Concepts for handling changing muck properties caused by TBM excavation

Transcription

1 ATS Concepts for handling changing muck properties caused by TBM excavation By Dr. Ulrich Rehm, Tunnelling Consultant GmbH; Germany & Davoud Falati 1. INTRODUCTION It is a well known practice most of the claims on tunnelling are due to unexpected geological conditions whereas in many cases where TBM have been used it should be clear that ground properties as described in the Geotechnical Baseline Report (GBR) are very likely to be changed during the excavation by the TBM. So a contractual line has to be drawn between the term unexpected ground conditions and expected permanent changes of muck properties caused by the TBM. From a contractual point of view it is very important for both sides owner and contractor to have the best assessment on the ground properties in the GBR which have been gained from traditionally quasi-static exploration tests in the laboratory. However, the excavation and mixing of the ground in the working chamber of a TBM is far away from being quasi-static but are highly dynamic to chaotic why traditional soil mechanics have to be judged critical in this case. Quasi-static ground properties change under dynamic conditions due to the complex 3-phases system of the ground (solids, water, air). The economic prediction of the TBM performance and efficiency of a tunnel project is mainly depending on these ground properties changes because they determine geotechnical phenomena like clogging, abrasion and face stability which are most important for the contractor s assessment. Furthermore he has to predict the penetration of the cutting tools into the face which again is a highly dynamic process. Since every TBM is unique and furthermore the human factor during operation of the TBM comes into play a generalization on how it will react on geotechnical phenomena can not be made. But the client has to expect from the contractors who are bidding on his project that they are aware of the ground property changes during excavation of the TBM and the associated phenomena. Therefore fundamental practical experiences of the contractor on how the ground will behave under dynamic conditions are the most important base for a realistic prediction of TBM performance and efficient TBM application. 2. GEOTECHNICAL PHENOMENA IMPACTING TBM TUNNELLING Both systems ground and TBM are at the same time two fascinating and very complex systems that have to be understood each separately by practical experiences as a base for assessing, planning, designing and consulting on TBM tunnelling. Traditional soil mechanics are pretty well understood within quasistatic ground conditions. In contrast to that TBM tunnelling happens under highly dynamic to chaotic conditions at the tunnel face and in the working chamber. Therefore approaches of assessing the dynamic interface reactions between ground and TBM on the base of traditional quasi-static soil mechanics should be questioned. The geotechnical phenomena which have important impact on the performance of a TBM are: - clogging - face/ground-stability - wear and tear - ground-water flow - electro-chemical loadings in the ground - convergency These geotechnical phenomena have to be predicted on the base of the GBR by the contractor who has to give his best price estimation on how to handle them

2 having in mind that the cheapest quotation will make the race. The geotechnical characteristics of the ground are depending on its composition of the 3-phases solids, ground-water and air and furthermore on the actual stress situation within the ground which determines the consistency of cohesive ground and the bedding of non-cohesive ground. The consistency and bedding are most important information for TBM tunnelling in softrock and vary permanently during the excavation. Little changes in water-content can turn non-clogging brittle hard clay into a sticky soft paste and the power of a 100-tons-cutterhead removing a grain out of a 10 m stable highly compacted tunnel-face can give the starting signal for a running avalanche. One approach to predict the sensitivity of soil changing its consistency is the plasticity index which does not take into account the actual stress situation. Based on [1] clogging is mainly depending on consistency and plasticity of the ground whereas the highest tendency for clogging appears at firm to stiff consistency for plasticity indices greater than 20%. The spoil of an EPBM should therefore be controlled within a certain range to reduce the risk of clogging. 2.1 EPBM It is a soil mechanical fact that the change of porewater pressure has a severe influence on the effective soil stresses which are responsible for the friction between the particles thus for face-stability and wear on the TBM. Furthermore the consistency of cohesive soil is influenced by the actual stress situation. For EPBM the active face-pressure in the working chamber has a certain inaccuracy of approx. 0,3 bar due to the dynamic and inconsistent mixing process of the spoil. This has impact on the pore-water pressure in the working-chamber and with that on the effective soil stresses and hence on the face-stability, the wear and the consistency of the muck. The more friction appears between the particles the more heat is generated which can lead to very disadvantageous effects in the working chamber like cooking and concreting of muck (affecting negatively the main-drive sealing). The addition of chemical additives like foam can decrease the effective stresses between the particles and therefore reduce friction and wear. The foam bubbles increase the air-phase within the ground which impacts on the pore-pressure and acts like elastic feathers helping to compensate face-pressure changes quickly (air-cushion effect). This requires stable very little homogenous distributed foam bubbles in the spoil. Despite spoil treatment with foam is a very helpful tool it has to be avoided to over-foam the chamber which can have negative effects due to loosing too much friction between the grains (low effective stress) resulting in a collapsing liquefaction of the face. In this context the contractor has to be familiar with the different operating modes of the foam generator of the TBM which controls the amount and the quality of the foam. Fig. 1: Classical geotechnical phenomenon clogging (left), disc cutter wear (right).

3 Fig. 2: Estimation of clogging behaviour after [1]. Fig. 3: Stable foam with homogenous small bubbles (left); instable foam due to inhomogeneous bubble size distribution (right). 2.2 SLURRY TBM In case of a slurry TBM the characteristics of the ground at the face are changing totally from solid into predominantly liquid immediately after being excavated and dissolved in the slurry. The original soil mechanical properties base lined in the GBR are far away from the properties of the spoil dissolved in the slurry. In this case the geotechnical phenomena of the spoil are supposed to result mainly from the phenomena related to the swelling clay types of the Bentonite-slurry and their interaction onto the spoil. This is again depending on the reaction of the claytypes (Smectite, Illite, Kaolinite) on water and furthermore by their electro-chemical sensitivity. The changing of electrical loadings within the slurry determines the interaction between the dissolved solids and with that the geotechnical phenomena like segregation, clogging and abrasion. These electrical loadings come from the nature of ground minerals (Cat- and Anion), the quality of the ground-water (phvalue) and chemical additives. The quality of handling the Bentonite slurry has significant impact on the quality of the tunnelling performance. The efficiency of a slurry TBM can only be as good as the Bentonite slurry technology is well understood and handled in an appropriate way. The geotechnical phenomenon of generating a so called mud cake at the tunnel face has decisive impact on the face stability and is existential for the stuff during maintenance works under hyperbaric conditions in the chamber. These phenomena of the slurry have to be well understood and controlled by the contractor otherwise it exists a high risk of fatality for the personnel during chamber interventions. Therefore frequently daily quality checks of the Bentonite slurry on-site in terms of viscosity and density as well as an appropriate separation technology are inevitable for an efficient tunnelling with a slurry TBM. The comprehensive knowledge of handling Bentonite slurry technology is also important for EPBM where often a certain amount of Bentonite slurry is used for spoil treatment.

4 2.3 HARDROCK TBM For TBM tunnelling in hardrock the rock mechanical properties of the original ground described in the GBR are also changing severely after being crushed by the disc cutters. Base lined uni-axial compressive strength and indirect tensile strength of the rock are an important base for assuming the penetration of the disc cutters into the rock and the primarily wear (direct crushing of rock). For the assumptions of the secondary wear (flow of crushed rock) which has a big impact on operational costs the strength properties of the rock does not give reliable indications. The geotechnical phenomena of wear on cutters is mainly influenced by the mineralogy of the rock, the chip and mineralogy size, the moisture content of the muck, the steel resistance of the components (quality of materials), the position of the tool on the cutterhead and the constructability of the cutterhead (smooth muck flow). The rock which is directly in contact with the cutterring in the so called crush zone turns into a fine power after being crushed. Depending on the geological genesis of the rock for sediments the amount of fines resulting from the crushing process can amount up to 30-50% oft the initial rock mass. Adding only a little amount of joint water can turn these fines into a highly abrasive paste which can furthermore jam the muck channels within the cutterhead. In the worst case the channels get blocked and have to be cleaned frequently. These phenomena can not be directly derived from the base-lined rock properties but have to be assessed by an experienced contractor. Unexpected high amount of ground water during a hardrock TBM drive and/or the occurrence of natural gases like methane or hydrosulphides is something which is difficult to predict but can have very negative impact on the whole tunnelling. A permanent measurement and warning system on the TBM is strictly recommended. From the TBM side most innovative techniques (housed components, emergency containers, water-mist curtains, additional ventilation, Ex-protection etc.) should be used for mitigation. Since such incidents are very difficult to derive quantitatively from the GBR they should be handled on a partner-ship like base between contractor and client. Geotechnical phenomena like squeezing due to over-stressing of the tunnel profile or swelling caused by water absorption can lead to dramatic convergences causing severely damage to the TBM and/or the tunnel lining. Based on several case histories and published research works in the field of rock-convergences the contractor has a base for planning tunnelling activities under these conditions. Innovative flexible rock support systems and tunnel lining, as well as innovative TBM technologies and a comprehensive pre-planning of work in order to avoid stoppages at very critical positions can help to reduce the risk of getting stuck within critical rock zones. Predicting disadvantageous tunnel convergences requires experiences in the field of rock mechanics whereas for overcoming them it is much more important to have a well organized job-site logistics. The changes of the ground properties during the excavation by the TBM have to be expected and well understood and assessed by the contractor for a reasonably prediction of the performance and efficient operation of the TBM. In any case a cooperative partnership-like relationship between client and contractor is very helpful to overcome and manage critical project phases together. 3. OPERATIONAL AND LOGISTIC ASPECTS The properties of the ground as base-lined in the GBR are changing in many ways until the spoil reaches the surface. There are two basic different approaches to handle the geotechnical phenomena like mentioned above: procedural by spoil treatment and/or technically by innovative TBM design. The technical application areas for the classical TBM types are well known and every contractor being awarded a TBM project should have comprehensive experience to handle these very complex tunnellingfactories. Every TBM can only be as good as the experience of its user allows it for. In this context the owner of a TBM should be aware of the latest innovative technologies and components which have developed quickly during the last decade in order to achieve optimum performance. One of these innovative procedural developments is the chemical spoil treatment which helped to extend the classical application area of EPBM from cohesive grounds into more non-cohesive grounds where normally slurry-tbm would have been used. This development helped to save extensive space (and costs) for a separation plant and associated logistics at the job-site. However, the use of the various types of chemical additives causes a wide spread range of changes of the geotechnical properties of the spoil. The advantageous change of the consistency of the ground into more soft for an improved excavation can be disadvantageous for its transport by conveyor belt or musk skips. Every logistic chain from the tunnel-face to the dumpsite requires special spoil consistencies for best efficiency. In case of different contractors being responsible for different logistic chains a close cooperation between them should be practised in order to find most suitable and efficient spoil treatment in order to avoid redundant works. In case of adding several different additives into the chamber their combination can have reversely bad impact in terms of increased clogging or increased corrosion on TBM components. In case of impropriate handling of the muck consistency the whole muckingchain does not work efficient and a chain will burst at its weakest link. Therefore the contractor has to know

5 and understand the available possibilities for changing the properties of the spoil in order to achieve best efficiency for the mucking process and furthermore for the total tunnelling performance. Beside the chemical spoil treatment the mechanical impact of the very heavy cutterhead on the muck should not be underestimated. Cohesive spoil consists of a so called sensitivity which is like the behaviour of modelling clay that is being kneaded over a certain time. Due to the induced energy resulting from the friction the hard modelling clay gets weaker and softer with the time. Finally it can be squeezed through the fingers. The behaviour of cohesive muck being mixed in the working chamber is very similar to that. The very high mixing energy of the cutterhead can weaken cohesive spoil into a soft paste which is on the one side positive for proper EPB-technology but in case of getting too soft the tendency of clogging will increase. This effect of mechanical spoil treatment by the cutterhead gets more serious with increasing face pressure due to increased friction between the particles. In one case it could be observed after cutting through a concrete wall which generated a lot of fines of the concrete the heating-up of the muck during mixing in the chamber caused a re-curing of the concrete in the muck which had bad impact on the sealing of the main bearing. For slurry-tbm operation the dissolved spoil in the muck can cause a lot of secondary wear on all components that are in touch with the slurry especially the slurry-pipes and pumps. Most sensitive are curved pipes. A major impact in this context has the density of the discharged slurry and its flow velocity. In case in the GBR the amount of abrasive ground minerals are mentioned wear can be predicted. Special slurry circuit design can help to reduce the operational costs. The application of different types of hardrock TBM is closely linked to the geotechnical information in the GBR and the contractor s assessment regarding the stability of the rock. In cases of long tunnels some contractors prefer a double-shield due to its potential of high performance. Since the advantage of a doubleshield in terms of performance can only be gained on the base of a very good logistic system on the job-site the contractor has to balance between this alleged advantage and the obvious disadvantage of a too long shield getting stuck within squeezing sections. As mentioned above the geotechnical phenomenon of squeezing rock conditions are predictable and the GBR should help to rule out some of the risks. From an operational point of view TBM have to be operated different in hardrock and softrock respectively. In case of a purely hardrock tunnel project it is questionable whether a contractor with only experience on softrock TBM tunnelling should be awarded (and vice versa). The logistics of a job-site and the operation of a TBM have decisive impact on its performance which unfortunately is often not as intensive planned and executed like setting up a GBR. If this would be changed a lot of claims could be avoided. Fig. 4: Bentonite mud cakes, left: 4% slurry, middle: 5% slurry, right: 6% slurry.

6 Fig. 5: Primary and secondary wear on disc cutters. Fig. 6: Intersection of best consistency for a job-site logistic chain. 4. CONTRACTUAL ASPECTS Any base of a contract are the facts of the actual situation which allows all parties involved a most realistic prediction of the future situation and the monetary effort to get there. For tunnelling projects the main contractual facts are the geotechnical information in the GBR which form the foundation for the contractor s evaluation and assumption of his costs for tunnelling. Therefore the GBR is the most important document for contractual agreements. Unfortunately most of the higher management for contractual negations are no geotechnicians. This would be very helpful to reduce misunderstandings and the amount of claims because for tunnelling projects challenging conditions have to be dealt contractually which have a clear geotechnical background: - The TBM will be confronted with every Millimetres of the geological conditions within the alignment whereas the distance between the bore logs vary much more widely (not seldom hundreds of meters). - The soil mechanical facts of the ground have been derived from quasi-static exploration measures whereas TBM tunnelling is a dynamic process; so the quasistatic facts of the GBR can not realistically describe the future situation in terms of TBM tunnelling. - Standardized dynamic test procedures for a more appropriate determination of the expected soil properties as a base of more realistic facts do not exist; sporadic dynamic test procedures can only be executed by few (mostly research facilities) at fixed places; all testing procedures should be checked critically whether they are compatible to TBM tunnelling. - The geotechnical characteristics of the spoil will be very likely different to the base lined facts of the expected ground conditions in the GBR due to the dynamic mixing process of the TBM and due to the various possibilities of the contractor to change these properties by adding chemical additives. - The influences resulting from the tunnelling process range directly from the tunnel alignment until indirectly to the surface; any destructive short- and long-term reactions on the surface (which in most cases have a geotechnical reason) are under the responsibility of the contractor who has to be aware of the chainreaction of his TBM.

7 - The GBR describes expected ground properties before excavating whereas the contractor s daily bread is the spoil after being excavated by the TBM.From these items it can be derived that if it comes to claims based on geotechnical reasons one has to distinguish between initial base lined ground conditions at the face and spoil conditions immediately after and during the excavation. In many discussions claiming parties could not come to a consistent solution because one side was talking about virgin ground and the other side about the excavated spoil. The owner of a project should therefore execute a pre-qualification of contractors to select those who can handle the difference between unexpected ground conditions and expected spoil properties. Despite there is actually a big gap between the theoretically quasi-static soil mechanic parameters given in the GBR and the more realistic dynamic conditions on the TBM any experienced contractor should be capable to calculate and to assess his expected realistic performance and realistic costs. The past showed that the philosophy of bid low claim high leads to a bad reputation of the contractor, to a low quality construction with huge cost overrun and intensive maintenance and operation costs which finally have to be paid by the whole community. 5. CONCLUSION The tremendous technical development in TBM tunnelling during the last decade demand beside comprehensive wide spread technical knowledge and practical experiences from the contractor an appropriate understanding of the interaction between TBM, Geology, job-site logistics and the human factor of the personnel. Geotechnical phenomenon which have a decisive impact on the performance of the TBM, the quality and efficiency of the whole tunnelling have to be understood by the bidders in order to come up with a reasonably assessment of the ground conditions and to quote in a most realistic manner. This will lead to a better understanding of the terms unexpected ground conditions related to the GBR and expected changes of spoil properties caused by the TBM which is the foundation of a reliable quotation and a fair contractual relationship. REFERENCES [1] Markus Thewes: Adhäsion von Tonböden beim Tunnelvortrieb mit Flüssigkeitsschilden; Bergische Universität Gesamthochschule Wuppertal; Bericht Nr. 21; July 1999