Excavations for the Lam Ta Khong power station 01 Dec 1998

Shani Wallis, TunnelTalk

Lam Ta Khong is one of three pumped storage schemes identified by Thailand's Electricity Generating Authority (EGAT) to meet peak power requirements. It was first studied in 1985 and its feasibility was confirmed in 1991 by the Japanese International Corporation Agency in conjunction with Japan's Electric Power Development Company (EPDC).

Heavy immediate and permanent support of the powerhouse cavern vault
Heavy immediate and permanent support of the powerhouse cavern vault

Project realisation was approved in the early 1990s and loans for US$100 million and ¥18 billion from the International Bank for Reconstruction and Development and the Overseas Economic Cooperation Fund of Japan were secured to complete the civil works and install the first two of the proposed four 250MW turbines.

EPDC completed the scheme's detailed design in association with EGAT's design division in the early 1990s and the civil construction contract was awarded in December 1995 to a consortium led by Vianini Lavori of Italy (the sponsor), with Dragados of Spain and local Thai company Nawarat Patanakarn. The winning bid of Baht2.47 billion (about US$100 million at time of award) was the lowest of nine international bidders elected from nineteen pre-qualified consortia invited to tender. The second bid was within 8% of the winning bid and the highest bid was nearly Baht4 billion. The Engineer's Estimate was double the lowest bid price, less 20-30% which is usual according to Kiyoshi Shiota, the Project Manager and Resident Engineer for EPDC.

Vianini/Dragados/Nawarat started work on site in December 1995 and the nearly four-year civil contract is progressing towards an on-schedule completion by the end of 1999. Excavation of the underground structures started in March 1996 and by December 1997 more than 500,000m3 of rock had been removed and only a few elements of underground excavation were left to complete.

Design features

The scheme lies some 190km north-east of Bangkok and takes advantage of a 400m elevation difference between the existing Lam Ta Khong Reservoir - built as a water supply/irrigation project in 1968 - and a point of the Khorat Plateau which rises steeply to the east of the reservoir. The existing reservoir is the project's tailrace outlet and a new upper reservoir is being built under the current contract. The particularly short 2.2km waterway between the two provides an effective head of 360m as the water passes from the 600m x 600m x 40m deep man-made upper basin through two 550m long steeply inclined underground penstocks (one for each of two units), through the underground powerhouse which, at 175m long, 25m wide and 49m high, is one of the largest underground caverns in South East Asia, and out through two 1.4km long tailrace tunnels (Fig 1). There is an underground surge tank chamber associated with each tailrace and nearly 6km of additional tunnelling is required for access, drainage and cable laying.

Fig 1. Isometric of the pumped storage scheme
Fig 1. Isometric of the pumped storage scheme

Design and excavation of the huge powerhouse cavern and the two inclined penstocks, were the two most challenging aspects of the project. Both lie in the gently dipping and moderately to slightly jointed sandstone and siltstone of the Khorat Plateau, which can be on the weak side. There are no major faults in the sedimentary rocks or abrupt geological changes, but the siltstone and sandstone formations can be thinly bedded and more highly fractured, and the siltstone can exhibit a degree of instability in the form of spalling and slaking.

Because of these geotechnical weaknesses and the need to respect safety, EPDC in conjunction with the project's IBE (international board of experts) specified that the 7.2m to 6.1m o.d. (6m to 4.9m steel lined i.d.) penstocks be excavated in the full-face, topdown method. This would avoid any uncontrolled ravelling of instability through the minimum 60m rock cover of the inclined penstocks. The method requires the muck to be hauled back up through the shaft, rather than allowing it to free-fall through the unlined excavation of conventional shaft raising techniques. It also allowed for each round of exposed rock to be supported with a systematic pattern of rockbolts and shotcrete to protect against weathering, damage or excessive stress relaxation.

Instability of the rock mass in the powerhouse region was minimised by the decision to incorporate the transformers within the one huge machine hall, thereby eliminating the juxtaposition of a separate large excavation for the transformer hall. The design of this huge machine hall, its shape and its position, was said by Eng Shiota of EPDC to be the most difficult aspect of the project. After much study and analysis, the cavern was located such that a 15m deep band of hard sandstone (of between 300-600kgf/cm2 UCS) passes through the middle to upper walls of the 49m high x 25m wide cavern. This, however, placed the vault in a band of weaker siltstone of between 100 and 600kgf/cm2 UCS in which weathering and deformation could be expected.

For the cavern vault, EPDC dismissed the conventional in-situ concrete lining and selected instead a more flexible but comprehensive system of steel fibre reinforced shotcrete, rockbolts and prestressed rock anchors of 10m and 15m long. A drainage gallery around three sides of the cavern further improved stability by reducing hydrostatic ground water pressures on the cavern.

Fig 2. Specially designed jumbos for excavating the 33m2 full-face x 48˚ incline penstocks
Fig 2. Specially designed jumbos for excavating the 33m2 full-face x 48˚ incline penstocks

The intense support system in the powerhouse vault starts with an 8cm layer of 100kg/m3 steel-fibre wet-mix shotcrete. Fully resin grouted rockbolts 5m long and on a 1.5m x 1.5m pattern were then installed followed by another 8cm layer of steel-fibre shotcrete. The six-strand 15m long rock anchors were then installed also on a 1.5m x 15m pattern between the rockbolts and these were tensioned to 65 tonne. A final 8cm layer of steel-fibre shotcrete finished the programme. The same combined support system was applied to the walls of the powerhouse but with the rockbolts on a 2m x 2m pattern and the rock anchors on a 4m x 4m pattern.

During the TunnelTalk site visit at the end of 1997, excavation of the cavern was in the final stages and the final concreting works had started. In-situ concrete walls of 1,000mm thick lined the lower 21m of the cavern walls and concrete pillars above and extending across the band of strong sandstone, support the beams for the machine hall mobile cranes.

Once the strongly supported vault of the cavern was completed, excavation of the subsequent benches in the 175m long x 25m wide machine hall progressed at an average of 15,000m3/month completing the total cavern excavation of approximately 120,000m3 in eight months. Deformation of the vault and walls as excavation progressed was said to be as expected with the prestressed rock anchors arresting any relaxation or movement.

All excavation, except for a roadheader used by a subcontractor for the upper drainage tunnel designed to reduce ground water pressures on the asphalt lining of the upper reservoir, was by drill+blast using Atlas Copco drilling rigs. These also drilled all the holes for rockbolts and rock anchors.

Within the specification, rockbolts are not tensioned but three bolts in every 100m of tunnel were subjected to a 10 tonne pull test. Less force was applied to pull tests in sections of weaker claystone of the project but any pull test failures resulted in the deduction from contract payments of all bolts in the 100m length. There was no tolerance for failure in the contract documents.

Casting of in-situ lining
Casting of in-situ lining

Shotcrete provided a major element of support on the project and more than 13,000m3 of wet and dry mix shotcrete has been applied using Aliva equipment. In the powerhouse cavern, two Aliva 285 mobile and remotely controlled units with telescopic nozzle booms were used to apply the wet-mix, steel-fibre reinforced shotcrete. Initially, the design specified 117kg of steel fibre/m3 in order to attain a high shotcrete strength. For workability this was reduced to 100kg/m3 which is still substantially higher than the more familiar 35-50kg/m3 but as the permanent and final lining for the wide cavern vault, EPDC required high strength and long-term support quality.

The consortium used Xorex steel fibre imported from Novocon of the USA through the Sika agent in Thailand. Vianini/Dragados/Nawarat also used Sika admixtures in the shotcrete using Sigunit accelerators, Sika LPP1 silica fume and Sika plasticisers.

Penstock excavation

Full-face top-down excavation of the two penstocks called on particularly innovative ideas. Two access adits allowed the possibility of excavating the shafts as four concurrent headings with two 360m drives from the upper access and two 180m sections from the lower. The joint venture, however, proposed working from two headings, which caused the engineer some concern for the programme. But Vianini/Dragados/Nawarat was confident in its chosen technique.

To tackle the challenge, Techniplant of Milan in collaboration with Atlas Copco was engaged to design a machine or jumbo from which excavation, mucking out, and immediate rockbolting and shotcreting support could be carried out as a complete cycle. To meet the programme, two sets of equipment were required to advance both penstocks concurrently. The first of the two machines had to be manufactured, delivered and mobilised within seven months of contract award, and both had to provide for an average rate of progress of at least 1.5m/day.

To meet the delivery deadline, the machines were designed using standard components with specialist equipment imported from overseas, and the working platforms and machine frames on their 48˚ angle to the horizontal were fabricated in Thailand.

Each machine was equipped with two Atlas Copco drill rigs comprising BUT 28 booms, BMH 2343 feeds and COP 1238 ME drifters. A 0.4m3 capacity backhoe mounted on the frame loaded muck into a single 4m3 muck skip which was designed to pass through the frame of the working decks. The drill rigs drilled the holes for the 2m long resin grouted rockbolts on their specified patterns and an Aliva 262 machine stationed at the top of the shaft conveyed dry mix shotcrete down a 75mm diameter steel pipeline to the hand held nozzle on the working platform.

Shotcreting in the portal zone
Shotcreting in the portal zone

The jumbos, or working platforms, and the muck skips were rail-mounted and winch- operated, each running on their own set of tracks. After charging each face, the jumbos were winched 50m back from the face for the blast and lowered again for the mucking and support cycles. All winching was controlled from a station at the top of the shaft. A small 75kW winch raised the 40 tonne working platform at a maximum speed of 9m/min and a larger 200kW winch raised the 14 tonne fully loaded skip at a maximum 90m/min. At the top of the shaft, a device tipped the skip up and over to discharge into a muck bunker.

Several safety features were built into the working decks, including mechanical brakes and stabilisers on the jumbos, which gripped the H250 rails front and back, and the muck skips were equipped with an automatic mechanical brake. Limit switches controlled line speeds on the winches and there was telephone and video contact between the winch and shotcrete operators at the top of the shaft and the workers on the working deck below. There was also camera surveillance of the travelling muck skip.

After placing the order for the two penstock excavation jumbos in March 1996, the first arrived in August 1996 and the second a month later. Following a relatively short learning period, production stabilised at about one 3m blast/24h day.

Both machines started from the top of the penstocks and were designed to carry on past the intermediate access. Only minor, but significant, changes were made at this intermediate level. The winches and the shotcrete machine and their control stations remained at the top of the penstocks, but the skip tipping device was relocated to the intermediate access to take advantage of shorter cycle times.

At the time of the site visit, excavation of both shafts was complete. The minimum 1.5m/day programme average had been bettered to a 2m/day average including all stops. With best advance rates of more than 20m/week on occasions, excavation of the penstocks, some 10 weeks ahead of schedule, was considered a major success.

Meanwhile concrete lining in the powerhouse and the two tailrace tunnels has continued through 1998. As part of the strict adherence to specifications, the contractor must trim back any underbreak of the shotcrete-supported excavations before the steel reinforcement of the lining can be installed.

As the civil contract draws to a close, the electrical and mechanical contractors are beginning work. The first two units of the 1,000MW installation are scheduled to be in operation by April and July 2000. The second two are programmed to be installed in the project's second phase planned for a date beyond 2006.


Thanks are extended to Eng Shiota of EPDC; Giuseppe Chiodetti, Project Manager for Vianini/Dragados/Nawarat JV and Eng Enrique Fernandez, Manager of the Under- ground Excavation Works for receiving TunnelTalk on site, conducted the underground tour and providing information regarding construction methods, design and progress.


  • Visit the TunnelTalk Archive and use the library search facility to access a wealth of information about tunnelling and underground engineering projects in Thailand.

Add your comment

Thank you for taking the time to share your thoughts and comments. You share in the wider tunnelling community, so please keep your comments smart and civil. Don't attack other readers personally, and keep your language professional.
In case of an error submitting Feedback, copy and send the text to Feedback@TunnelTalk.com
Name :

Date :

Email :

Phone No :

   Security Image Refresh
Enter the security code :
No spaces, case-sensitive