Engineering geology

Underground structures in karst

Characterization of karst

The name “karst” is used for sedimentary rock mass, which is apparently influenced by leaching processes of its water-soluble components. Mainly limestone, dolomite and gypsum are the subjects of karstification, which can mean as well superficial loss of strength, as the formation of subsurface cavities. The process of subsurface erosion starts along joints, which are widened more and more, especially at their crossings.

karst landforms
Figure 1. Karst landforms

Finally, the ground is intersected by a lot of dykes and shafts, locally enlarged to big caverns (Fig. 2). Where the surrounding rock is still stable, even huge caverns can last for thousands of years. Under changed hydrogeological conditions they can fall dry perma­nently or temporarily and be filled with stalactites and stalagmites as carbonate precipitation or with sediments spilled into the opening from above.

When the span of a cavern becomes too large or the surrounding rock becomes weak by the time, caverns start to disintegrate. This can happen step by step as a long lasting sequence of rock falls, or suddenly as total collapse. The collapse of a cave deep under the surface will stop, when the opening is filled by the loosened material from above. Only the collapse of caves near to the sur­ face can include all the overburden, creating a sinkhole.

On their way from big depth to the surface, karst caverns can have a very eventful history. Periods with and without water, with erosion and new sedimentation, in a stable state and with collapses may have changed several times. For that reason the variety of manifestation of karst is wide, and it is difficult to make a prediction without exploration.

Sometimes karst just means stable rock with locally enlarged joints and small caves without any importance. n other cases all the ground con­sists of big blocks of hard rock in a matrix of sticky clay. Some caverns are totally open, filled with air or water. Other caverns are partly or completely filled with hard, soft or mixed material, which can be dry or wet.

Figure 2. Karst cavern, met during tunnel driving

Karst water

In the karstified ground, there is a lot of open joints, dykes, shafts, natural galleries and caverns, form­ing a system of very effective waterways and water reservoirs. Such a ground is totally different in view of its hydraulic behaviour, compared with all other types of ground. If the karst cavities are already filled with water, the water inflow during tunnelling can be extremely big and long-lasting.

If the cavities are empty and no water is found during tunnel driving, a flood can happen within some minutes or hours after heavy rainfalls. The time between rainfall and flood is especially short if the underground waterways are directly connected with the surface by open natural shafts. When the precipitation of a large area is collected in dry valleys, hollows or the crater of a doline, shafts ending in such structures will lead big quantities of water directly into deep zones.

Such constellations are often found in karst areas. The groundwater level can suddenly arise by some 10 m, flooding areas, which are usually situated far above the groundwater table. That can occur even if there is no visible opening at the surface, but only a thin permeable cover of soil above the karst rock.

Special properties of karst in the view of the tunnelling

The heavily karstified ground is characterized by its extreme inhomogeneity. Hard, compact rock ab­ ruptly alternates with clay and sand or zones with totally broken rock. The size of cavities reaches from some cm to several 10 m. The quality of in situ rock varies between fresh and totally weath­ered. The soil filling of cavities can be cohesive or not, soft, stiff or even hard, and it can be mixed with stones and blocks. For that reason, the strength of the ground can be 200 MN/m 2, but also only 1MN/m2 or even 0 MN/m2. The stiffness can be 20.000 MN/m2 as well as 0 MN/m2.

Problems and dangers in tunnelling because of the inhomogeneity are considerable. Excavation has to be done partly by blasting, partly be digging, sometimes both within the same round. One part of the face is stable, the other needs immediate support. Such works are dangerous and time-consuming, if not done in a proper way.

The bearing capacity of the ground and the loading of the temporary and final lining is highly de­pendent of local differences in strength and stiffness. t is necessary to include a variety of differ­ent assumptions in the structural calculations, for example, asymmetric loads and embedding. An open cavity means locally no loading and no embedding, a soft filling in the roof can act as a big load, besides the tunnel or below it means a total lack of bedding.

Possible modes of failure and dangers during tunnelling in karst

There is a big variety of possible modes of failure when underground structures are excavated in karst. They yield numerous problems and dangers in the different stages of work. Especially large caverns, empty or filled, are highly dangerous when they are met unexpected and the miners are not prepared for the sudden changing of the ground conditions.

Where caverns are filled by loose or soft material, the filling may flow out and bury the crew. Figure 3 shows such a case. The slope of soil had to be secured with shotcrete and anchors, after that the emptied cavern was filled with concrete.

Figure 3. Loose material, flown out of a karst cavern above a roof of the tunnel

Besides instabilities at the face, the high inhomogeneity of the ground can create different kinds of deformation, displacement of the already stabilized opening and failure. Examples are vertical movement by settlement, horizontal movement by unsymmetrical loading or bedding and the col­ lapse of hollow caverns below the actual working floor.

Sometimes opening created by such an effect are so large, that not only persons but also trucks or excavators can fall into them. Figure 4 shows the entrance of a large system of cavities, exposed during bench excavation. The roof of some parts of it was only 0,5 m below the invert of the top heading. t was just good luck, that no accident happened.

Depending on the position of a karst cavern relative to the underground structure to be built, there can happen a collapse into the natural space or a breakdown of material out of it. In the latter case stones or even big blocks can fall out of a matrix of soft soil, or free-hanging parts at the roof separate from their surroundings and fall into the tunnel.

Another typical mode of collapse is the failure of a thin pillar between the tunnel wall and an un­ known caverns behind it.

After strong rainfalls the tunnel can suddenly be flooded, even its total filling with water is possible. That can be a serious danger for the tunnelling crew if it cannot escape in time. Also, the stability of the tunnel can be influenced seriously, when the strength of the ground is diminished by water­ ing and the loading of the lining is increased at the same time by water pressure.

In such a case it is possible, that an opening stable till then collapses by the failure of the face or by yielding of the foundation of the shotcrete vault. In Fig. 5 the lining is destroyed and the siltstone behind it has been liquefied after a karst flood event.

Figure 4. Entrance of a large system of karst caverns. met during benching
Figure 5. Loose of the strength of water sensitive ground under the influence of karst water

In the karstified ground, the water level can arise by some 10 m within a very short time. The existing state of equilibrium, in which no water pressure acts because of the drainage effect of the opening is disturbed suddenly. The high water pressure cannot be borne by the shotcrete lining. If the tun­nel does not collapse in this situation, at least the lining will be deformed and damaged seriously.

Fig. 6 shows the top heading of a tunnel, which nearly collapsed due to a sudden karst flood. The tunnel could be saved by means of some 100 timber props. However necessary ground improve­ment by grouting and total reprofiling took half a year.

Figure 6. Vault of the top heading nearly collapsed because of sudden karst flood

Special measures to overcome bad ground conditions in karst

There are different possibilities to reduce the risk of a collapse at the face. The simplest one is to subdivide the cross-section to be excavated. After each of the partial excavation steps a first stabi­lizing is done (Fig. 7). Having excavated all parts of the cross-section of the respective heading (as a rule of the top heading) installations of shotcrete, steel ribs and anchors is completed, before a new round is started.

If such a procedure is not enough for safe driving, the cross-section can be divided into two or three separately driven headings. Most common is the division in a left and a right part or into two lateral galleries and a final central drift. The parts are driven independently with their faces at different chainages. The distance between the most advanced drift and the final connection of the openings and completion of the provisional lining should be 10 m as a minimum; it can be also 30 m or more.

The small cross-sections of the partial drifts reduce the risk of collapse and accidents considerably. Besides, the first drift can be used as an exploratory gallery. If found to be useful, ground improvement measures can be executed in the pilot gallery, making easier driving of the other parts of the cross-section. Also, the drainage effect of a pilot gallery can often be very advan­tageous.

Figure 7. Partial opening of the face in disturbed karst ground

Following changing types and shapes of karst structures, the mode of subdivision can be changed. Sometimes a lateral pilot gallery will be best, sometimes a central one. n this way driving is flexi­ble adapted to the local requirement.

Another method to stabilize the face area consists in the installation of longitudinal advancing steel elements, mainly long anchors at the face (Fig. 8) and pipe umbrellas above the roof (Fig. 9). Drill­ing the holes for these elements can be used as a kind of sounding at the same time, giving in­ formation on the ground conditions to be expected along the next 10 – 15 m.

Figure 8. Long anchors supporting the face
piperoof in karst
Figure 9. Installation of pipe umbrella in the karstified ground

Instead of special efforts in driving and support measures often it can be much easier to improve the ground conditions before driving. For example, loose material like a heavily broken rock can be stabilized by cement grouting. Such a work can be done either from a pilot gallery or from the surface. Also, big hollow caverns can be filled with mortar or concrete in this way.

If such struc­tures have been found in advance by exploration from the surface, their filling should be done in advance also from the surface. The advantage is obvious: Logistics of concrete delivery and pump­ ing is often easier on the surface, and a filling in advance does not make necessary interruption of tunnelling works. When large caverns are found only during driving, their filling is mostly done from the existing opening. Concrete can be applied as shotcrete, as pumped concrete behind formwork or pressed into boreholes (Fig. 10).

tunneling in karst
Figure 10. Filling of a karst cavern below the top leading through boreholes

Soft, cohesive ground, mainly clay, as a filling of karst cavern, cannot be improved by grouting. When it is met during excavation and its properties found to be not acceptable, it can be taken away step by step and replaced by shotcrete or concrete (Fig. 11).

karst tunneling
Figure 11. Excavation of soil out of a karst cavern below the face

Adapted from D.Kolic, UnderCity

Leave a Reply

%d bloggers like this: