Nepal is a country that can take a visitor by surprise. Mention it to most people, and it is the heart of the Himalayas. Perhaps the Himalayas and temples.

This view is not entirely unfair; the mountain range that spans the northern border of the country is the most spectacular in the world, and is a key component of Nepal’s geopolitical situation as a smaller power caught between the two giants of China and India. Combine that with the country’s several thousand recognised deities, the spiritual reputation too is well-earned.

Yet its bustling capital Kathmandu is now the third most expensive city in Southern Asia, contrasting with the monastic image and a sign that industry may be building to something And it is.

The country has an incredible resource in its topography. The water held in Nepal’s share of the ‘Abode of the Snow’ (the literal meaning of Himalayas in Sanskrit) is immense, more than the country’s population of 30 million could possibly need for drinking water and irrigation, and yet many of the country’s farmers have difficulty accessing enough water in the drier months.

Similarly, the hydropower potential could supply Nepal’s current needs and have leftover for an industrial renaissance.

Despite this, many of energy-importing Nepal’s regions suffer chronic power cuts. The problem: Nepal is stepped. It has a highly-populated southern lowlands, while the water and energy resources are currently locked away in the middle and upper heights.

The solution, which doubtless comes as no surprise to our more experienced readers, is tunnelling.

Nepal’s Ministry of Energy, Water Resources and Irrigation has identified that less than 10 per cent of the country’s surface water has been made available for irrigation. Most of the irrigation systems in the country draw on small- and medium-sized rivers, which have an excessive flow in the wet months (June-September) but many dry up in the remainder of the year.

Enter the Bheri Babai Diversion Multipurpose Project (BBDMP) in the Dang Deukhuri District of Midwestern Nepal. The hydrology of the region is dominated by two rivers: the mighty Bheri and the periodically dry Babai.

Local agriculture depends on drawing water from the Babai, so a tunnel is under construction to divert 40m3/s of the average 332m3/s Bheri flow into the Babai. This will unlock approximately 51,000ha of agricultural land, while also generating 48MW of electricity.

Although first identified in 1977, and despite the project being seen as one of the most obvious projects to pursue for remaking Nepal, progress has been slow. The country has had 11 prime ministers in the past decade, for example, which causes some political challenges.

But now Nepal has designated a series of ‘National Pride Projects’ which removes some of the administrative blocks to progress. It also marks a move away from small/medium-sized projects, to larger schemes. Things are beginning to move in the ‘Roof of the World’.


The tunnel itself is a 12.3km-, 5.06m-excavated-diameter drive incorporating a 152m change in level (including shaft) between the rivers. The drive involved curves with a minimum radius of 700m, maximum overburden of 820m and a maximum gradient of 3 in 1,000 upwards along the drive.

The powerhouse at the Babai end and the headworks at the Bheri form a separate contract, which is behind schedule compared to the tunnel.

It can be a point of pride for the industry that the client underestimated the efficiency of modern tunnelling techniques and so scheduled more time than has been necessary for the tunnelling, only to find that tunnelling is ahead of all other project segments. Breakthrough is scheduled for mid-March as Tunnels and Tunnelling goes to press. This is about one year ahead of schedule.

In fairness to the client, this is the first use of mechanised tunnelling in the Siwalik geology, a highly folded and faulted formation of the Young Himalayas, comprising sandstone, mudstone and conglomerate.

Variable rock mass quality characterises this formation, with bedding and foliation in the sedimentary and metamorphic rock making its behaviour highly sensitive to the direction of tunnelling in terms strength and deformability. This results in a considerable reduction in the self-supporting ability of the rock mass.

Additionally, climatic conditions in the Himalayas combined with active tectonic movements can cause deep weathering effects that can reach up to 100m below the surface, causing rock mass to lose cohesion and creating a risk of collapse during excavation.

Geology was relatively well-understood and as-described by the Geotechnical Baseline Report. Throughout the drive, whenever the contractor felt appropriate, probing through the shield was undertaken by an Atlas advance drilling rig. Although probing was not mandated by the contract under any specific conditions.

Two thrust zones cross the alignment, the Babai thrust (100m from the launch portal) and the Bheri thrust (5,800m from the portal). The GBR predicted that high water ingress would be a problem at the Bheri thrust zone in particular, but the rock was found to be Class IV and perfectly dry. The TBM passed through without problems.

There were two incidents of high water ingress (2,000 litres per minute) on 27 December 2017 at ch. 1+174.782 from the 8 o’clock to the 11 o’clock positions of the ring, and on 6 January 2018 at ch. 1+337.457 in the 12 o’clock position, but the TBM was able to move through at a reduced speed and the flow was stopped with grout injections.


The TBM selected for the first assault on the Siwalik by mechanised methods was a Robbins double gripper shield, it is equipped with 33 no. 17in disc cutters, has main drive power of 6×330 (1,980kW) for a total power of 3,027kW (see box on page 43 for full specification).

It was launched from a 150m drill-and-dig step-in tunnel, lined with lattice girders and shotcrete. There was then a 500m-long ‘trial bore’ where drilling parameters were assessed, given the total lack of knowledge of TBM performance in the Siwalik. A three-month advance of 750m was considered a learning curve target.

The project is seen locally as a test case for how effective mechanised tunnelling can be in the region.

Robbins project engineer Missy Isaman led the team that designed and built the TBM. She said of the project: “The process of designing the Bheri Babai TBM was very interesting. There were two big concerns with this tunnel. One being geological and the other being political.

As far as the tunnel and the machine design, it was critical to focus on the geology.

“The Himalayas have notoriously difficult ground conditions. Any time you bore through a mountain it is difficult to get an accurate idea of the geology. In the drive for this project, we knew there was at least one large fault zone that the machine had to pass through. Ground faults pose unknown ground and water pressure that can severely damage or destroy a machine, if not prepared.

“The water alone can flood a machine with large amounts of silt. These small particles can fill in the invert of the machine making it very difficult to clean out around the cylinders and other equipment. Squeezing ground can also pose a problem. Getting a machine moving once it is stuck, can be an expensive and time-consuming process.

We spent a lot of time planning for the best ways to combat situations like this. Multiple drilling locations were included, in order to ensure that the machine had plenty of options for probing and grouting. Other preventative measures were also taken for the preparation additional equipment that could be added in the future if needed.

“The other concern for this project was the relevance of this machine to the country of Nepal. Years of government debate and planning went into the decision to use a TBM for the first time in Nepal. We always plan for every machine to succeed, but the success of this project held a lot of importance. With the kick off of this project, came the hope and plans for many more TBM projects in the near future for Nepal. If the Bheri Babai machine encountered too many difficulties or delays, the future plans for the TBM industry in this part of the world could be negatively impacted for a long time.

“Boring has gone better than expected. Except for a few encounters of water, which were dealt with fairly quickly, the machine has been making great time. [Everyone is happy, and we are] already in the planning stages of more TBM projects. As an engineer, it is pretty special to be a part of something like this. Having this be a successful first in the country, is something I am really proud of.”


In terms of machine performance, an average advance rate of 400-500m per month was expected by the client. The site teams had achieved around 700m on average at the time of Tunnels and Tunnelling’s visit, with a best daily advance of 62m, a best weekly advance of 325m and a best monthly advance of 1,063m.

Challenging Moments During Excavation

Front shield sticks

At ch. 8+588.860 the TBM’s front shield became stuck. The axis of the tunnel was positioned just about perpendicularly to the strike of the rock mass, until around ch. 8+400 (or 10 October 2018). At this point, the dip direction changed and the tunnel axis was nearly parallel to the strike, and groundwater was seen to be dripping from the face.

The tunnel alignment began to deviate consistently from 14 October and had reached 131mm when the TBM became stuck on 16 October. Geodata believes that one or a combination of the following factors caused this: unfavourable ground conditions, misalignment of the TBM due to deviation, a snag in the guidance system, a lack of ground engineering works.

The contractor applied a thrust of 18,500kN to free the TBM but this was unsuccessful, so a bypass was excavated from the telescopic shield window to the cutterhead to release the TBM. This excavation was 6.5m long, 0.7m wide and from the 5-12 o’clock position. The TBM was successfully released with 10,000kN of thrust. Boring resumed on 23 October.

Cutterhead jamming

The cutterhead jammed at ch. 8+606.262 when loosely cemented sandstone and water ingress from the 11 o’clock position triggered overbreak at the left side of the crown and jammed the cutterhead.

To control the ingress, 1,287kg of polyurethane was injected through a 16m-deep probe hole drilled from the 12 o’clock position, this almost entirely stopped water ingress and torque of 440kNm was applied, which freed the jammed cutterhead. The whole operation to unjam the cutter took 15 hours, from about 4am to 7pm on 24 October 2018.


The tunnel is lined with a precast segmental ring comprising four hexagonal segments, with which it is possible to perform a rapid ring build, so that lining operations do not slow down a speedy advance. Sources tell Tunnels and Tunnelling that this method has in the past been associated with higher water ingress as a TBM alignment correction can result in larger than normal gaps between the circumferential joints, so it is generally used in situations where this is not a concern for the long-term design or the construction of the tunnel.

The segments were placed sides first, then bottom, then top, for a 4.8m outside ring diameter, 4.2m inside, have a 300mm thickness and are 1.4m long. The ring is backfilled with pea gravel and grout along the bottom and sides of the annular void to stabilise the ring, with full mortaring to fill the void at the top taking place at the end of excavation.

Two types of segment have been designed for the job: type A and type B. Type A. Both types are reinforced with rebar cages; type A containing 95kg/m3 of steel, while type B has 153kg/m3. This translates to a rebar thickness of 12mm for type A and 16mm for B.

Type B is for complex areas, curves and anywhere with rock mass classification five, while type A is for normal tunnelling. By the 9km point in the drive, 92.52 per cent of the segments used were type A, compared to the 66.58 per cent expected during the planning phase, owing to better-than-expected conditions.

The segments are also bolted in the critical areas, but elsewhere guiding bars have been sufficient. Glued-in, narrow EPDM gaskets from a Chinese supplier were also used on all segments.


Covec purchased seven sets of moulds (for a total of 28 moulds) from a Chinese manufacturer. An additional set was purchased and is being held in reserve for emergencies.

At the time of Tunnels and Tunnelling’s visit in November 2018 the casting yard had achieved an average seven-day strength of 53.57MPa and a 28-day strength of 67.38MPa. Segments are steam- and water-cured.

The outside air temperature is temperate in the winter, owing to the relatively low altitude (in Nepali terms) and was not at all close to approaching freezing during Tunnels and Tunnelling’s visit. It can reach 48°C and 60 per cent humidity in the summer. Vibrators were used to agitate the concrete as it was placed in the mould, vacuum lifts were used for de-moulding and nothing unusual in terms of admixtures has been required.

Production has managed to reach 100 segments per day. In line with Nepalese government requirements it employs a large local workforce of around 360 people, with six Chinese overseers per shift.

According to project manager Hu Tianran, the local workforce needed training from scratch, and had little knowledge of construction work, but has performed well. The cracking percentage has been low, less than 1 per cent. And with the storage on-site for 1,500 segments, supplying the TBM has not been a problem.

The location has allowed for a generous site footprint, which has allowed for clearly defined areas for different tasks, such as aggregate storage, reinforcement cage preparation, curing and final segment storage, and so it is possible to keep less trained workers on appropriate jobs.

The Bheri Babai project has been a success for the introduction of mechanised tunnelling to Nepal, and hopefully the first in a series of projects for the Nepalese tunnelling community, which looks likely to enjoy more enthusiastic backing from key national clients. The Nepal Ministry of Energy, Water Resources and Irrigation is now reconsidering a number of its projects that were previously considered non-viable, due to the slow progress associated with nonmechanised tunnelling in the Siwalik.