History & analysis · Long read · 17 min read

Mullaperiyar: A 130-Year-Old Dam, Two States and the Question of Risk

A colonial lease, a lime-and-rubble dam in an earthquake zone, millions living downstream and millions more depending on its water. The history, the science and the hard choices behind one of India's longest-running disputes.

By GK Global, Adimali, Munnar · Trails & Tales by GK

Deep in the forests of Periyar Tiger Reserve, a few kilometres from Thekkady, a wall of stone and lime holds back the waters of the Periyar. Tourists on the boat at Thekkady are floating on its reservoir. The dam was opened in 1895, when Queen Victoria was on the throne and the Kannan Devan Hills above Munnar were still being cleared for tea. It stands in Kerala, on Kerala's river, with a catchment entirely in Kerala; but it is operated by Tamil Nadu, and its water flows east through a tunnel to irrigate the dry plains of the Vaigai basin.

For more than four decades the Mullaperiyar dam has been the subject of court cases, expert committees, protests and political promises in both states. For people in Kerala's central districts it is a question of survival; for farmers in Theni, Madurai and Ramanathapuram it is a question of livelihood. This article sets out the history and the evidence as fairly as we can, including the parts that remain uncertain or disputed. It does not try to settle the dispute; that is a matter for the courts, engineers and governments of both states.

A river turned east

The idea is older than the dam. The Periyar carries the heavy rain of the Western Ghats west to the Arabian Sea, while the Vaigai basin on the other side of the hills lies in a rain shadow and has suffered repeated droughts and famines. Diverting the Periyar eastwards was first explored in 1789 by a minister of the Raja of Ramnad, but abandoned as too costly. British engineers studied the site in 1808, 1850 and 1862, and an earthen dam was begun and abandoned. After the great famine of 1876–78, which killed millions across the Madras Presidency, the scheme was revived, and in 1884 a revised plan by Major John Pennycuick of the Royal Engineers was approved.

The 1886 lease: a colonial bargain

Because the dam site lay in the princely state of Travancore, the British needed its consent. On 29 October 1886, an indenture was signed between the Maharaja of Travancore, Moolam Thirunal Rama Varma, and the Secretary of State for India, representing the Madras Presidency. It was signed by Travancore's Dewan, V. Ram Iyengar, and the Madras Presidency's J. C. Hannington. Its main terms were:

• a lease of about 8,000 acres for the reservoir and 100 acres for the works,
• for a period of 999 years,
• at a rent of ₹5 per acre per year, giving the Presidency the right to "all the waters" of the Periyar above the dam for irrigation.

Historians have often noted that the agreement was signed between a princely state under British paramountcy and the imperial government itself, and that it is hard to call it a bargain between equals. A dispute over the use of the water arose even in the colonial period: Kerala's advocates cite a 1941 arbitration award by the umpire Sir Nalini Ranjan Chatterjee, which they read as confining the water to irrigation.

Independence and the reorganisation of states in 1956 created the legal puzzle that remains today. Travancore became part of Kerala; the Madras Presidency became Madras State, later Tamil Nadu. Kerala argued that the colonial lease lapsed with the end of British rule. After failed negotiations in 1958, 1960 and 1969, the two states signed supplementary agreements in 1970, under Kerala Chief Minister C. Achutha Menon, which confirmed the lease, raised the rent to ₹30 per acre and allowed Tamil Nadu to generate electricity at its Lower Camp (Periyar) power station on payment of a fee. Kerala has since questioned whether those agreements are valid, whether Tamil Nadu's use of the water for power, towns and industry fits within an irrigation lease, and whether the deed binds a state whose rights rest on constitutional federalism rather than imperial treaty. In 2024 the Supreme Court agreed to re-examine the validity of the lease.

The result is a structure without parallel in India: a dam that sits wholly within one state, on a river whose catchment lies wholly in that state, but which is owned, operated and maintained by another state under a lease that runs until the year 2885.

Building it: lime, rubble and malaria

Construction began in May 1887. The site was remote, deep in malarial forest, and supplies came up from the plains by bullock cart and on foot. The monsoon repeatedly washed away the temporary coffer dams that held the river back while the foundations were laid. A story often told is that when funds ran short, Pennycuick sold his wife's jewellery to keep the work going; it is widely repeated, though hard to document. What is documented is the human cost: 483 workers died, mostly of malaria, and lie in a cemetery north of the dam.

The dam was built as a gravity dam, holding back the water by sheer weight. Its faces are of uncoursed rubble masonry set in lime-surkhi mortar, and its interior, the hearting, is of lime-surkhi concrete. Surkhi is burnt brick powder; mixed with lime it makes a strong hydraulic mortar, and it was the standard material for large works in India before Portland cement became widely available. It opened on 10 October 1895. Water passes east through a tunnel about 1.7 km long into the Vaigai basin, where it is stored in the Vaigai dam, completed in 1959.

By the standards of its time it was a remarkable piece of engineering, and it has now stood for more than 130 years. In Tamil Nadu, Pennycuick is remembered as a benefactor: his portrait hangs in many farmers' homes in the Cumbum valley, and a memorial to him was opened at Lower Camp in 2013. The question today is not whether the dam was well built, but whether a structure designed with nineteenth-century materials and methods can meet twenty-first-century standards for a dam with so many people below it.

From 1979 to today: strengthening and the courts

The modern dispute began in 1979, after the failure of the Machchhu II dam at Morvi in Gujarat, where estimates of the dead range from about 1,800 to as many as 25,000. Kerala raised safety concerns about Mullaperiyar, and the Central Water Commission directed Tamil Nadu to lower the water level, eventually to 136 ft, and to carry out strengthening works. These included grouting the masonry with cement, guniting the upstream face, cable anchoring, a reinforced-concrete cap on the crest and a concrete backing on the downstream face, together with additional spillway capacity. Mullaperiyar is therefore no longer a purely nineteenth-century dam: it is an old masonry core wrapped in later concrete, and engineers disagree about how well the old and new parts work together.

The legal history since then is long:

The safety debate: what the studies say

Few dams in India have been studied, and argued about, as much as Mullaperiyar. The evidence falls into four areas.

Age and materials

Mullaperiyar is more than 130 years old, well beyond the design life usually assumed for dams of its kind. Kerala has long pointed to seepage and to the leaching of lime from the masonry, which over decades can weaken mortar from within. Tamil Nadu and the Empowered Committee pointed to the strengthening works, monitoring and test results to argue that the dam remains sound. Critics in Kerala have challenged the Committee's testing as too limited, and in 2025 the Supervisory Committee itself flagged cracks in the upstream wall among the issues needing urgent attention. A comprehensive safety review under the Dam Safety Act is due, and its findings will matter a great deal.

Earthquakes

The dam lies in a region of moderate seismicity. A magnitude 4.5 earthquake struck within about 20 km in June 1988, and smaller tremors have been recorded since. A study by Kerala's Centre for Earth Science Studies concluded that the structure would not withstand an earthquake above magnitude 6, and an IIT Roorkee study in 2009 assessed that a magnitude 6.5 earthquake could damage it with the reservoir at 136 ft. Earthquakes of that size are rare in peninsular India, but not unknown; the 1993 Latur earthquake (magnitude 6.2) struck a region previously considered stable.

Floods

The second hydrological question is whether the dam can safely pass an extreme flood. In August 2018, rainfall over the Idukki and Periyar catchments was estimated by researchers at IIT Gandhinagar to be rarer than a once-in-500-years event over periods of one to fifteen days, and the reservoir reached its full permitted level. As the climate warms and extreme rain becomes more frequent (see our story A Changing Sky), the design flood of a nineteenth-century dam is a legitimate concern.

The UN University report

In January 2021, the United Nations University Institute for Water, Environment and Health (UNU-INWEH) published Ageing Water Infrastructure: An Emerging Global Risk, by Duminda Perera, Vladimir Smakhtin and colleagues. It cited Mullaperiyar as an example, noting that it was built more than a century ago with surkhi, sits in a seismically active area, shows significant structural flaws and has about 3.5 million people at risk if it were to fail. The report is a global review rather than a site investigation, and Tamil Nadu disputes its characterisation; but it has given Kerala's concerns international weight.

If the dam failed: the direct impact

Any discussion of dam failure must start with a caution. A failure is a low-probability event; the purpose of studying it is to prepare, not to predict. With that said, the questions are serious.

A dam-break analysis by IIT Roorkee, as summarised in material placed in the public domain in Kerala, modelled an overtopping failure using standard US Federal Energy Regulatory Commission breach parameters. It estimated that a breach would develop in about 12 minutes; that the flood wave would reach Vallakkadavu in about 26 minutes and Vandiperiyar in about 31 minutes; and that it would arrive at the head of the Idukki reservoir, some 36 km downstream, in a little over two hours. The peak flow, about 88,900 m³ per second, was estimated at roughly ten times the probable maximum flood, with water depths of 20 to 40 metres in the river gorge.

The most important technical question, and the most contested, is what happens next. The Idukki reservoir, behind the Idukki arch dam and the Cheruthoni and Kulamavu dams, holds several times the volume of Mullaperiyar. The Empowered Committee suggested that in the unlikely event of a failure, Idukki could absorb the water. Kerala's experts and commentators reply that this depends on how full Idukki is at the time (a failure is most likely in the monsoon, when it is also full), on whether a sudden surge carrying debris could be passed safely through its spillways, and on whether the Idukki dams themselves could withstand the impact. Tamil Nadu maintains that the scenario is remote and that Mullaperiyar is safe. There is, as yet, no publicly agreed, independently reviewed study of the combined Mullaperiyar–Idukki system under extreme conditions. That gap is itself one of the strongest arguments for a joint, transparent study.

If the flood were contained at Idukki, the toll would fall mainly on the settlements of the upper Periyar valley, among them Vandiperiyar, Chappathu, Upputhara and Ayyappancoil, with tens of thousands of people. If it were not, the flood would continue down the Periyar past Neriamangalam, Bhoothathankettu, Malayattoor and Kalady to Aluva, into the most densely populated and industrialised part of Kerala, the Kochi region. This is the scenario behind the estimates of several million people at risk.

Beyond the flood line: the cascade

Official estimates of dam disasters usually count deaths and the value of damaged property. History shows that the consequences of a major dam failure travel much further, in space and in time. Using Kerala's own experience and case studies from India and abroad, several chains of impact can be traced.

The industrial belt and the toxic flood

The lower Periyar at Eloor and Edayar is lined with about 280 industrial units, among them fertiliser, chemical, pesticide, mineral and metal plants such as FACT, Hindustan Insecticides, Merchem and the mineral sands and titanium industries. Some store ammonia, acids and other hazardous chemicals; the river here has long been one of India's most studied pollution hotspots. A violent flood through this belt could release stored chemicals and contaminated sediment into the river, the backwaters and farmland.

The destruction of Ukraine's Kakhovka dam in June 2023 shows what this looks like. The UN Environment Programme's rapid assessment described a "far-reaching ecological disaster" whose effects would be felt for decades, as floodwater swept through chemical depots and released fuels, lubricants and fertilisers across a huge area and into the Black Sea. The Kochi region, with its refinery, LNG terminal, port, fuel depots and chemical industries on the same estuarine system, has the same kind of exposure.

Drinking water for millions

The same stretch of the Periyar is the main source of drinking water for more than four million people in and around Kochi. A failure upstream, or even a major flood carrying chemicals and debris, could take that supply out for weeks. Survivors far from the flood line would still need water, and in the meantime would turn to unsafe sources.

Disease: what history really shows

It is widely believed that the dead bodies of people and animals after a disaster cause epidemics. The evidence is more specific. The World Health Organization and the Pan American Health Organization state that people killed by drowning or injury do not carry epidemic diseases such as cholera or typhoid simply because they died, and that rapid mass burial is not justified on public-health grounds; bodies can contaminate water, but routine disinfection of drinking water controls that risk. Carcasses and bodies are a serious problem of dignity, identification, sanitation and trauma, and they must be managed properly; but they are not the main disease threat.

The real killers after great floods are contaminated water, broken sanitation, overcrowded camps, disease-carrying animals and hunger. Kerala learned this in 2018: in the weeks after the floods, more than 50 people died of suspected leptospirosis, spread through water contaminated with the urine of rats, cattle and other animals, until a mass campaign of doxycycline prevention brought it under control. After the collapse of the two dams above Derna in Libya in September 2023, which killed thousands (estimates range from about 4,000 to more than 11,000), the UN warned that water contamination could become "a second devastating crisis" as people fell ill from drinking contaminated water.

The starkest example is the failure of the Banqiao and Shimantan dams in Henan, China, in August 1975, when a typhoon brought extreme rain and 62 dams failed in a cascade. The official toll from the flood was about 26,000; an estimated 145,000 more died in the epidemics and famine that followed, and some estimates put the total above 220,000. The second disaster was larger than the first, and it was a disaster of water, food and health systems, not of the flood alone.

Animals, forests and farms

The reservoir lies inside Periyar Tiger Reserve, one of India's most important tiger and elephant reserves, and the river below runs through forest, cardamom and tea estates, rubber, paddy and the backwaters. A dam-break flood would kill wildlife and livestock, strip soil, bury farmland in debris and sand, and contaminate fields with chemicals. Kerala's own 2018 experience, and studies of other floods, show that agricultural recovery takes years, not weeks.

Clearing the mess

The physical aftermath of large floods lasts much longer than the flood. At Kakhovka, the pollution and flooded land will take years to recover. At Derna, recovering bodies and clearing debris took many months, complicated by landmines left from Libya's conflict that the flood had shifted. Kerala's 2018 floods left vast quantities of mud, waste and damaged goods, and the Kochi airport, beside the Periyar, was closed for about two weeks. A failure of Mullaperiyar on top of a major monsoon flood would test every system the state has: health, water, power, transport, police, disaster response and finance.

Tamil Nadu's loss

The impact would not stop at the Kerala border. Five districts of Tamil Nadu depend on Mullaperiyar water for irrigation, and towns including Madurai rely on the Vaigai system for drinking water. The original scheme was planned to irrigate about 68,500 hectares, and the cultivated area served by the Periyar–Vaigai system is larger today. A failure would destroy the source of that water overnight, and Tamil Nadu would lose its power station at Lower Camp. For farmers in the Cumbum valley and the Vaigai basin, the dam is not an abstraction: it is the reason the land is green. Any solution that ignores their dependence on this water will not hold.

Reconciling safety and need

The positions of the two states are deeply held, and each rests on real interests.

Tamil Nadu's case is that the dam has been strengthened and repeatedly examined; that the Empowered Committee and the Supreme Court (in 2006 and again, through a Constitution Bench, in 2014) found it structurally and hydrologically safe; that the 1886 lease and 1970 agreements are valid; and that a new dam, built and controlled in Kerala, could erode its water rights. It also argues that Kerala has hindered maintenance work that the courts have allowed, and appeals to the constitutional value of cooperative federalism.

Kerala's case is that no dam lasts forever; that the structure is old, built with lime mortar, in a seismic zone, and that the testing on which the safety findings rest was limited; that a failure, however unlikely, would be catastrophic for millions of its citizens; and that the solution is a new dam downstream, which it has offered to build at its own cost while guaranteeing Tamil Nadu's water.

Engineers, lawyers and policy analysts have proposed several ways forward:

• An independent, transparent safety review under the Dam Safety Act, 2021, with international experts, published in full, including a joint dam-break study of the Mullaperiyar–Idukki system.
• A tested emergency action plan, with flood maps, sirens, evacuation routes and drills in the downstream villages, shared real-time rainfall and reservoir data, and agreed rules for releases.
• Lower risk while decisions are made, for example through a lower operating level in the monsoon, or the second, lower tunnel recommended in 2014 to allow faster drawdown in an emergency.
• A new dam with a treaty-strength water guarantee. A new structure would only be acceptable to Tamil Nadu if its water entitlement were legally protected, jointly managed and independent of political change, perhaps through a binding agreement endorsed by Parliament or the Supreme Court and a joint management board.
• Long-term water security in the Vaigai basin, through better irrigation efficiency, groundwater recharge and storage, so that Tamil Nadu's farmers are less vulnerable to any single source.

None of these is easy, and each requires trust between the two states that has often been missing. But the history of dam disasters elsewhere suggests that the greatest danger is not any single engineering flaw. It is delay, secrecy and the assumption that because a disaster has not happened, it will not.

Mullaperiyar was built to end famine in the plains, and for 130 years it has done that. The question for this generation is how to keep that promise to Tamil Nadu's farmers while also keeping a promise to the millions who live below the dam in Kerala.

Note on sources: figures for people at risk downstream vary widely between sources, from about 3.5 million to 10 million, depending on how far downstream the impact is assumed to reach; we give the range rather than one number. Dam-break results depend heavily on the scenario modelled. The positions of Kerala and Tamil Nadu are described from court records and public statements; this article is a summary for general readers and is not legal or engineering advice. In an emergency, always follow the instructions of the Kerala State Disaster Management Authority and district administration.