A considerable amount of information has been published regarding cement grouting of jointed rock masses and there are good text books covering this topic(1,3). However, in more recent times claims have been made as to the extent to which rock mass permeability can be decreased by cement grouting, claims that the writer finds difficult to accept given experience on some recent projects in Australia. Before addressing this experience it is necessary to note some fundamentals regarding rock mass permeability and cement grouts.

Rock mass permeability

There are practitioners in this field, e.g. Ewert(4) who think it inappropriate to consider mass permeability values, with the associated connotations of homogeneity, in a jointed rock mass. The writer acknowledges their experience but believes that retaining Darcy’s Law, and the associated calculations that can be made using this theory, provides a good basis for evaluating design decisions regarding grouting. An analogy is the use of elastic theory in soil and rock mechanics; we all know these materials are not elastic but it is well established that wise use of this theory can provide good design guidelines, and equally well prevent some silly conclusions.

Notwithstanding the limitations of borehole packer permeability testing, almost all assessments of rock mass permeability in Australian tunnelling projects are made from such tests and expressed as Lugeon units. As a reasonable approximation, 1 Lugeon = coeffecient of permeability of 1 x 10-7m/sec.

Table 1 sets out a general guide as to rock mass permeability characteristics in terms of measured Lugeon values.

It is important to note that 1 Lugeon does not represent a very low permeability, being one or two orders of magnitude greater than the typical clay core of an embankment dam. This is further illustrated by the following example of calculated inflows into a tunnel.

A recently completed 3.4km long, 3.5m diameter, cable tunnel in Sydney is located in sandstones and shales about 25m below the water table. The project criterion was for a long term inflow of 3.3 lit/min/100m (i.e. 0.00055 lit/sec/m).

Steady state tunnel inflow may be calculated using the equation: (Equation 1)

The effective average mass permeability should be taken as the log mean of measured Lugeon values and Fs, the shape function, may be estimated by using sketched flow nets, by computer analysis, or by the equations: (Equation 2)

It can be readily found that the parameter Fs is quite insensitive to the geometric conditions and ranges between about 3 for no drawdown of the water table, to 2 if drawdown occurs to tunnel level.

If in the case of this Sydney cable tunnel we assume no drawdown, then equation 1 shows that the project inflow criterion implies an effective mass permeability: (Equation 3)

This is an extremely low value and a question that needs to be asked is, could such a low mass permeability be achieved by grouting? However, before proceeding to this question, we need to address a few fundamental considerations regarding cement grouts.

Modern cement grouts claims

Table 2 sets out the typical particle sizes of cements used for grouting. Penetration of cement grouts into joints and cracks in a rock mass is considered to be substantially controlled by the D95 particle-size with the following being reasonable guidelines:

  • OPC Cracks to = 500 micron (0.5mm)

  • Microfine = 80 micron

  • Ultrafine = 30 micron

    In passing it should be noted that silica fume (amorphous SiO2), which is a common additive to cement grouts, has a typical particle size <2 micron.

    There appears to be no controversy regarding the mass permeability achievable in jointed rock using Ordinary Portland Cement. Houlsby’s (1) guidelines remain valid in that it can be expected to achieve:

  • 5 – 7 Lugeon with reasonable effort (primary and secondary holes)

  • 4 – 5 Lugeon with considerable skill and effort (probably involving tertiary holes)

    When it comes to the use of microfine and ultrafine cements in grouting around tunnels, matters are more controversial. Scandinavian practice suggests that effective mass permeability values of 0.1 Lugeon, or less, can be achieved. For example, Finnish practice(3) includes grouting for what are termed Class 1 tunnels, where the allowable inflow is <2 lit/min/100m. From the example given in Section 2, above, it can be seen that this would normally imply an effective mass permeability of <0.1 Lugeon (i.e. k<1x10-8).

    Proper scientific evaluation of this claim requires accurate measurement of the mass permeability before grouting (measured by Lugeon test) compared with that after grouting. It is not sufficient to say that after grouting the inflows into a tunnel were less than, say, 5 lit/min/100m and therefore the grouting was effective. One has to know what conditions would have been without grouting.

    An important aid in assessing this matter of ‘groutability’ of rock is given by Fell, MacGregor and Stapledon(9). This is based on the:

  • relationship between Lugeon value, fracture width and number of fractures per metre, and

  • relationship between grout particle size and crack width that can be penetrated.

    The conclusions from this analysis are summarised in Table 3. It can be seen that it is only in the case of a rock mass with very wide fracture spacing that penetration of ultrafine cements can be expected where measured Lugeon values are less than about 2µL. While Table 3 does not indicate what permeability can be expected after grouting, its clear inference is that achievement of Lugeon values <1 would be very difficult.

    As discussed later, the writer’s experience suggests that the expectation of achieving log mean permeability values of the order of 0.1 Lugeon using microfine and ultrafine cements is unrealistic.

    Tunnel layout and rock mass characteristics

    The engineering geology of the three-lane Domain and Burnley tunnels in Melbourne is described in (5), and the hydrogeology (6). The tunnel passes at a depth of 60m beneath the Yarra River (Figure 1) and is located entirely in folded Silurian siltstones and sandstones (the Melbourne Mudstone). A substantial paleovalley was cut into the Silurian rocks in the late Tertiary, and the base of this valley comprises an alluvial aquifer referred to as the Moray Street Gravels. Key geotechnical features are summarised in Table 4.

    Grouting works

    Computer simulations was undertaken to predict the movement of the permanent concrete lining under full hydrostatic loading of 60m. As with any lining under such loading the deflections create water paths behind the tanking membrane and as a result of some leaks through the lining it became necessary to undertake a programme of rock mass grouting through the completed lining. This grouting was designed to target areas of high permeability rock, and was done in conjunction with several other measures which eventually reduced total inflows to the design target. It was undertaken after an extensive programme of work involving anchoring of the invert slab of the tunnel.

    There are two facets of the rock mass grouting in the Burnley tunnel that provide valuable information as to mass permeabilities that can be expected using modern grouting techniques.

    The first was a grouting trial where two specialist contractors (one cement, one chemical) were given access to 50m lengths of tunnel sidewall wherein careful measurements had been made of pre-grouting permeability values. The contractors were allowed to use any materials in their “armoury”, any hole spacing and any techniques; with the aim being to achieve a final log mean permeability of <=1 Lugeon. The one contractor used acrylates and polyurethanes, the other microfine and ultrafine cements.

    The second facet was production formation grouting, using microfine and ultrafine cements, over a length of about 1km of tunnel wherein permeability measurements were made prior to grouting and at completion of the work. The target permeability had been set at 3 Lugeon on the basis of the grouting trial

    Grouting trial

    Lengths of tunnel for cement and chemical grouting were established so that the two grouting approaches could be compared on a reasonably equal basis. For both grouting approaches primary drilling and pre-grouting water pressure testing were undertaken on a set pattern. From this point the contractors were given a free hand as to how to conduct the grouting. After the grouting had been completed water pressure testing was carried out in selected holes to test the rock mass permeability reduction achieved.

    Multiple boreholes were drilled along sections perpendicular to the tunnel walls at spacings of approximately 1m, resulting in 24 sections being drilled in each wall. These were drilled by a track mounted rig using a top hole hammer; the hole diameter was 75mm.

    Each section comprised a “fan” of boreholes drilled in the same direction but with varied inclinations and lengths. This design was used to give a one metre vertical spacing between boreholes at a point approximately 3m behind the concrete sidewall of the tunnel. The effective borehole pattern for the entire wall, 3m behind the concrete facing would therefore be a grid of boreholes at 1m centres.

    A total of 13,300 lit of cement grout was injected at pressures of up to 700kPa. Table 5 gives a comparison between initial and final Lugeon values at different “fan” locations. Discussion of the chemical grouting trial is outside the scope of this article. Suffice it to say that it was decided to proceed to production grouting using microfine cements.

    Production grouting

    Production grouting was undertaken beneath the tunnel floor and behind the sidewalls over some 600m of tunnel. Primary, secondary and in some cases tertiary holes were drilled to give a typical hole spacing of one hole every 2.5m² of tunnel perimeter. All primary holes were water pressure tested and the contractor was required to drill and grout secondary and tertiary holes until it was demonstrated that 3 Lugeon or less had been achieved. Maximum grouting pressures were between 600kPa and 800kPa.

    5600 grout holes were drilled and 280,000 lit of cement grout were injected. Despite the very close hole spacing, about 20% of the final closure holes gave permeability values greater than 3 Lugeon. Overall the grouting work achieved its objective but the facts of this work led the writer to conclude that it would be difficult to be confident of achieving a mass permeability of <1 Lugeon using cement grouting. A reasonable target for the log mean permeability is 2 Lugeon. It is acknowledged there was a limitation on the grouting pressures because work was being done behind the walls of an existing tunnel. Grouting ahead of an advancing tunnel allows the use of much higher pressures, but then the hole spacing is normally much greater than the 1 hole/2.5m² used in this case.

    Other experiences

    Within tunnels in the Triassic Hawkesbury Sandstone that underlies much of Sydney, there have been several uses in recent times of microfine/ultrafine cements to reduce rock mass permeability. These include:

  • the 6.6m diameter Northside Storage Tunnel beneath the Middle Harbour paleovalley(8)

  • two nominally 4m diameter cable tunnels in Sydney, and

  • gas storage caverns beneath Botany Bay

    Fresh Hawkesbury Sandstone typically exhibits Lugeon values ranging from <0.01 to about 50, with log mean values of about 0.2 to 1 Lugeon. However, at the Northside Storage project, valley bulging effects had generated high mass permeability values of greater than 50 Lugeon. Groundwater inflows of about 200 lit/sec were expected over a length of 210m. Forward probing and grouting was carried out over this length with up to 54 grout holes being drilled prior to each 4 - 6m TBM advance (8). Total inflows into this length of tunnel after completion of cement grouting amounted to 18.5 lit/sec. This equates to an average permeability significantly greater than 3 Lugeon.

    On the two cable tunnels most of the grouting has been post-excavation and this is known to be far less effective than forward, canopy, grouting. However, some forward grouting has been performed but in no cases have mass permeability values of less than 3 Lugeon been achieved.

    Unfortunately the writer has no detailed data from the extensive grouting programme carried out around the gas storage caverns beneath Botany Bay.

    Conclusions

    Ultrafine and microfine cements, when properly used whilst undertaking systematic grouting, are very valuable materials with respect to reducing inflows into tunnels.

    However, factual data available to the writer indicates that it is very difficult, if not impossible, to achieve mass permeability values of less than 2 Lugeon in rock masses of much higher initial permeability.

    It would be of great value to the underground construction industry if practitioners could provide permeability data from tunnelling projects, “before” and “after” grouting, so that designers may have a good database as to what can be achieved.

    Related Files
    Equation 3
    Figure 1 – Longitudinal sectio of the Burnley Tunnel showing a simplified geology
    Equation 2
    Equation 1
    Figure 2 – Cross section of the Burnley Tunnel