Civil works at Dhauliganga will start in 1999. I D Clarke* and S P Sen**describe the scheme.

THE Dhauliganga hydroelectric project is a medium head, run-of-river, diversion scheme to be built on the river of the same name in the far north of Uttar Pradesh. It is the lowest stage of a proposed cascade developing the power resources of this river of the Indian Himalayas. The Dhauliganga is a tributary of the river Sarda (or Kali) which, over almost 200km, forms Nepal’s border with India.

The lowest stage of the Dhauliganga cascade will develop a head of slightly more than 300m and will have an installed capacity of 4x70MW. Water taken from Chirkila daily storage reservoir, about 4km upstream of the Sarda confluence, will pass along a 5.6km pressure tunnel to an underground power station on the western side of the Sarda valley. The project will generate 1285GWh annually, (90 per cent dependability). The construction programme sees the start of main civil works in 1999 and commissioning of the four generating units during late 2003 and early 2004.

Five-stage development of the Dhauliganga was many years ago the subject of master plan studies. In 1980, the National Hydroelectric Power Corporation (NHPC) carried out a feasibility study of the lowest project of the cascade, which was completed in 1985. Since then, NHPC has continued studies and has obtained clearance for project realisation from the Central Electricity Authority and the Public Investment Board. Work has recently started on the tender design.

Development of the four upper stage schemes has been halted, pending further environmental studies.

Conditions of the site

The catchment of the proposed dam site at Chirkila is 1360km2, and is an extremely rugged area with elevation ranging from 1300m asl to more than 6000m. The average river slope is almost 50m/km but locally much steeper. Valley slopes are often unstable and landslides are common. This, combined with freeze/thaw in the spring and torrential rainfall during the monsoon, contribute to high erosion rates and sediment load.

The principal hydrological studies covered dependable 10-day mean flows, to compute energy production, construction diversion and spillway design floods, and sediment transport, to assess reservoir storage, flushing and desilting requirements. The mean flow of the Dhauliganga at Chirkila is 84m3/s, ranging from about 20m3/s in February to 200m3/s in the July-September monsoon. Suspended sediment is estimated at 1.40 million tons per year, or a mean concentration of 340g/m3, but may be disproportionally higher in wet years.

Review studies for the final design also have to include the potential for glacial lake outburst floods and whether these could exceed natural flood inflows and threaten the dam. Little information is currently available for this analysis but surveys are in progress based on satellite images of glaciers and moraine-dammed lakes in the upper catchment.

As regards geology, the dam site is a deep, asymmetrical gorge in sound but highly foliated pre-cambrian biotite gneiss bedrock, with extensive areas of slump debris upstream and downstream of the dam axis. Alluvial deposits (60m maximum thickness) fill the valley; these are fine-medium sand (10-4-10-5m/s), generally with boulders and gravel, but containing one lens solely of sand.

The 5.6km pressure headrace tunnel will pierce generally massive crystallines. The tunnel at its deepest is about 1000m below ground, but its final third is just 100-200m deep. The surge chamber, twin high pressure shafts and power caverns will be in strong, generally massive but locally-fractured biotite gneiss.

Completed site investigations include core drilling and grouting tests at the site, excavation of drifts into the dam abutments, surge tank area and power cavern area (with rock mechanics testing) and laboratory testing of materials.

The main factors affecting design of the project are:

• The remote project area and the consequent very difficult logistics.

• The difficult topography and geology of the Dhauliganga valley at the site of the dam, spillway and intake structure.

• The sediment load carried by the river, principally during the monsoon season.

The dam design

The preferred dam site is characterised by the asymmetric valley with rock exposed on both banks above about El 1320m. The right flank is a steep high outcrop of bedrock, but on the left bank the alluvium forms a gently sloping terrace 150m wide. The overall width of the valley at dam crest level is about 270m.

The larger part of the dam foundation area will be the river channel alluvium, the upper and lower heterogeneous strata of which are separated by a distinct lens of medium to fine sand, of 3-20m thickness on the dam axis. This lens extends upstream and downstream and investigations of the longitudinal variation of its thickness are in progress.

Even though the dam’s maximum height will be only 50m above ground level, and it will only impound a small reservoir for daily storage (6 million m3), its site offers design challenges:

• Suitable clay core material is only found about 40km from the site, along the narrow, mountainous access road used by all local and other construction traffic.

• The sealing of the deep alluvial deposits and their deformation when surcharged, and the risk of liquefaction of the confined sand lens (the site is in one of the seismically most-active regions of India).

• The need to site the large spillway adjacent to the intake structure, against the steep right abutment of the dam, to facilitate sediment flushing.

• The restricted working area in the narrow valley and the extensive excavation needed in the valley flanks.

An earth-core rockfill dam (ECRD) founded on the alluvial deposits, with a cut-off or special grout curtain, would be the solution of choice for the site, were it not for the problem of the transport of core material. This haulage would cause great disruption of traffic on the sinuous access road up the Sarda valley, which passes through many villages, with the risk of accidents, construction delays and claims by other contractors and road users. During two dry seasons (300 working days), clay haulage lorries would have to pass several thousand times daily on a carriageway less than 6m wide.

Design studies now concentrate on the alternative concrete-face rockfill dam (CFRD). Although there are few precedents to date for a CFRD with the plinth wall not founded on bedrock, an adequate design can be developed. Plentiful rockfill for such a dam will be available from the spillway area, but very careful co-ordination of excavation work will be necessary.

Apart from avoiding haulage of core material, a CFRD offers an easier construction programme (the cut-off does not have to be completed before placing can start of the dam itself, and more work will be possible in the rainy season) and simplified river diversion arrangements.

The second important design feature is the cut-off to seal the deep alluvial foundations of the dam. In such material, grouting requires special procedures and its cost and effectiveness are unpredictable. The alternatives of piled concrete or panel cut-off walls have been constructed for many dams on similar foundations, even where this has required the trench to be excavated through alluvium containing large boulders.

A grout curtain could be constructed in the alluvial deposits, using the latest methods, but is not considered a desirable solution for a CFRD as several rows of curtain would be needed and would call for a very wide and expensive plinth slab. For this reason, detailed design is concentrating on a positive cut-off in the form of a concrete wall; for which the main difficulties to be overcome will be the excavation or removal of boulders from the slurry trench and the accurate placing of adjacent panels of the wall.

Whilst the concrete face itself poses no problem, the highly-stressed connection of the plinth to the cut-off, along the heel of a CFRD, will be a critical design feature. The design has to ensure satisfactory behaviour of this connection under the foundation settlement likely following first filling of the reservoir.

Among the other design considerations:

•The spillway, on the right abutment of the dam, will have three low-level radial gates, 6m wide and 10m high, and will be located adjacent to and at right-angles to the power intake structure. No overflow weir is proposed so during the monsoon, when both spill and power production are at their maximum, floods discharging through the low-level gates will prevent sediment accumulating in front of the power inlets. For detailed design of the spillway and intake, model tests are being done at the Central Water and Power Research Station in Puna. Because of the height and angle of the valley flank, spillway construction will require more than one million m3 of rock excavation, much of which will be placed in the dam.

•A conventional river diverson will be via a 10.5m diameter tunnel, 700m long, through the right abutment of the dam. Detailed studies are in progress to determine whether it would be feasible, during the one critical monsoon season of the construction period, for the river to be diverted through the partially-completed spillway channel. If so, the diversion tunnel could be dimensioned only for a much smaller dry season flood.

•Underground desilting in the form of two 300m-long parallel chambers will be constructed in the right abutment of the dam, immediately downstream of the intake structure, and will be dimensioned to remove material down to 0.2mm diameter. Sediments will be flushed back to the river along a low-level, free-flow tunnel. These de-silting chambers are currently also being modelled.

•The concrete-lined pressure tunnel, 5.6km long and of 6.5m lined diameter, will carry diverted water to the two, vertical pressure shafts, each about 250m deep. These are located just downstream of the surge chamber, which is a 15m diameter vertical shaft, about 90m deep. The lower parts of the pressure shafts will be steel-lined and branch penstocks will connect them to the spherical valves of the Francis turbines.

•The main power cavern, about 100m long, will contain the four turbine generator groups, the spherical valves and control room. A smaller, parallel cavern will house the unit and reserve transformers as well as, in a separate extension, the gas-insulated switchgear. Generated power at 220kV will be carried to the outgoing overhead transmission lines by cross-linked, polyethylene insulated cables, located in the ventilation tunnel. An inclined and curved construction gallery will be converted to form a surge chamber at the upstream end of the 480m long tailrace tunnel.

•The four Francis turbines will have a rated capacity of 70MW under a net head of 297m and will run at 428.6rpm. The generators, each of 80MVA and 11kV, will be air-cooled. Transport capacity limitations meant that only single-phase transformers can be considered.

Although its detailed design and construction does not form part of the Project for which NHPC is responsible, the transmission line serving the station will be a 220 kV double-circuit line about 330km long. It will connect with the interconnected system near Bareilly and over much of its length will cross difficult mountain terrain.

Construction planning

A period of about five years is required for construction of the Dhauliganga project, including the river diversion tunnel which may be constructed under a separate advance contract. Key construction sequences are:

• Excavation of the underground power station, concrete placing and erection and testing of the units.

• Construction of the spillway and adjacent dam (the aim here will be to minimise stockpiling of excavated rockfill prior to its placing in the dam).

The environmental impact of the project will be limited. Only a small area will be flooded by the reservoir, most structures will be underground and there will be no significant change of the regimes of the Dhauliganga or Sardar rivers. All related procedures under Indian law are in progress and resettlement plans for the small affected population are being drawn up. A catchment protection and management programme is also being prepared. Construction of the project will, however, have a major impact on the region, in particular the traffic and influx of workers, and appropriate mitigating measures are being drawn up during the present design work.

Related Articles
Spotlight on… CFRD