The Cityringen Metro is expected to be running in 2019 and will consist of two single tracks in twin tunnels, each approximately 16.5 km in length. A total of 17 underground stations, 4 cross-over facilities and 3 construction and ventilation shafts have been made when the project is finished.
The Cityringen is built entirely in the urban area of Copenhagen; most of the stations are located in congested areas, surrounded by historical building. One of the most sensitive activities during construction is the break-in/-out of the station/shafts. It is mandatory to ensure a dry or at least a controllable groundwater inflow during the break-in/out phases. Between various technical solutions, the contractor developed a grouting plug to reduce the inflow of water, which has proved efficient in the Copenhagen limestone.
The Geology of the Copenhagen area consists of Quaternary Deposit overlying the Paleogene Deposits, primarily consisting of Danian Limestone and Locally of Selandian Greensand . The Copenhagen Limestone Formation is divided in 3 stratigraphically units: Upper Copenhagen Limestone (UCL), Middle Copenhagen Limestone (MCL) and Lower Copenhagen Limestone (LCL). The tunneling and excavation activities are entirely in UCL and MCL.
The Copenhagen limestone is a carbonate rock that can be generally described as a weak rock with very hard layers or nodules of flint. The Upper Copenhagen Limestone is horizontally bedded with layers of different indurations and flint beds with thickness ranging from a few centimeters, up to 1 m which in some cases can be followed continuously over long distances of up to 1 km. The Middle Copenhagen Limestone has a more nodular distribution of flint and more evenly distributed silicification. The unit is characterized by a lower frequency of strongly indurated limestone layers and flint bands compared to the Upper Copenhagen Limestone.
The upper part of the limestone is often disturbed or fractured by glacial processes. The disturbed zone is highly variable and ranges from a mixture of indurated limestone clasts in a less indurated, muddy limestone matrix, to a zone of higher fracture intensity than the limestone below. In a few areas, no glacially disturbed limestone is observed, and the undisturbed limestone is directly overlain by Selandian Greensand.
Layer fracturing in the horizontal plane usually occurs along the surfaces between weakly indurated limestone and strongly indurated limestone beds or flint bands. Such fractures are probably caused by relaxation, due to unloading of the ice or by isostatic uplift. The intensity of the layer fracturing usually decreases with depth, but also reflects the differences in the constituent limestone units. Horizontal fractures are usually narrow with apertures sizes in the range of 20 μm to 2 mm, although openings locally up to 10-20 mm have been observed. The layer fractures are often associated with the peak flow zones registered in flow logged boreholes.
The permeability of the Copenhagen limestone is derived primarily from fissure flow (secondary permeability).
The fissuring is related to the structural history of the region combined with the effects of glaciation. In general, the permeability is likely to be higher in the UCL than in the MCL with peak flow zones in the uppermost part of the glacially disturbed UCL. The vertical extent of this flow zone is usually around 2 to 3 m but thicker zones are also observed. In general a hydraulic anisotropy of 10 has been adopted for the UCL, while 5 has been used for the glacially disturbed UCL.
Rock Grouting Method
The rock grouting has been carried out in stages of maximum 5 meters ascending from the base of the borehole.
This method, called upstage grouting, has been utilized on all rock grouted drillings at the Cityring project.
The process is figuratively shown in Fig. 2:
a) In-situ conditions of the limestone, starting steel tube installed before drilling if below groundwater table.
b) Drilling to full length, usually using DTH (Down the Hole) hammer. Casing is used in the quaternary soil and in the top of limestone (if highly fractured).
c) The bottom stage is injected, using a single packer placed in the limestone and the bottom of borehole as boundaries.
d) Next stage is grouted after end of previous stage, using a single packer and previous stage as boundaries.
Due to the composition of limestone, where the majority of water flow is seen in the fractures and fissures, it was assumed that the required reduction in water ingress could be achieved by permeation grouting alone. In rock, permeation grouting is a technique where grout is injected into the pre-existing fissure network, replacing the water in the fractures and fissures with stable cement or chemical grout, hence reducing the permeability of the rock.
As such, mechanical and chemical weakness of the grout was not of particular concern, while the stability and viscosity was.
Three stable mixes with bleed of less than 5 % and different viscosities were produced. To achieve this, various w/c ratios together with bentonite and plasticizers mixes were tested. In terms of effect of the different components, these are simplistically summarized in Table 1:
On the base of the previous table, the following grout mixes were used at Cityringen for all rock grouting campaigns
A grout pressure and volume stop criteria has been defined and later field tested in a subsequent trial test with the three predefined mixes.
The aim of the stop criteria was to define when the grouting is adequate and further grouting would be uneconomical. The volume criterion was defined as 100 l/m of a given grouting stage while the grouting pressure was set to 1 bar/meter of overburden depth. In relation to the refusal pressure, for permeation grouting the process dictates that the grout enters the formation without displacement of the ground adjacent to the fissure.
As the rock, especially limestone is very blocky and consists of both high and low strength rock it is hard to confirm this assumption and to assume no hydro-jacking or fracturing, however no movement in terms of heave or deflection has been recorded as a result of rock grouting at the Cityring project. The graph in Fig. 3 shows various grouting pressures in the US and Sweden in terms of rock quality and depth. It can be seen that the chosen grout pressure is between good and normal rock according to Swedish practice.
For the three defined stable grout mixes, the following work procedure applied. If the refusal pressure Pa is reached before the volume criteria Va (P Pa), the stage is seen as complete and grouting of the next stage can commence. If the volume criteria Va is reached before the refusal pressure Pa (P Pa), a thicket mix is chosen and the process is repeated. The flow chart in Fig. 4 shows the entire grouting procedure for a given stage.
It should be noted that at the Cityringen project, grouting with mix G3 has only been repeated a couple of times and the pressure refusal is usually achieved beforehand.
The grouting layout is defined using the split spacing principle. Initial injections are performed through primary holes followed by a second injection if deemed necessary. As the required work is often to perform an entire grouting plug, a triangle primary spacing has been selected with a distance of 2.6 m between primary-primary holes.
In theory, by starting with the widely spaced primary holes, the largest fractures and fissures are treated. During subsequent permeability tests it is determined if a secondary grouting campaign is needed. If required, the grouting layout is intensified to grout injection holes 1.5 m between primary-secondary and secondary-secondary. A typical layout plan used at Cityringen can be seen in Fig. 5.
In principle a tertiary campaign could be performed if deemed necessary, however this has never been the case at Cityringen.
A deep station has been built adjacent to the existing church known as “Marmorkirken”. The station is 38 m deep and located only 3 meters from the foundation on the western side of the church. The station was built as Top-Down to minimize the settlements due to the deflection of the R-walls. Due to the space constraints from the fully structurally realized station, together with TBM’s entering at multiple levels, a grouting plug solution was chosen in order to save time compared to the demolition and subsequently retrieval of a false tunnel.
Both the upper and lower grouting plugs were carried out from one platform with multiple inclined drillings to cover the lower part as shown on Fig. 7. Subsequently, a cavern was excavated to allow the drilling of vertical boreholes to grout behind the back wall for the lower tunnel break in/out. The grouting layout was designed as an envelope with horizontal drillings around the tunnel, with injections in the middle area to allow the excavation of the cavern.
To allow safe drilling and grouting beneath the groundwater table, pre coring were drilled in the diaphragm wall to a depth of 70 cm, leaving 50 cm to the limestone. Steel flanges were then cast into the cores and pneumatic preventers were installed. With the preventers, it was possible to close the gap between the flange and drill string during the drilling phase and subsequently between the flange and grout injection pipe, during the grouting phase.
Unlike the vertical grouting, the drilling rig is not occupied by the grouting activity as a grouting pipe is used instead and no casing is present, allowing grouting and drilling to run as parallel activities. No grouting and drilling in adjacent grout holes were allowed due to risk of migration.
The grouting sequence was performed to confine the grout mass as much as possible, starting from the outer grouting rings, and progressively working towards the center.
As the horizontal drillings complicated the permeability tests performed by conventional method, evaluations were performed on the water flow by gravity into the horizontal drains (grout boreholes) prior to grouting. Both primary and secondary grouting was carried out to reach the required permeability. Pressure used around 25 bars.
In the middle of the “Sortedamssø”, an artificial temporary island spanning 30000 m2 has been established as tunnel construction site to handle muck trains and tunnel linings. In addition, the shaft has later been redesigned to facilitate the connection between the Cityringen line and Nordhavn branch off. The shaft consists of 41 m deep diaphragm panels imbedded into the middle Copenhagen limestone with the bottom slab 32 m below ground level.
In total, four TBM’s will break-in to the station at the lake side with two break-outs at street side.
Due to the many break-ins at the end wall from two different contractors, a grouting plug has been chosen as the preferred solution to mitigate interfaces. The entire scope of work was:
4 In-situ permeability tests;
112 Primary Drillings with 4 subsequent permeability tests
224 Secondary Drillings with 8 subsequent permeability tests
The grouted envelope was approximately 3 m or more in thickness on all sides of the TBM, with the grouting performed in three grouting stages, each 4.5 m in length. The permeability tests were used to determine the conditions before, during and after the various grouting phases. The results, listed in Table 4, show the average permeability for each grouting phase at the three stages of injections.
It can be seen, that the target permeability of 1e-6 m/s was more or less reached after the secondary grouting campaign. The grouting layout can be seen in Fig. 9.
The entire grouting campaign was carried out over a span of 11 weeks and a total of 279 m3 grout was injected. Maximum pressure used was approximately 30 bars.
Evaluation and Conclusion
It can be difficult to assess when a sufficient permeability reduction has been achieved in an inhomogeneous rock such as the Copenhagen limestone. This consequently can lead to very low permeability test result close to ungrouted fractures if the fracture system is not connected. Observations and evaluations have been performed, based on the various grouting campaigns performed at the Cityringen project in addition to the already proposed Lugeon and horizontal gravity flows. It is noted that all evaluation is post completion with all information available.
From the experience in Copenhagen Area, it has been generally accepted that a cementitious grout with normal Portland cement will not penetrate apertures of sizes less than 0.5 mm or 3 times the max particle size. As such, when the initial pressure refusal is reached with little to no volume, it has been hypothesized that either no fractures larger than 0.5 mm was present or a fracture was present with a very minimal reach.
The frequency of when volume criteria have been reached and subsequently a thicker mix utilized has been graphed below (the example is for the Østersøgade shaft).
As can be seen on Fig. 10, the frequency of reaching the refusal volume criteria was quite high in the beginning of the campaign with generally good grout takes during most of the Primary campaign. At week 6 and 7, the Secondary campaign was commenced based on performed permeability tests. As the grouting is always performed in a confined manor, week 6 and 7 mostly focused on the periphery of the grouting block. The grout takes again are reasonable with around 20 % of grouted boreholes surpassing the volume stop criteria. As the grouting campaign moves inside of the grout block during and after week 8, the frequency is reduced down to 10 % or lower.
Compared to the initial frequency of occurrence, the chance of encountering a large fracture has been reduced by 75 % or more and in the final two weeks of work, the volume criteria was not surpassed once. Looking at the frequency, it can be argued that the grouting campaign could have been suspended sooner, hereby accepting some possible untreated fractures.
Apparent Lugeon Permeabilitiy
Using the grout takes (volume, pressure and flow rate), the Apparent Grout Lugeon has been calculated using the following formula  which normalizes the flow of the grout to water:
The Lugeon values have been recalibrated to permeability [m/s] and are shown in Fig. 11 together with the actual lugeon values for Østersøgade. As can be seen, a huge spread is present in the data but generally the calculated Apparent Lugeon decreases together with the Lugeon test results.
It is therefore recommended to look at the grout takes as well to enhance the available data to give a more accurate description of the conditions in the grouting block.
Many grouting campaigns have been carried out at the Cityring project to facilitate break-in and break-out at various stations. It is possible to achieve an average permeability of 1E-6 m/s using normal Portland cement.
However, the work required is very extensive compared to injections needed to reach 2 to 2.5E-6 m/s which has also been discovered at a previous project . It is recommended to carefully analyze and determine what amounts of acceptable inflow during a break-in/-out and determine if some fractures could be allowed to exist.
Rock Grouting In Copenhagen Limestone – The Cityringen Experience
Authors: Enrico Paulatto*, Sune Carstensen