Turkey’s Tough Geology Inspires Robust Machine Design

The Genesis of Crossover TBMs at Kargi
By The Robbins Company


Driven through a mountain side with 600-m of cover in Central Turkey, the Kargi Kizilirmak Hydroelectric Project is one of the most challenging tunneling projects ever completed in the region. The project is located on the Kızılırmak River, near the Beypazarı district of Ankara province in Turkey. The Kızılırmak (Turkish for “Red River”), also known as the Halys River (Ancient Greek), is the longest river in Turkey. The project includes a 13-m high, 450-m wide earth dam with a concrete spillway in the southern end and a water intake construction on the northern side. The dam creates an artificial lake with a high-water head of 405 meters above sea level. A headrace tunnel diverts the water from the dam to the powerhouse situated east of the village of Maksutlu. The water from the powerhouse will flow into the Boyabat reservoir at 330 meters above sea level. Once online, the project will generate 470 GWh of power annually for project owner Statkraft, an amount sufficient to supply approximately 150,000 homes.


Initial Machine Supply & Geology
The project area is located within the Northern Anatolian Fault System, which is primarily responsible for earthquakes in Turkey. The tunnel was driven through a mountain side with up to 600-m of overburden. The expected geology along the tunnel alignment consisted of Kırazbası complex Kargı ophiolites (including sandstone, siltstone and marl) for the initial 2,300 meters, followed by 1000 meters of Kundaz metamorphites (including marble, metalava and metapelite), and the remaining 8,500 meters consisted of Beynamaz Volcanites (including basalt, agglomerates and andesite). The anticipated strength of the rock was up to 140 MPa. Multiple fault zones and transition zones added to the complexity of the geological conditions.


The Robbins Company supplied a 9.84-m diameter Double Shield TBM and continuous conveyor system to Turkish contractor Gülermak for excavation of the 11.8-km headrace tunnel in 2012. Initial geological reports predicted softer ground for the first 2.5 km, which would be lined with pre-cast concrete segments. The remainder of the tunnel was to be supported by a combination of shotcrete, rock bolts, and wire mesh.

Excavation & Difficult Geology
The machine almost immediately encountered geology that was substantially more problematic than was described in the geological reports. The geology consisted of blocky rock, sand and clays. “The Kargi TBM was launched into poor geology in early 2012. After boring 80 meters, the TBM became stuck in a section of collapsed ground that extended more than 10-m above the crown, loaded onto and around the cutterhead,” said Glen Maynard, Robbins Site Manager. As a countermeasure that was immediately put into place to avoid the cutterhead becoming stuck in the blocky material, crews began boring half strokes and half resets.


Even with these measures, the machine encountered a section of extremely loose running ground with high clay content. A collapse occurred in front of the cutterhead and the cathedral effect resulted in a cavity forming that extended more than 10-m above the crown of the tunnel. The weight of the collapsed material trapped the cutterhead. After several unsuccessful attempts to clean out and restart the cutterhead, consolidation of the ground above and in front of the machine was carried out. Injection of polyurethane resins via lances inserted through the cutter housings and muck buckets was the method utilized for consolidation operations; however, injection locations were restricted to the available openings and subsequent attempts to restart the cutterhead proved to be unsuccessful.

After assessing all the available options, it was decided that a bypass tunnel would be required. Robbins Field Service assisted Gülermak with bypass tunnel design and work procedures to free the cutterhead and stabilize the disturbed ground. Blasting techniques were ruled out due to concern over further collapses caused by blast induced vibration; hence, the excavation was undertaken using pneumatic hand-held breakers. Bypass tunneling was successfully completed and at that point the section of bad ground was believed to be an isolated one.


These hopes were proved wrong, however, as six more bypass tunnels were needed within the first 2 km of tunneling. At this point it became apparent that the actual geology was far more complex than originally stated. Both the contractor and manufacturer worked together to develop and improve bypass tunneling and hand tunneling techniques, resulting in an average bypass tunnel construction time of just 14 days. All tunnels were completed safely and in a timely manner, though there were of course significant delays associated with the downtime. Despite the setbacks of these multiple events, the TBM did succeed in crossing numerous faulted sections that would have trapped a machine with less power. In fact, the crew measured cavity heights above the cutterhead in some of these fault zones at over 30 meters.


TBM Modifications & Significant Progress
In order to improve progress in the difficult conditions, the contractor, owner, consultants and Robbins engineers worked together to generate solutions. The contractor, with the assistance of the field service team, installed a Robbins custom-built canopy drill and positioner to allow pipe tube support installation through the forward shield. Drilled up to a distance of 10-m ahead of the cutterhead, 90-mm diameter pipe tubes provided extra support across the top 120 to 140 degrees at the tunnel crown. Injection of resins and grout protected against collapse at the crown while excavating through soft ground. As a result of successful use of the probe drilling techniques, Gülermak was able to measure and back-fill cavity heights above the cutterhead in some fault zones to over 30-m and, in addition, was able to help detect loose soil seams and fractured rock ahead of the face.


To further mitigate the effects of squeezing ground or collapses, custom-made gear reducers were ordered and retrofitted to the cutterhead motors. They were installed between the drive motor and the primary two-stage planetary gearboxes. During standard boring operations, the gear reducers operate at a ratio of 1:1, offering no additional reduction and allowing the cutterhead to reach design speeds for hard rock boring. When the machine encountered loose or squeezing ground, the reducers were engaged, which resulted in a reduction in cutterhead speed, but the available torque was increased by nearly double. The net effect of the modifications was to allow the Double Shield TBM to operate much like an EPB in fault zones and squeezing ground with high torque and low RPM—these methods effectively kept the machine from getting stuck. In addition, short stroke thrust jacks were installed between the normal auxiliary thrust to double total thrust capabilities.

What is a Crossover TBM?
Crossover TBMs are also called hybrid or Dual Mode machines and are able to “cross over” into distinct types of geology. Crossover TBMs feature aspects of two TBM types and are ideal for mixed ground conditions that might otherwise require multiple tunneling machines. The XRE (standing for Crossover between Rock/EPB) features characteristics of Single Shield Hard Rock machines and EPBs for efficient excavation in mixed soils with rock. Other types of Crossover machines include the XSE (Crossover between Slurry/EPB) and the XRS (Crossover between Rock/Slurry).

“The cooperation and trust between the contractor, project owner, and Robbins Management, Engineering and Field Service resulted in the correct modifications being successfully installed on the Kargi TBM,” said Glen Maynard, Robbins Site Manager.

Despite the slow progress initially, the Robbins Double Shield TBM made some remarkable advances once modifications were in place, which essentially made it a Double Shield TBM with Crossover features. An advance rate of 600 meters in one month was achieved in March 2013 and a project best of approximately 723-m was achieved in spring 2014, including a daily best of 39.6 m in April 2014 (see Table 1 for advance rates). In doing so, the TBM significantly outperformed a drill and blast heading progressing from the opposite end of the tunnel. Crews at that heading progressed in relatively good ground conditions for 4 km, where they achieved advance rates of nearly 300 meters per month. The TBM bored 7.8 km of the tunnel in total, making its final breakthrough in July 2014.

Table 1: Average Advance Rates in Various Ground Conditions
	TBM	TBM	D&B
	(Before Modifications)	(After Modifications)
Good Stable Rock	N/A	22.5 m/day	8 m/day
Fairly Stable Rock	9.7 m/day	19 m/day	5.5 m/day
Non-self-supporting rock	4.7 m/day	8.5 m/day	1.5 m/day
Running/squeezing ground	1.3 m/day	1.8 m/day	N/A

Implementing the Lessons Learned
The TBM design changes gleaned from Kargi were not considered to be isolated, and immediately began to inform the design of Robbins’ new line of dual-mode type machines, termed Crossover TBMs. Mexico City’s TEP II Project is a complex wastewater conduit that promises some challenging ground conditions during excavation. The 5.9 km long tunnel is expected to include volcanic rock such as tuff, lava, and breccias, giving way to abrasive sand and clay in the last 10% of the tunnel. “Kargi was a great experience for us and an example of unexpected geology that now informs some of our engineering decisions. We went through a rock mass with a Double Shield TBM, but it would have worked best using a Crossover machine or even EPB—the actual geology turned out to be very poor and weathered rock, it could not really be called solid rock,” said Robbins Project Manager Martino Scialpi.

Robbins has thus provided an 8-m diameter Crossover TBM, optimized toward rock excavation, for the ALDESA/PROACON/RECSA JV. The canopy drill from Kargi was pre-installed at TEP, providing another ring for probe drilling close to the cutterhead, or for fore-poling, if needed. The canopy drill, like at Kargi, can operate in the top 120 degrees of the tunnel, while a second probe drill is located further back on the machine. Two different patterns of holes are thus present: one in the rear shield for probing and one in the forward shield for fore-poling.

Extremely high torque/breakout torque is another feature added to the TEP machine after lessons learned at Kargi. “We are expecting soft ground under pressure with a rock section and water in the middle of the TEP—so we have two-speed gear boxes that can be activated to achieve higher torque at a lower speed, like how an EPB operates,” said Scialpi. Two-speed gear boxes can be activated in bad ground so as to free the cutterhead. If the cutterhead becomes stuck, the only way to free it is with a bypass tunnel, which can hopefully be avoided with the modifications. The 8.7-m diameter Crossover XRE (a type of EPB/rock Dual Mode machine) was launched in August 2015 and has been progressing well.


Conclusions & Thoughts on Tunneling in Difficult Ground
Prudent pre-planning by all parties involved is a common theme running through most case studies of challenging conditions. Planning includes conducting accurate geological testing and developing a detailed Geotechnical Baseline Report (GBR) in order to create a tunnel design and specify a construction method that can mitigate risks even when difficult ground is expected. In most cases, the client or owner takes care of geological investigation, but the contractor takes the risk, while equipment suppliers are asked to share risk with the owner. In difficult ground, it is prudent for contractors and TBM suppliers to become highly involved in the geological investigation, for the best interest of all parties involved.

Choosing the proper TBM type is another piece to the puzzle: High performance tunneling in hard rock requiring little or no long-term ground support is best performed with open-type, main beam TBMs. In many projects, however, where the majority of the formation is in good ground, there may be areas of faulted zones or otherwise poor ground. These challenging ground zones can be a real obstacle for main beam machines.

The problem of selecting the proper machine type is only solved using a balanced approach: The type of TBM selected should not be unduly constrained by cost or scheduled TBM delivery; factors that can be politically driven. Fully leaving the choice of means and methods to the contractor in a very tight cost environment may be problematical for unique or state-of-the-art projects, however. Furthermore, specifying machine characteristics, or even the owner pre-purchasing the machine, may also be problematic. Each agency has to find a balance between prescriptive and performance requirements for complex projects with unknowns.

Once geological tests have been conducted and the machine design has been carefully weighed and selected, there is still potential for unknown ground conditions. With new developments such as Crossover TBMs, the risk of varying zones of ground can be mitigated as widely different types of geology can be excavated with one TBM. As the industry evolves to meet the challenges of mixed ground, and becomes more educated through increasing numbers of projects and case studies like that from Kargi, the risks of tunnelling in such conditions will decrease.

NBM&CW October 2019