrojacking or hydrofracturing occur:
• In a surface near seam weathering reduced the strength along the discontinuities and disconnected the rock bond, i.e. the rock units bordered by discontinuities are separated from each other. The separation increases with growing disintegration making the rock units movable against each other. The test – or grouting – pressure initiating a heaving of the rock units above a largely disconnected plane depend on the weight of the overlying rock causing the overburden pressure. That pressure increases with depth (Figure 6A).
• Weathering decreases with depth and the rock bond becomes connected. The overburden pressure does not determine the critical pressure any more while the strength of the rock becomes increasingly effective controlling the dilation of existing paths or the fracturing of latent planes. Different courses occur depending on the types of discontinuities, the degree of separation and the strength of the rock around the section of the borehole being tested or grouted (Figures 6B, 6C, 6D). Hydrojacking depends on the compressibility of the rock bordering the paths. Hydrofracturing depends on various factors: the tensile strength, the compressibility of the rock and the relation between the orientation of the borehole and the orientation of the most susceptible discontinuities. The various types of discontinuities have a different susceptibility to fracturing: bedded rocks and slates are particularly prone but also metamorphic rocks possessing a pronounced parallel texture of their crystals.
A total of 2272 WPT’s carried out in 448 boreholes of 22 projects has been evaluated yielding P/Q-diagrams. They have been classified using the system given in Figure 5. Only four of these projects are presented here serving as examples to demonstrate particular features: Almendra, Cernadilla, Santa Eulalia and Las Portas. For P/Q- Types 3, 5 and 6a (hydrofracturing, hydrojacking and washing out) the relations between critical pressures and depths are presented in point diagrams.
The project comprises a combined gravity and arch dam, 180m in height. The subsoil consists of massive granite, often slightly metamorphic with parallel orientation of crystals. The limited space does not permit the evaluation of the many thousand WPT’s schematically executed in the groutholes; this contribution considers the 137 WPT’s carried out in 11 investigation holes.
Approximately half of the tests recorded a proportionate relationship between pressure and water take (P/Q-Type 4). Almost one fourth of the tests showed dilation or fracturing (P/Q-Types 3 and 5). Some of those occurring up to a depth of approximately 20m were caused by heaving at pressures between 1 and 5 bar already. Further below fracturing took place due to strength with critical pressures ranging between 5 and 13.5 bar. Their scattering is considerable, it does not increase with depth. One fifth of the tests absorbed no water though high tests pressures up to 25 bar were applied.
The gravity dam, 45m in height, was founded on Gneiss. This analysis considers 86 WPT’s carried out in 11 investigation holes up to 63.5m in depth. The essential results are summarised as follows:
• In several zones intense weathering reaches to a depth of approximately 30m, thus permeable rock was encountered in about 63% of all tests – P/Q-Type 4. Large water takes often kept the achievable test pressures small – Pmax: = 1 bar, Qmax = 100 LU.
• In some of the permeable test sections the paths were dilated at pressures between 3 and 5 bar – P/Q-Type 5.
• Besides exist impermeable zones: almost one fourth of the tests were tight and did not fracture, even at pressures between 9 and 12 bar – P/Q-Type 1.
• The impervious rock included test sections cracking at pressures between 6 and 10 bar – P/Q-Type 1.
• Up to a depth of 10m the critical pressures are controlled by overburden pressure, further below by strength; there they are scattering but do not increase with depth.
The data come from 321 WPT’s systematically carried out in 72 holes before grouting. They were drilled from the gallery of the 75m high arch dam into metamorphic slates down to a depth of 32m. The test pressures ranged between 9 and 11 bar. With the exception of a few test sections the rock was tight or just a little permeable.
The most important features are as follows:
• A considerable portion of the test sections remained impermeable up the maximum pressure of 11 bar (P/Q-Type 1) while in other zones hydrofracturing took place at critical pressures between 6.5 and 8 bar (P/Q-Type 3), thus, latent discontinuities susceptible to fracturing are evidently rare.
• About one third of the test sections was permeable and remained unaltered during higher-pressure steps (P/Q-Type 4).
• Several further tests also encountered permeable rock but were dilated during the higher-pressure steps (P/Q-Type 5).
• The range of scattering of the critical pressures causing dilation or fracturing is rather small and does not increase with depth.
The arch dam is approximately 120m in height and was founded on schists and quartzite. The available information comprise the detailed grout takes of the entire curtain. Here only the few WPT’s carried out in two investigation holes are considered because these tests comprised an extraordinarily great number of pressure steps with maximum pressures up to 23 bar, instead of the few ones usually applied. Such sequences disclose much better how the relationship evolves between pressure and water takes. Due to the high pressures mostly applied, hydrofracturing prevails (60 % of P/Q-Type 3) and cases of hydrojacking are still noticeable (20 % of P/Q-Type 5).
Within the scope of this analysis the most important features of this project are the detailed P/Q-diagrams resulting from the many pressures steps, exemplarily displayed in Figure 25: hydrofracturing is obviously a gradual process, comparable to wedging off a trunk. In fact, this is not really surprising and well comprehensible but has not been experienced and presented so clearly.
Results of WPT’s carried out in other countries
The results obtained in the Spanish projects correspond well to those obtained elsewhere as the comparison with a large number of WPT’s worldwide carried out in 31 ‘international’ projects demonstrate. The ranges of scattering of the critical pressures obtained in a total of 52 projects are listed in Table 3. It is evident that both groups yielded similar – or even equal – results, e.g. all results are representative.
The results obtained with the analysis of the many examples in almost all possible rock types confirm the courses of hydrojacking and hydrofracturing postulated earlier:
• Hydrojacking of existing paths causes an over-proportionate increase of the water take. It begins once the critical pressure is reached, it continues as long as the test – or grouting – pressure grows. The critical pressures increase with depths as long as they are controlled by the overburden pressure. Furthermore they are ruled by strength; there they are scattering but do not increase with depth.
• Hydrofracturing of latent discontinuities takes place if the critical pressure is reached being able to overcome the tensile strength across the planes and the compressibility of the rock. The various types of discontinuities are unequally susceptible to fracturing: bedding planes are prone to cracking as well as cleavage planes or parallel texture of planar crystals as mica, for instance.
All examples prove that in unweathered rock owning a connected bond, the critical pressures do not increase with depth but vary within an individual range of scattering because susceptible discontinuities are not completely identical but differ with respect to: homogeneity of rock types; shape of planes and tensile strength; angle between borehole and planes; intensity of relaxation underneath slopes of different steepness.
Fresh igneous rocks usually own larger compression strength and modulus of elasticity than young sedimentary rocks. Consequently the latter ones should have substantially lower critical pressures, bedded sandstones or limestone, for instance. Surprisingly that is not true. Perhaps granites and similar strong rocks show a wider range of scattering, however the order of magnitude of the critical pressures observed in all the rock types analysed differs just a little: latent discontinuities are also fractured in granites at pressures between 5 and 10 bar as they are in slates or young sediments. Nevertheless, within the same order of magnitude appear certain differences reflecting the geological conditions (Table 4):
• The low critical pressures between > 3 < 7 and > 3 < 5 bar, respectively observed in the programs CAN and VIL correspond to their rock types – cleavable slates.
• By contrast, the thick banks of dolomites and limestone found in the program PAL yielded higher critical pressures (> 8 < 12 bar) obviously reflecting the larger strength.
The critical pressures causing hydrofracturing are surprisingly low in most rock types. Assigning the rock types considered to the seven groups between < 5 and > 30 bar, their portions are distributed as follows (those three cases with Pcrit > Pmax are disregarded):
Pcrit = < 5 bar: 17.8 %
Pcrit = 5 –10 bar: 64.4 %
Pcrit = 10 –15 bar: 11.1 %
Pcrit = 15 –20 bar : 0.0 %
Pcrit = 20 – 25 bar : 2.2 %
Pcrit = 25 – 30 bar : 2.2 %
Pcrit = > 30 bar: 2.2 %
This relationship may not apply to all rock types existing worldwide but its tendency is probably representative. Conclusively, approximately 82 % of all cases of hydrofracturing begin at critical pressures smaller than 10 bar and 93% of all cases are fractured at pressures smaller than 15 bar. Rock types with larger strengths of their rock bonds are evidently very rare.
Mechanism of hydrofracturing processes
Since many of the rock types analysed are quite strong, the critical pressures seem to be surprisingly low. However, the decisive factor is the tensile strength across the discontinuities. It is evidently not as large in strong but cleavable rocks as it is in well-bedded sequences. By contrast, massive rocks without such planes are not susceptible – granites without prevailing texture or massive young limestone. The susceptibility to hydrofracturing depends on several factors: the availability of latent discontinuities; the angle of intersection between borehole and susceptible plane; the drilling technique; the stress field at the entrance of a susceptible plane; the stiffness of the rock bordering the plane being fractured; the development of pressure during fracturing and subsequent grouting.
Conventional grouting pressures
The conventional concept for curtain grouting recommends an increase of the pressure with depth (Figure 30). This concept compromises between two contrary requisites:
• To avoid dilation or fracturing due to heaving of the overlying rock the grouting pressure shall be small near to the surface in order not to exceed the low overburden pressure.
• To achieve a wide spreading of the grout suspension injected, higher grouting pressure should be applied in deeper zones to allow for subsequent series of groutholes of staggered depths.
This compromise presupposes that the overburden pressure always controls hydrojacking and hydrofracturing. This analysis revealed rather the opposite: the overburden pressure is effective only within weathered rock, while in fresh rock dilation and fracturing depend on the strength. It varies due to local conditions but does not principally increase with depth. Conclusively, groutholes of staggered depths are not appropriate and grouting pressures increasing with depth soon exceed the critical pressures. In such cases latent discontinuities intersected in tight borehole sections are fractured and subsequently grouted causing – often huge – additional expenses and grouting time. Thus, whenever we have to reckon with latent planes, the grouting pressures should not increase with depth. In penetration grouting aimed at filling open voids the pressure should be related to the fracturing behaviour.
The idea that the overburden pressure determined the hydrojacking and hydrofracturing behaviour of deeper zones was misleading from the very beginning: If the wall of a borehole is compressed the deformation of the rock does not depend on the weight of the overlying rock but on the strength of the rock around the wall. The reduction of the pressure around a borehole occurs principally similar to the reduction of the pressure around a tunnel: as long as the rock is solid or just slightly weathered and the rock bond is largely connected, the load of the rock rests on the roof, forming an arch which diverts the pressure around the tunnel into the ground. If weathering reduced the rock bond, technical support is required to keep the roof stable. If the rock bond is completely disturbed, the support has to carry the entire overburden pressure. Under extreme conditions the rock collapses completely resulting in an open shaft up to the surface.
Changes of absorption capacity
Instead of the conventional Lugeon-values, the author prefers to use the absorption parameters Q1’/Q10’ and Q1*/Q10* [4, 7/2, 8, 9] because they permit recognising the deformation of water conducting paths either due to hydrojacking or hydrofracturing, i.e. the increase of absorption. The comparison between the mean values (Figure 37) and the relative frequency distribution (Figure 38) for the parameters Q1’/Q10’ and Q1*/Q10* demonstrates whether – and in how much – the effect of hydrojacking or hydrofracturing increases the absorption capacity – and possibly the permeability.
Comparison between critical pressures
This contribution analyses the critical pressures determined by means of WPT’s. The question arises whether the critical pressures occurring in grouting share the same level in spite of the larger density of the grout mixes. Unless modern equipment is available to record directly the pressures effective in the borehole section tested or grouted, the comparison is difficult when the data are recorded at the top of the borehole, as is often still common practice. The true pressure applied in WPT’s can be determined by ‘pressure correction’ but the true pressure effective in grouting is hardly determinable due to unknown factors. Thus, in evaluating testing and grouting data we are often confronted with the combination of true test pressures and grouting pressures referring to the top of the hole. This combination also applies to the examples used for the analysis.
The main article discusses in detail the relationship between the different types of pressures and a possible approach to figure out the true grouting pressures. Besides the different densities of water and suspensions, the latter one need for their flow higher pressures to overcome the head losses due to friction. Conclusively, it is distinguished among the following groups: If the grouting pressure gains more due to density than it looses due to friction, the critical pressures observed in grouting seem to be larger than in WPT’s; If the grouting pressure looses more due to friction than it gains due to density, the critical pressures seem to be smaller in grouting than in WPT’s; If head losses and gains compensate each other, the critical pressures in grouting and testing are similar.
The increase due to density depends a) on the W/C-ratio and b) on the depth of the section treated. The head losses due to friction are determined by the flow properties of the suspension and the geometry of the paths: the narrower the paths and the larger the viscosity the higher the pressure to initiate the penetration of the paths and keep the suspension flowing to reach sufficient spreading of the grout.
In testing and grouting of open paths, pressures to overcome the head losses due to friction are needed to penetrate them and to maintain the flow. The height of the pressure required depends on the size and the shape of the paths, the roughness of their walls and the flow properties of the materials injected. The narrower the paths and the larger the cohesion of the materials the higher the pressure required – and vice versa. The injection of water and grout develops differently: the flow of water can be endless, the flow of grout ends off as soon as the path is clogged due to sedimentation – as intended. Since the Lugeon-criterion plays such an important role the author carried out lab tests to find out for various sizes and shapes those paths being capable of absorbing that specific amount of water defined as Lugeon-unit and to quantify the pressure required to grout them.
The tests revealed that a planar shaped conduit of 0.3mm in width and 20.8mm in breadth yielded water takes between 0.9 and 1.3 LU, depending on the length of the path: 300mm and 900mm, respectively. A grouting pressure of 12 bar was needed to penetrate that path using a suspension of W/C-ratio = 1. The flow through that path stopped once the pressure dropped below 9 bar.
With growing width of the paths the absorption capacity increases yielding larger Lugeon-values while the required grouting pressures decrease accordingly. Other widths and shapes yield different results, of course. Each real fissure possesses its specific relation between pressure required versus geometry and conductivity.
In respect to the relation between width, conductivity and pressure required, the lab tests confirmed what had to be expected. The lab tests also identified the geometry of those paths absorbing the water take of 1 LU which is very small. They also quantified the pres-sure required to grout such paths, they increase inversely proportional to the width of the paths (Figure 39). The cohesion of the suspension is also effective: thicker suspensions need higher pressures – and vice versa. The results are in conformity with the fundamentals of hydraulics as calculations confirmed. The respective details are presented in .
Hydrofracturing and individual groutability
Relating the high grouting pressures needed to grout narrow paths to the critical pressures leads to the conclusion that each rock type possesses its own individual groutability. It originates from the interplay of: the critical pressures (Pcrit); the grouting pressures required (P); and the actual grouting pressure (Pgrt).
Since each of these factors differs, we have a great variety of individual conditions. In general it is summarised that narrow paths are not groutable if the grouting pressure required is higher than the critical pressure fracturing latent planes (Figures 39 and 40).
Hydrofracturing restricts width and depth of groutable paths
The following factors form an important interplay: pressure developed by the pump and recorded at the surface (PP); additional pressure caused by the weight of the suspension due the W/C-ratio; and depth of the stage being grouted. Two examples are presented to demonstrate the consequences considering a grout mix of W/C = 1 causing a weight of g = 1.52 g/cm2 and no counter pressure due to groundwater:
• With a pump pressure of PP = 2,5 bar, the effective grouting pressure reaches 10.2 bar at a depth of 50m.
• With a pump pressure of PP = 5.0 bar, the effective grouting pressure reaches 9.6 bar at a depth of 30m.
Furthermore narrow paths necessitating higher grouting pressures cannot be grouted if latent discontinuities exist cracking at Pcrit < 10.2 or 9.6 bar, respectively. The main article discusses the details.reveals that increasing pump pressure and increasing weight of the grout mix reduce the depth of the zone in which narrow and even medium sized paths are still groutable provided latent discontinuities possessing low critical pressures exists.
Hydrofracturing restricts grouting pressure
The results of this analysis proves that the conventional increase of the grouting pressure with depth and staggering depths of successive groutholes are not appropriate whenever latent discontinuities exist. If the rock mass is susceptible to fracturing rather the contrary is more likely to be expedient because increasing depth and increasing weight of the suspension cause the grouting pressure to easily exceed the critical pressure.
Hydrofracturing causes economical disadvantages
Hydrofracturing followed by a re-filling of fractured planes causes greater expenses, which impair considerably the economic efficiency of a grouting programme without improving the rock significantly.
Assessing the GIN-Method considering results of this analysis
Deere & Lombardi  proposed their GIN-Method to carry out grouting of rocks. It uses constant Grouting Intensity Numbers. GIN suggests grouting pressures between 15 and 50 bar The minimum grouting pressure proposed for the smallest GIN-Number is already 15 bar. Thus, the author suggests the grouting pressures proposed provoke hydrofracturing because they exceed by far the critical pressures of more than 90% of the rock types.
Proposal for grouting design
Besides the project conditions (value of water, height of reservoir, type of dam), each rock mass has its own permeability, hydrogeological regime, strength and groutability. In assessing the grouting requirements, and designing and executing the grouting work all components should be duly considered to achieve an optimal solution. If the given permeability and hydrogeological regime require an impermeabilization of the subsoil and if a favourable individual groutability is promising, the grouting work should be designed utilising the concept ‘Grouting based on facts’. .
Figure 42 illustrates this concept.
The results of this analysis are summarised as follows:
• The various types of discontinuities have a different deformability. Hydrojacking occurs preferably in regular joints, hydrofracturing in latent discontinuities owning a pronounced fissility due to bedding, cleavage or laminar texture of planar crystals.
• Hydrojacking enlarges the grout takes. As long as the consumption of cement and the distance of the travel can be kept in reasonable limits, grouting of dilated joints is advantageous because it leads to a better impermeabilization. Within the surface-near zone of weathered rock the widening of open joints is caused by heaving of the overlying rock, i.e. the overburden pressure is the decisive factor. In that zone the pressures widening the joints increase with depth. Weathering ends at different depths: the weathered rock can reach down to a depth of several ten meters or ends after a few meters. Furthermore, where the rock bond is still connected, the dilation of open joints depends on the deformability of the rock surrounding the borehole. There the pressures are scattering within an individual range not increasing with depth.
• Hydrofracturing of latent discontinuities should be avoided – or minimised at least. Hydrofracturing causes very large grout takes and extends the grouting time, often enormously. With the exception of a few cases, the re-filling of originally latent planes does not reduce the permeability of the rock mass: the water seeps through the joints left open and around the groutstone pillows deposited along the formerly closed planes. The great effort and expense required for such measures are spent in vain.
• Grouting pressure: As the critical pressure causing fracturing determined by strength does not increase with depth, the grouting pressure should not increase either whenever the rock possesses latent discontinuities as is the usual case. The custom of a grouting pressure proportionately increasing with depth is not in harmony with the conditions of most rock types; all rules of thumb concerning the grouting pressure should not be applied any more. The appropriate grouting pressures have to be determined individually from case to case by means of WPT’s, taking into account the critical pressure of susceptible planes as well as the density of the grout, the depth of the zone being treated and the counter-pressure due to groundwater.
• Staggered depths to be applied for groutholes of subsequent series are not expedient if the grouting pressure does not increase with depth. The holes of all series should reach to the same depth.
• Pressures = 10 bar are required to penetrate and to grout those narrow paths yielding water takes on the order of 2 LU. The impermeabilisation of these paths by means of cement suspension grouting is impossible if latent planes fracturing at lowers pressures already exist in the same borehole section. According to the result of this analysis that condition applies to the majority of all cases. The unfavourable relationship between the high pressures required to grout narrow paths and the comparatively low critical pressures fracturing latent discontinuities identified for the majority of rock types has serious consequences: narrow paths cannot be grouted because before the high pressure required to penetrate and to grout these paths is reached, latent planes are fractured at a lower pressure. Large water takes do not necessarily indicate groutable rock: high grouting pressures are also required if these water takes result from a multitude of narrow paths found in the same section; they cannot be grouted if latent planes fracturing at lower pressures are also available.
• Restriction of depths: If latent discontinuities exist narrow paths can be grouted only to a limited depths. The achievable depth varies: it decreases with growing grouting pressure imposed at the top of the borehole and increasing weight of suspension.
A final statement: It is rather easy to press large quantities of grout into the rock; that needs simply a sufficiently high pressure. Unfortunately, just a large cement consumption is often seen as a confirmation for both the necessity and the correct execution of a grouting programme. However, a meticulous and sceptical analysis discloses that a large number of grouting programmes were not carried out appropriately: on the way to the high grouting pressures designed, latent planes are fractured which before subsequently absorb large grout takes. Such a process wastes lots of money and time. This analysis demonstrates that latent discontinuities already fracture at much lower pressures than realised so far. The due consideration of that fracturing behaviour should help to avoid its harmful consequences.
The author is Friedrich-Karl Ewert, Geologist, retired professor of Geotechnics, Germany.
The author thanks all Spanish authorities involved and, above all, his colleague and friend Alberto Foyo who supported this project with great efforts.