AS CHRISTMAS 2013 approached, the 17.5m Hitachi Zosen EPBM on the SR99 project in Seattle received an unwelcome present as it advanced into an unknown obstruction and stopped in its tracks. Washington State Department of Transport (WSDOT) programme administrator Todd Trepanier told local press that the world’s largest TBM wasn’t stuck, but had been halted by a mystery obstacle, later identi_ ed as an 8in (200mm)-diameter pipe after difficult probing. Dewatering wells were sunk, the ground was improved and any voids stabilised, and then a 1.5m-diameter vertical bore from the surface removed the pipe debris. Arguments rage over the cost of the delay, and responsibility for it.

Meanwhile, local media lampoons the tunnelling industry’s biggest new toy, and the engineers in control, for being stopped by so tame an impediment.

Standard geotechnical investigations can only do so much – exploratory drilling from the surface or into the face can only tell geologists the nature of the rock in the boring diameter.

In addition, surface geotechnical investigations are limited by topography, while exploratory drilling from the tunnel face is another delay, and cautious, extensive probing even more so. A risk-balancing act ensues: whether the contractor advances tentatively, assuming the worst and carrying out extra investigations at a cost, or act in accordance with the known data from geotechnical surveys.

There are options to mitigate this. Real-time and semi-realtime geotechnical prediction technologies allow advance to continue relatively unabated, without having to "drive blind".

LOOKING BACK AND SEEING AHEAD _ THE BIRTH OF GEOLOGICAL PREDICTION

It all started, according to Amberg Technologies’ geophysics business unit manager Thomas Dickmann, in Switzerland in 1990 with early geological investigation work on the 21km Vereina Tunnel.

The demands placed on the engineers for forecasting technology led to the early development of the company’s Tunnel Seismic Prediction (TSP) technology for detecting geological discontinuities ahead of the face.

The year 1994 saw TSP 202 launched, a precursor to TSP 203 in 2001 which featured three-component signal reading, and allowed the evaluation of both P and S waves.

This meant rock mechanical parameters such as Poisson’s ratio, bulk and shear moduli, and Young’s modulus of the rock ahead was available to the contractor. Last year, the company unveiled its ‘3D’ system, TSP 303.

This newest system integrates 3D data acquisition and processing software, containing routines for optimal seismic imaging with respect to tunnelling requirements. It exploits the information in the seismic wave field by separate compression (P) and shear (S) wave analysis and the 3D-Velocity based Migration & Reflector Extraction technology (3D-VMR).

The 3D-VMR technology provides an adequate and detailed 3D image of the ground leading to a more reliable interpretation compared to conventional 2D approaches (Dickmann and Krueger, 2013). Sound principles

Given the limitations of probe/core drilling as an investigative measure, there was a need for a wider field, deeper-penetrating assessment of geology in advance of the face.

…If this could be done with minimal downtime, that would be a key bonus.

The procedure takes 60-90 minutes, and so is ideally taken in brief production breaks. It involves a series of 24 ‘shots’ containing 0.025-0.1kg of explosive to be set off behind the excavation face, which causes an acoustic wave to project. Four sensor probes consisting of tri-axial receivers are contained in protection tubes, the ends of which are cemented into boreholes of 45-55mm in both sidewalls. These detect discontinuities in the return wave – which are caused by geological interface-caused reflections – and pass the data to an evaluation software package. Results are apparent within a matter of hours.

On the small bores required, Dickmann muses, "The use of a precast segmental lining does mean a restriction for seismic surveys, since the rock mass isn’t accessible at all. In order to avoid large-scale drilling measures through the segments, it’s helpful to use the grouting and lifting inserts [for shot or receiver boreholes]."

The company states that as long as segments are 300mm thick, there is no chance of a stability issue. Induced loads from blasts are momentary and there is only a chance of small fissures. Regarding the analysis of returning waves, Dickmann adds, "Elastic body waves passing through homogeneous, isotropic media have well-defined equations of motion.

Utilising these equations, computations for the wave speed may be uniquely determined. Field surveys can readily obtain wave velocities, Vp and Vs; velocities are in units of length per time, usually meters/second (m/s). A homogeneous, isotropic medium’s engineer- ing properties of Young’s or elastic modulus (E) and shear modulus (G) and either density (r) or Poisson’s ratio (n) can be determined, if Vp and Vs are known.

"The units of these measures are: moduli in pressure, usually Pascal (Pa); density in mass per volume, grams/ cubic centimetre (g/cm3) and, n, dimensionless.

n = [(Vp/Vs)2 -2]/{2[(Vp/Vs)2 -1]}

Poisson’s ratio E = rVp2(1-2n)(1+n)/(1-n)

Eelastic modulus G = E/[2(1+n)]

Shear modulus rho= G/Vs 2

Density

"Density rho can be determined from samples or from boring samples or downhole logging. Estimates may be assumed by material type. Laboratory values may be offered of all these parameters, but field values will vary considerably from the lab estimates, because intact rock behaves differently to rock mass in-situ."

Passivity at any price (a non-percussive option)

Bore Tunnelling Electrical Ahead Monitoring, better known by its acronym Beam, was patented in 1998 and has been in use on projects, in various forms, since the Lötschberg base tunnel in 2000.

As with TSP, Beam targets the main geotechnical problems of TBM tunnelling, i.e. the sudden occurrence of unexpected high porosity ground (e.g. cavities, karst rocks, fault/fractured zones, and coarse material). Aside from the long standstills that can be avoided through the reduction of major risks, advance rate increases are also possible through the reduction of probe drilling occurrences, and also prior warning of the nature of ground to be traversed allows for optimal tunnelling practice to be adopted.

Beam method

Beam combines the principles of focusing electrode logging (Davies et al, 1992) (Schlumberger, 1972), and frequency domain induced polarisation (IP) measurements [5]. Low frequency alternating electrical fields are generated by galvanic injected currents through an excavation-specific focusing electrode configuration.

The electrodes being either the cutter wheel as a giant measuring electrode in the basic ‘BEAM Integral’ package, or additional cutting tools being adapted as electrodes in the further ‘BEAM Scan’ package. This allows more detailed imaging of 2D and 3D targets, as well as an advanced lateral resolution ability.

By adjusting the same voltage of the same polarity simultaneously between the guard electrode A1 (+) and the return electrode B (-), and between the measuring electrode A0 (+) and the return electrode B (-), the measuring current is forced ahead of the face, even if the electrode resistance between A0 and A1 is very low [3].

Therefore the realised constant measuring conditions of applied frequency dependent voltage U (f) and electrode configuration cause the magnitude of the sensed measuring current I0 (f) to be solely influenced by the electrical properties of the ground. Thus, when the tunnel face is advancing towards a ground change, I0 (f) directly images the ‘coming’ new geological situation. The obtained measuring parameters are the frequency dependent resistivities:

R (f1) = k x (U (f1)/I0 (f1)

R (f1) = k x (U (f1)/I0 (f1)

With: K= constant, and the PFE (Beam) = 100 x (R(f1) – R(f2))/R (f1) [%],

with frequency f1 < f2 For imaging the electrical field distribution around a tunnel, and forwards relative to the bore, a 3D finite element model is carried out using a software package, ‘Fracture’ [6]. The model simulates an excavation with one measuring electrode A0 positioned on the face and the enclosing tunnel diameter large guard electrode A1 (armed lining, shield mantle) encountering different ground changes and 2D/3D targets compared to a rock mass without ground change. The distribution of the electric field and the current vectors are presented in distances of 1 and 10m within successive sections ahead of the face.

The assessment of modelling results and geological experience from tunnel projects are indicating a distinct sensitivity zone for ground changes in a prefield distance of about three times the guard electrode A1 diameter (three times the tunnel diameter).

Components of BEAM include:

  • The Beam geoelectric unit in the TBM cabin
  • Measuring electrodes
  • Guard Electrodes
  • Return electrode

Tears for fears

For contractors operating in hard rock, high porosity zones are of extreme relevance to the drive.

In soft ground, clay layers, sand/gravel aquifers, pyroclastics, wood, boulders, piles and archaeological remnants are of interest for excavation. And Regarding 200mm steel pipes of the sort encountered by the coastal SR99, Wolf Boening says, "[Although] in general [the] Beam real-time ground prediction system is a tool for minimising… geo-hazardous risks [in particular], of course obstacles like wood piles or metal pipes could be detectable and indicated as warning on [the] Beam screen, whereby a sufficient contrast to consisting geology is required."

Boening adds, "Our main advantages are the disadvantages of seismic predictive systems like Amberg.

"Beam is installed only once onto any TBM type, and independent of any geology.

It performs permanent automatic high-resolution forward prediction and perimeter exploration while the TBM is boring. There is no disturbance of tunnelling work and the TBM doesn’t need to stop for any measurements. Every stroke is measured and directly evaluated as well as indicated. Beam’s measuring unit is placed in the operator cabin and it has an integrated screen for indicating real-time ground prediction results ahead of the face.

"The contractor receives continuously update forecast ‘pictures’ at a distance of three-times the tunnel diameter ahead of the TBM. At the same time, the Beam system can be controlled from outside the tunnel by using any accredited computer for showing simultaneously forecast results. Also online maintenance, or potential online support is possible by the manufacturer (us), if required.

"When using a seismic system the tunnelling work has to stop, and also TBM needs to stop rotating during measurements. Results are not real-time and experts are needed to evaluate the data.

"In our opinion you can’t achieve high advance rates with seismic systems. Modern tunnelling under time and cost pressure should use a real-time technology like BEAM, if geophysical measures are required."

Beam case study: Sparvo highway project

The second largest TBM in the world was manufactured to construct a new bypass for the A1 motorway in Italy. The project will alleviate congestion on the key route between Naples and Milan in the Emilia-Romagna region.

A major project though complex conditions, such as those in the Appenines, consisting of frequent squeezing clay sections, methane presence and fault zones require a deep and precise knowledge of the geological scenarios to be face with the TBM.

The BEAM system, developed and provided by Geo Exploration Technologies (GET), was applied as a geophysical method for real-time ahead investigation continuously during the tunnelling activities.

Two operation modes are available. The integral mode is the basic system for 1D forefield exploration of resistivity and IP-data along the tunnel. The scan mode provides 2D forefield exploration perpendicular to the tunnel axis for lateral resolution.

System setup for high-diameter clay drive

The system setup for the Sparvo project comprises the following main components:

The BEAM measuring instrument performs fully automatic data acquisition, processing, and visualisation simultaneously for realtime data presentation including self- explanatory geological-hydrogeological interpretation. It is placed in the TBM steering cabin with cable connections from the electrodes, guidance system (VMT), the PLC (boring signal switch), rotary encoder, network internet access and uninterruptible power supply.

For the BEAM integral mode the whole cutter head is used as one large "moving" measuring electrode (A0Int) coupled to the face due to the rotational forward advancement of the TBM.

For the BEAM scan mode 8 insulated cutters are used as rotating and forward moving measuring electrodes (A0). The cables are protected inside channels and connected through an electrical rotor. Rotational electrode position is provided by a rotary encoder.

The guard electrode (A1) consists of the ground coupled TBM shield with fixed contact and cable connection for the whole drive time.

The return electrode (B) is earthed by a steel rod in a far distance to the TBMface outside the tunnel, leaving fixed for the whole drive. Thus the B-cable has to be prolonged from time to time.

Two switchable chart screens are provided displaying measuring results of the integral mode and the scan mode. The integral chart represents in the upper part the "Ground Change Indicator" graph along the tunnel axis with the resistivity curve and the PFE (IP-parameter) curve build-up of survey points allocated about 47m ahead of the face (3-times the Sparvo TBM-diameter). These are ‘moving’ stroke wise from right to left always refreshing when new data is recorded during the drive. The actual face (yellow line) divides between forecast data and passed data.

The bar chart visualises the changing ground classification during a drive, which is based on resistivity and PFE combination classes. The attached interpretation table in the lower part is customised according to the geology of the Sparvo project.

The scan chart is based on the measuring data obtained at the rotational survey positions of the A0 cutter electrodes. It provides forecast and passed PFE contour plots within spherical cross-sections successively aligned by a stroke wise spacing.

Because of the reciprocal relationship between both polarisation and porosity it indicates a relative porosity distribution whereby high porosity (low PFE) anomaly zones containing potentially open cavities and open fractures could be laterally located.

Application

To operate under the prevailing complex ophiolithic geological conditions a customised petrophysical classification was developed. It is based on the Beam cross correlation matrix using four PFE ranges (P1 – P4) and three resistivity ranges (R1 – R3) resulting in 12 combination classes (P/R) reflecting hard rock and soft ground, fault and fracture properties as well as water-inflow and gas-inflow potential (methane in this case). These P/R-classes are incorporated within the dynamic bar chart using the matrix colour/hachure combination and the geological and hydrogeological characterisation within the interpretation table of the integral screen.

Resistivity and PFE measuring data which are falling into the P1/R1-R3 matrix fields are indicating the most driverelevant faulted, fractured and brecciated zones. Clayey and weathered ground will be mainly expected in the matrix fields P3-P4/R1 and compact rock corresponds mainly to the P4/ R2-R3 matrix fields.

Along the 2,500m long TBM-drive there have been several successfully early warning forecasts of critical rock sections, which have been confirmed and safely cleared up by using different geotechnical tunnelling measures

Conclusion

The technological improvements in passive (and semipassive) geological predictive methods are well publicised by the suppliers, particularly in the case of TSP. And the cost as a percentage of project value, or measured against any reasonably major shutdown from geological surprises seems almost negligible.

Encouragingly, there are advantages to both systems, which seem to cover the other’s disadvantages – and neither excludes the other. For mountain projects with difficult investigations conducted from the surface, both could be vital inclusions for the contractor’s arsenal.