Features
Raising of the Kotmale dam – misconceptions and facts
By Asoka Herath
BSc (Geology Hons); MSc (Engineering Geology), CEng.
Over the past few months, there has been considerable interests among not only geoscientists and engineers but also ordinary people in the potential risks faced by two of the major hydropower projects constructed in the central highlands, namely the Victoria and Kotmale projects. These projects have come into the limelight owing to some minor seismic tremors felt in the Victoria area, together with the recent announcement of the intention to raise the Kotmale dam. Alarm bells are ringing that seismic activity may jeopardize the integrity of the Victoria dam and the raising of Kotmale may result in dam failure. Should these dams fail, they will be the biggest manmade disasters our country has ever faced. This risk has prompted suggestions from the people higher up that we should seek advice from foreign experts.
Various opinions from the leading geoscientists of the country, regarding the causes and effects of this unusual activity were aired through the media recently. Some have attributed the seismic tremors to limestone quarrying in the Victoria reservoir area, some as the result of neo-tectonic movements of the central highlands, yet some others said they are reservoir-induced earthquakes. Those who are interested in the fundamentals of such activity may find the presentation in this link useful (https://www.youtube.com/watch?v=7huzvuU9TEI).
The intention of this article is to provide an insight into the geotechnical issues that influence the Kotmale Project, the risks, and to address some misconceptions people may have regarding the project. I take the liberty of writing on these issues as a person who has an in-depth knowledge of the Project by working from the site investigation phase through to construction completion, the later years as the Senior Engineering Geologist, leading the project geoscientist team.
The Project
Kotmale was meant to be the main upstream storage reservoir under the Mahaweli development scheme launched in the 1960s. It will regulate the Mahaweli waters, while providing a substantial energy component. The capacity of the reservoir will be more than doubled by the proposed raising of the dam, approximately by 30 m. The project was inaugurated in 1979 and funded by Sweden. Sir William Halcrow and Partners of UK (Halcrow) with Central Engineering Consultancy Bureau of Sri Lanka (CECB) was the design engineer and Skanska of Sweden was the contractor.
To understand the project risks, it is essential to understand the geology and structure of the project area. The area is underlain by Pre-Cambrian age metamorphic rocks comprising predominantly of two rock types, charnokite (gneiss) and crystalline limestone (dolomite or marble) with minor quartzites. The limestone is sandwiched between layers of charnokite gneiss.
The rock units are folded into a gentle anticlinorium (an arch like structure-convex shape) which has a slight plunge downstream. The axis of the anticline trends northwest and falls on the right abutment in the dam area. As such the dam foundation rocks have a slight inclination (15°-25°) downstream and into the abutments.
Charnokite is a very strong resistant rock, which does not breakdown easily while limestone deteriorates and erodes more easily. The initial site investigation drilling in the Kotmale valley found limestone thickness varied from 20 m to 130 m within the project area.
The limestone consists mainly of calcium carbonate (CaCO3) and partly magnesian carbonate (MgCO3). Carbonates dissolves in acidic water (acid rain) and forms solution cavities. These cavities have the potential to create leaks from the reservoirs.
Formation of the Kotmale Valley
The Kotmale valley is an amphitheatre-like valley reshaped by the erosional activity of the Kotmale Oya for millennia. It is approximately 5 km wide at the widest point, is surrounded by high rock cliffs up to 300 m high on both flanks, which converge at the Kadadora village; the site of the dam construction. The valley forming process was a combination of differential weathering and erosion of the underlying limestone unit, resulting in the collapse of overlying more resistant rock units. The basal slopes of the cliffs consist of accumulated debris of rocks and soil derived from the cliffs (termed as scree or talus). Historically these talus slopes were in perpetual slow motion (creep) towards the river. They were subjected to numerous landslides, typically during or after heavy rain events. Generally, the Kotmale valley was infamous for landslide activity.
History of Kotmale Project investigations
The location of the dam site had been selected for the seemingly favourable topography formed by the converging valley flanks. One of the first jobs assigned to me when I joined the CECB in 1978 as a young engineering geologist, was to map four exploratory tunnels driven into the abutments of the proposed dam and report the rock conditions. Our evaluation indicated the left abutment had very poor rock conditions not suitable for a dam foundation.
In 1979, to evaluate the landslide risk to the project, Halcrow carried out a landslide survey of the project area. This survey was done due to the frequent landslide activity in the Kothmale area, and its similarities to the Vajont Dam disaster that occurred in Italy in 1963, which wiped out several downstream towns resulting about 2500 deaths.
This study identified the left abutment as a settled block (subsided en mass); limestone solution the likely cause. It strongly recommended to move the damsite downstream and monitor the impressive cliffs on the left flank of the valley for potential movements.
Evaluation by a panel of experts
As these findings were major issues impacting the viability of the Project, the Mahaweli Authority of Sri Lanka as the owner of the project, in early 1980 engaged a panel of foreign experts who were the leaders in the fields of rock and soil mechanics, to evaluate the potential issues and provide recommendations. This panel endorsed the relocation of the damsite.
Subsequently, the dam axis was moved approximately 200 m downstream from the original location.
Nature and distribution of the limestone
At the new location, the limestone was at a much greater depth below the dam foundation compared to the original site as shown in Figure 1. Even so, solution cavities were observed in the contact zone and within the limestone itself identified by drilling and close circuit television (CCTV) inspections carried out in the drillholes. Our best estimate indicated small cavities, however a drillhole represents only a small area, therefore the existence of larger cavities could not be ruled out.
Leakage through cavernous limestone is widespread from the reservoirs built in our hill country. The Samanalawewa Project is the best example where significant leakage had been occurring through a cavernous limestone under the right abutment since the impoundment, which could not be sealed.
Continuity of the Kotmale limestone
Investigations were carried out to establish the continuity of the Kotmale limestone to downstream areas in view of potential leakage which were inconclusive. The recent geological map for the Kandy-Nuwara-Eliya districts (GSMB), I am convinced that the Kotmale limestone continues to Gampola-Atabage Valley and further to the Victoria Reservoir as a major rock unit. The limestone quarrying in the Victoria Reservoir area occurs in the same unit, which caused the recent furore about seismic events allegedly caused by blasting. Whilst I do not believe there is (or will be) significant leakage from Kotmale Reservoir, future raising of the dam may enhance leakage if it exists which needs evaluation.
Water tightness of the dam foundation
To eliminate the risk of water leakage through the limestone under the dam foundation, it was extensively grouted from a grouting gallery constructed below the dam level (see Figure 1). Generally, the grout intakes were low. Nevertheless, few holes required significant amounts of grout before they could be sealed. Subsequent water pressure testing indicated the limestone unit was properly sealed and the foundation was watertight.
Monitoring of the cliffs and other landslides
During the initial impoundment, the cliffs were inspected by walkover surveys and by periodic monitoring with geodetic surveying. Movements of any significance, precursor of an impending major failure were not recorded during the early years after construction.
Reservoir Induced Seismicity (RIS)
One important consideration in the construction of large reservoirs is the potential for reservoir induced seismicity (RIS). RIS is the incidence of earthquakes triggered due to the impoundment of water behind a dam. A simplistic explanation is that reservoirs trigger earth tremors due to the load of water which could activate otherwise dormant faults, the energy released causing earth tremors.
Halcrow decided to investigate the risk of movement of some major structures present in the Kotmale reservoir area upon impoundment, which could potentially induce seismic events. Consequently, a micro-seismic monitoring network (MSMN) consisting of four monitoring stations was established around Kotmale project in 1981. The monitoring program was managed by the CECB and continued until the early 90s.
Published data (Fernando & Kulasinghe 1985) show that the maximum event recorded at Kotmale during the first two and half years of monitoring, was an event of 2.25 magnitude (Richter Scale), located far away from the project area. The monitoring led to the conclusion that the potential for RIS at Kotmale was extremely low or non-existent. However, in view of the recent seismic events recorded around Victoria, it will be prudent to revisit the monitoring records from the Kotmale network, presumably archived by the CECB.
Potential risks from future raising
The Kotmale project infrastructure, were designed and constructed for future raising. Theraised dam simply achieves the original vision of the project, as the main upstream storage reservoir and increases power output.
In my opinion the geotechnical risks arising from the raising of the Kotmale dam are as follows:
= The raised water level in the reservoir will fully inundate the scree slopes and reach the base of the left bank cliffs with potential to create major landslides. Cliff monitoring should be re-established using modern methods such as satellite based synthetic aperture radar interferometry (inSAR).
= There is potential for leakage through the limestone with raised water levels. All evidence suggests the limestone under the dam was properly sealed. Leakage from the reservoir flanks cannot be prevented, which is not a fatal flaw to the project. Requirement for abutment sealing should be revisited.
= It will be useful to re-establish the monitoring program of stream gauging in the Gampola-Atabage valley to investigate if any leakage will occur from Kotmale to Atabage valley. This should be initiated before the dam raising.
= All evidence points to reservoir induced seismicity is a non-issue for Kotmale Project.
Finally, I strongly advocate the raising of the dam for the following reasons:
= The original project infrastructure was developed with the intention of future raising.
= There are potential risks but not fatal flaws to the viability of the Project. These risks can be mitigated.
= People were evacuated from their ancestral lands going under the reservoir and if the project is not completed as originally intended, then this relocation of people was without merit and a grave injustice to them.
= The country unnecessarily lost approximately 30 MW of power for more than 35 years.
The potential risks with the raising of the Kotmale dam discussed here need to be evaluated, and if the need arise, they should be mitigated and managed using local expertise. In my opinion, what Sri Lanka lacks is good project managers vital in successful implementation of projects of such enormous national importance. In hindsight, the Kotmale reservoir should have been built to the full capacity at the first instance.
(The author currently works as a Principal Geotechnical Engineer with AMC Consultants Pty Ltd, Perth, Australia, and can be contacted at hma1904@gmail.com).