Problems associated with tunnelling through karst are well documented in China. In order to properly understand the variability in karst we need to understand the processes that control solutioning and associated hydrogeology of these formations. We also need to know the processes that shape karst landscape and the geomorphology of such terrains to appreciate the landscape complexities and their impact on our interaction with this type of ground.

China has large areas of karst terrain with over 23 provinces reported to have significant areas of limestone susceptible to solutioning and karst formation. It is estimated that carbonate rocks outcrop close to the surface, or exist at shallow depth, over about 1.3Mkm² of China’s 9.6Mkm² total land surface. If you include deeper carbonate rocks, up to a third of the country is underlain by these deposits.

Karst is a term that describes the landforms that develop on rock types readily dissolved by water. Approximately one tenth of the Earth’s land surface is karstic limestone and some 25% of the population lives in these regions. This presents a significant geohazard to structural, civil and tunnel engineering.

Karst can be described as either bare, covered or buried. Bare karst is exposed at the surface while covered karst is usually overlain by recent soil deposits. Buried karst is found where soluble rocks lie below other less soluble capping rocks such as shales and mudstones. Palaeokarst is another type of karst which has been formed in a similar manner to covered karst where the surface has been buried by surrounding soils which have then been deeply buried and have undergone diagenesis to become rock.

In any project in karst it is important to define the nature and variability of the rock. It is essential to understand the weathering process, rocktype, structure, hydrogeology, karst activity and geomorphology as these usually define the main problems experienced when tunnelling in these terrains.

Karst and tunnels

Geohazards are common, from unstable limestone hillsides to solutioning and sinkholes to deep water and large underground reservoir systems.

Cavities in karst are frequently encountered in tunnels excavated through limestone. During construction of the 88km of tunnels for the Wan Jia Zhai Yellow River Diversion project the tunnels encountered some significant cavities that stopped progress of the TBMs on several occasions for several months. Numerous smaller cavities also significantly hampered progress. Drill and blast highway tunnels have also encountered cavities on many occasions causing delays and extra cost. Clay filled, open voided and flooded cavities all pose their own problems in tunnelling. Plastic clays or “terra rosa” can occur in the cavities as complete infill or partial infill and can disrupt TBM progress with mixed hard soft material.

Where tunnels run through bare karst the potential effects of dry, flooded and seasonal variation of water levels within cave networks must be understood in detail. Construction could encounter dry cavities that could be subsequently filled with water during rainy seasons or during storms, resulting in large water inflows to any tunnel or higher than expected water pressures on a lining. Groundwater problems in karst areas include the sudden flooding of the tunnel when a flooded cavity is punctured and the continuous flow of water through cavities which can prevent concreting or shotcreting and possibly may prevent grouting.

When a cavity is encountered in a tunnel there are several potential difficulties to be considered. One is the continuity and stability of the tunnel invert. In a small cavity this can be easily overcome by pouring concrete to partially fill the cavity to provide access to the point on the opposite wall where the wall continues. If the cavity is too large then a bridge may be required. A sufficiently large cave or cavity may significantly influence the viability of the tunnel, especially if the roof of the cavity cannot be stabilised or supported.

Collapse structures within caves can have debris lying on the floor of cavities in bare karst areas and may even be partly or wholly infilled with clays. Terra rosa can cover the upper surface limestone areas and can also infill some of the collapse structures. This material is sometimes very soft and can be problematic. Covered karst collapse structures are usually infilled with soils and the mix of hard and soft ground can be difficult to tunnel through. Disturbance to these materials may even induce further collapse and subsidence. Where a TBM tunnel encounters the base or the side of an infilled cavity forming a mixed face this will cause problems in maintaining directional control of the machine. It is also likely to cause significant over-excavation in the soft material which may result in more settlement at the surface than predicted.

Covered karst terrain

Tunnels that are constructed within alluvial areas or above covered or buried karst terrain can also experience indirect affects from karst, including ground movements from underlying karst collapse and sinkhole development induced by below tunnel dewatering.

Portals are usually located within steep limestone cliffs and slopes with very poor stability. Often, to ensure portal stability, large scale stabilisation has been required above these areas. The portal location should be chosen carefully to ensure that these expensive measures are minimised.

Radon gases and CO2 are other potential tunnelling hazards. Cavities and cave systems may have pockets of these gases that should be considered. Drill and blast and open shield TBMs are more exposed to this.

There is also a potential problem of alkali-carbonate (dolomite) reaction of some rocks, particularly dolomitic type rocks with cement grouts and concrete. These may be minimal, although problems would tend to affect the surrounding rocks, as long as the concrete lining does not contain dolomitic limestone aggregates. Effects are more pronounced when swelling clay minerals lie within the rock or aggregates. If the concrete contains dolomitic limestones then the durability and longevity of the lining may be affected.

Identifying geohazards

Karst implications for engineering are often studied using risk assessment with the level of acceptable risk increasing with decreasing sensitivity of the structure.

Tunnelling should be considered as an operation that should attract low acceptable risk levels due to inherent safety considerations during initial construction and future operation of the tunnel.

Risk investigation should focus on hazards that produce the most significant impact on the tunnel. A thorough desk study must be undertaken to determine physical and indirect hazards and external influences. From this a site investigation (SI) should be developed and implemented to test geological models and areas of uncertainty and identify hazards along the route. Finally a strategy should be developed at the construction option stage to deal with the identified hazards to ensure precautionary measures and procedures are implemented.

The benefit of a thorough desk study cannot be overstressed as it provides information in targeting and planning any SI efficiently and cost effectively.

Engineering geologists familiar with karst terrain should be employed to review and assess the information gathered during the desk study phase. The aim of the study would be to understand the current and past karst processes that have shaped the area and it should include a range of techniques from a site walkover and mapping of the karst terrain, a study of the geological maps of the area, reference to the karst collapse GIS database of China and other karst related data to a study of all historical data relating to the area under study. These parameters can vary depending on the job’s complexity and the initial assessment on the ground conditions along the tunnel alignment.

After the desk study process a preliminary geological model identifying the areas and degrees of certainty of the geology and geohazards along the tunnel alignment should be identified. This will allow the model to be tested by the SI phase, if considered necessary.

Site investigation

The cost of investigation should be linked to the importance of the tunnel and level of hazard identified during the desk study. The investigation and procedures in undertaking the investigation of the ground along the tunnel may vary according to the nature of the project, the complexity of the geology, the background of the consultant and /or the contractor, and the experience of the geologists or engineering geologists involved.

The fact that geological formations are spatially variable, and that only limited measurements or observations can be made, has important consequences. Ground variability can be extreme and there should be a balance between value for money in investigation against the development of contingency plans to deal with the variability and hazards during tunnelling. These may even preclude tunnelling and other methods such as realignment may be considered.

In the case of bare karst areas, a detailed mapping exercise should form the majority of the site investigation. Mapping and exploration of caves can be undertaken where necessary and the use of tracer tests to determine the source and flowpaths of various swallow holes and springs along the alignment should also be carried out wherever possible.

Geophysics can be a very useful tool in the assessment of karst terrain for tunnel alignments, especially where the karst lies below recent soil deposits including, micro-gravity surveying, ground penetrating radar, electromagnetic conductivity surveying, electrical resistivity imaging, seismic reflection surveys and soundings, and cross-borehole radar surveying and seismic surveying.

It is necessary to understand the nature of the target of interest to determine whether it will contrast from its surroundings or “stand out” in the geophysical survey data set. When planning a karst detection survey the likely range of depths, lateral dimensions, vertical dimensions, nature of overburden, degree of infill etc. of the typical target cavities should be assessed beforehand. This allows a detailed, site-specific appraisal to determine which geophysical techniques may be of value in site karst detection. Numerical modelling can be used to assess the ability of a technique to detect karsts having the typical properties and dimensions that are of concern in the project.

Combining the results from several methods allows a clearer picture of the subsurface, by helping to distinguish between the signature of the karst features within the data, and other ground features.

Boreholes should be used to confirm the geological model and ground conditions at key locations along the route. In bare karst areas horizontal directional drilling or horizontal drill holes can confirm the ground conditions along the actual tunnel alignment. The use of cross hole seismic and tomographic methods to identify voids could also be used for areas where void definition is important and the risks are high. Cheaper probe drilling techniques may be used to confirm cavities in critical areas where better definition of the cavity is required.

Surveys can also be carried out within the cavities. Downhole CCTV and pulse laser surveying techniques can be used in dry cavities and ultrasonic surveying in flooded cavities. These can provide useful confirmation of the extent, size, direction and type of cavity. Pilot bores may be useful where risks are high, but costs are excessive and need to be considered before implementation.

Tunnel construction

Drill and blast would appear to be more flexible in bare karst areas in dealing with cavities and ground variability. Probing ahead can be carried out more thoroughly and proper fan probing around the tunnel profile from crown to invert overlapped and developed for the whole tunnel alignment. TBMs are more restricted for probe drilling. A TBM required to bore karst should be designed with whole tunnel circumference probing. TBM tomography techniques should be considered for early warning of cavities.

Mitigation measures to deal with karst hazards depend on early identification. If the hazards identified early present a large risk to the tunnel, then a realignment could be considered. In significant karstic areas this may not be possible and planning and implementing mitigation strategies is the best way of dealing with such problems.

Where known features such as cave networks are identified along the alignment, then pre-treatment can be carried out where direct access can be gained to carry out strategic concreting, support or infilling with appropriate materials to allow the tunnel to be excavated through the treated zone. Internal barriers and walls can be constructed if the cave network is large enough. Grouting and void filling can be carried out with bulk fillers and other proprietary systems to seal fissures and cavities from surface probeholes and from probe drilling at tunnel level during construction. Bypass tunnels, ground freezing etc. may be used depending on the scale and type of hazard encountered. Where soft clays infill cavities, ground treatment such as jet grouting can strengthen the ground and allow the tunnel to penetrate the treated cavities.

Conclusions

There is an increasing need to develop infrastructure in China. The large amount of karst terrain within and close to urban areas, means that this type of ground will continue to influence tunnelling. Karst problems should be well understood and its impact on tunnel projects with regard to direct and indirect hazards should be clearly investigated and studied.

Tunnelling in karst terrain has many hazards in addition to the usual range of normal tunnelling hazards. The prediction of ground conditions along a tunnel in karst is extremely difficult, a phased approach to investigation and background study searches should minimise such risk. Probing ahead of the tunnel is essential in karst and contingency procedures and systems should be developed to deal with a range of different hazards that could be encountered during construction of the tunnel.

Related Files
Hazards associated with covered karst tunnels
Hazards associated with tunnelling through bare karst conditions
Map showing China’s carbonates