Floods in Indian Rivers: Are Dams and Embankments the Solution or the Problem?

Flash floods, suspected to have been precipitated by a glacial lake outburst, have wreaked widespread destruction in Uttarakhand’s Chamoli district, starting on 7 February 2021. Over a 100 people were reported to be missing in the days after the glacial lake outburst flood, as two hydropower projects were severely damaged—the missing included labourers working at the project sites.

This is not the first flood disaster in the country to be linked to a river project or a dam. Do dams and hydroelectric projects on rivers damage the riverine ecology and exacerbate damage due to floods? We look for answers in the EPW archives, with this reading list.

Floods—Natural or Man-made Disaster?

In the classification of disasters into natural and man-made, floods are often categorised under natural disasters. For instance, the National Institute of Disaster Management lists “floods” under the subhead of “natural disasters.” Similarly, the World Health Organisation calls floods “the most frequent type of natural disaster.”

Yet, such a characterisation masks the human causes of flood damage. Shekhar Pathak (2020) observed: 

Floods have been a part of the natural system of the earth from the earliest times, along with earthquakes, landslides, avalanches, hurricanes and tsunamis. However, since the introduction of agriculture and urbanisation, human activity has been directly contributing towards floods.
 
… Man-made structures such as dams and barrages, hydropower projects, unsustainable mining, deforestation, catchment degradation and encroachments in the riverbeds and climate change have also contributed to the cause and nature of floods.

In the context of the annual flooding of the Brahmaputra River in Assam, an EPW editorial (2017) wrote:

Although the recurrent floods are a natural phenomenon, they are also an outcome of anthropocentric interventions. It is natural that the high precipitation in the Himalayas—the catchment of most of the tributaries of the Ganga and Brahmaputra—coupled with the sudden fall in altitude results in a large volume of water gushing down river channels from the Eastern Himalayas into the floodplains. This water exceeds the carrying capacity of the river channels resulting in a spillover into adjoining areas. But with increased deforestation in the Eastern Himalayas, the surface run-off has increased at the cost of infiltration leading to tons of sediment being deposited on the riverbed as the river reaches the plains. This further reduces the carrying capacity of the river and enhances the risk of flooding. The plan to build large dams in upstream areas, largely in Arunachal Pradesh, is likely to exacerbate this process.

Role of Dams

While dams are often touted as a method to control flooding of rivers, they are also at the root of many flood disasters.

Himanshu Thakkar (2018) explained:

In theory, every dam can help moderate floods in the downstream areas, as long as it has space to store water, and depending on the amount of space available. In fact, every action that helps store, hold, recharge (to groundwater aquifers), or delay the flow of rainwater from the catchment to the river helps moderate its flow, and, in turn, moderates floods in the river. However, our catchments are fast losing this capacity, due to the continued destruction of natural forests, wetlands, local waterbodies, and also the soil’s capacity to hold water.
… In practice, the potential capacity of dams to help moderate floods can be realised only when they are operated with this objective in mind. When dams are not operated with such an objective, and are, instead, filled up as soon there is water available, there is no space left to store more water. The only alternative then is to release all the inflow into the downstream river. Due to this, in downstream areas, which are already facing floods due to local rainfall or other reasons, the dams end up increasing the magnitude of the flood disaster.

He supplemented his explanation with examples of floods in India that have been exacerbated by dams: 

There are numerous instances of this kind, including the floods in Uttarakhand (June 2013), Tehri (September 2010), Hirakud (2009, 2011, 2014), Damodar dams (multiple years), Krishna basin dams (2006, October 2009), Ukai (August 2006), Chennai floods (December 2015), Bansagar dam (August 2016), Kurichu dam in Bhutan (2004, 2016, others), and Ranganadi (2017) and Doyang (2018), among other dams, where flawed operation of the dams created or worsened flood disasters in the downstream areas.

Building on Thakkar’s analysis, Pathak (2020) wrote:

Floods caused by a dam are sudden, and given its intensity and the unpreparedness of people living in surrounding areas, its impact is more destructive to lives and property. The sudden nature of floods caused by dams, together with the reluctance of authorities to share information with people, makes communities vulnerable in a two-term monsoon region like Kerala. In an undammed river, the flood water rises over a period of time, which allows people to respond. Moreover, due to the dam the contours of the riverbed, the flow of the river downstream and even the floodplains change. Besides, dams tend to give a false sense of security because people tend to forget even the normal impact on rivers due to the monsoons.

Focusing on the phenomenon of floods in Kerala, Pathak analysed:

A combination of heavy rains and breaching of the dams triggered the 1924 floods that killed more than a thousand people and damaged property, including houses, roads, bridges as well as livestock and standing crops. The 2018 floods have been called a “repetition of this tragedy,” as they were also caused due to the combined strength of heavy rains and inefficient dam management.

Even though he identified separate causes behind the floods in Kerala and Uttarakhand based on the different ecological factors at play, the common thread remained dams. Speaking of the 2013 Uttarakhand floods, he wrote:

Natural silt, which comes with rivers from glaciers, debris created by faulty road-building, construction debris and muck deposited by hydroelectric companies near riverbanks accelerated the flood. The collapse of many hydroelectric power projects in different valleys increased the devastation. The course of rivers were blocked by the immense amount of debris that caused major overflow. 

The Ravi Chopra Committee (2014) stated that dams played a role in worsening the disaster in the Uttarakhand floods.

Provided further examples of “dam-induced” floods, Mirza Zulfiqar Rahman (2020) elucidated the problem of “infrastructuring floods”:

The downstream communities of Assam of the 405 megawatt (MW) Ranganadi hydroelectric project in Arunachal Pradesh, for instance, experience catastrophic dam-induced flash floods in the monsoon months, while for the rest of the year they see the river as a trickle. The communities talk of “theft of their river,” of “run-away-with-the-river” while they were promised a “run-of-the-river” dam project, and had no experience of dam-induced flash floods before the dam was built in 2002 (Rahman 2014). This underlines the aspect of infrastructuring floods, and of creating new flood discourses.
 
Such dam-induced flood hazards are spread across the Brahmaputra river basin, for instance, the Doyang and Kopili hydroelectric projects in Nagaland and Assam respectively, and the several dam projects being built in upstream Bhutan, China, and the multitude of proposed small, medium and mega dams in Arunachal Pradesh.

The problems of flood damage due to dams need to also be seen in light of the government’s disaster management preparedness. In this regard, Pathak highlighted:

Around 5,254 completed dams in India are an integral part of flood management, apart from storing water for irrigation and generating power. The Comptroller and Auditor General of India (CAG) report of 2017 submitted in Parliament said that there is an emergency action plan available for only 7% of these 5,254 dams (CAG 2017). For the 61 dams in Kerala, there are no emergency plans. 

Role of Embankments

Apart from dams, another oft-touted method of flood management is embankments. But these too have been known to be ineffective. 

For instance, an EPW editorial (2020) observed:

In both Bihar and Assam, the approach of the government policies to mitigate the impact of floods has been mainly focused on building embankments. However, the increasing intensity of floods had made these embankments largely ineffective. Moreover, no cost–benefit analysis has been done so far to determine the effectiveness of embankments. 

Instead, embankments have been criticised for their counter-effective impact of exacerbating floods and causing flooding over a more expansive area. J Albert Rorabacher (2008) wrote about the expansion of flood-prone areas in Bihar after the large-scale construction of embankments post-independence: 

The British had attempted to “tame” the Damodar, beginning in 1854 but by 1869 they had abandoned the idea of embankments completely. Never again would the British attempt to control the flooding of Indian rivers using levees or embankments.
In 1954, when the Bihar flood policy was first introduced, Bihar had approximately 160 kms of embankments. These embankments were referred to as zamindari and maharaji bandhs and were administered by the department of revenue. At this time, the flood-prone area in the state was estimated to be 2.5 million hectares. Upon the completion of the system of embankments, 3,465 kms of embankments had been constructed and were administered by the Water Resources Department (WRD) [Krishnakumar 1999]. However, the amount of flood-prone land increased to 6.89 million hectares [Karunakaran 2004:1-2]! All one needs to do is look at a topographic map of Bihar, to see that the flood-prone area is now also the area the government says is already protected!

An EPW editorial (2017) described how embankments can be dangerous both in areas that they divert water away from and in areas that they divert water to:

Embankments were constructed to create a “safe” area for habitation and they provide these in areas where the embankments are new. But in areas like North Bihar and Assam, where there has been a fairly long history of embankments, the situation is complicated. Large populations continue to stay inside the embankment, that is, outside the “safe” areas, at the mercy of the imminent flood. Unfortunately, people located in the “safe” area also live in constant fear of embankment breach. Their fear is not completely unfounded—the floods in Assam this year and the Kosi River flood of 2008 were outcomes of embankment breach. Furthermore, people living inside the embankment face the risk of flash floods and sailaab floods. The latter is due to a gradual increase in water level, while the former, as was seen in Uttarakhand in 2013, occurs when there is sudden high discharge from a reservoir into the river channel.

Rorabacher listed the main problems with the use of embankments in flood management in Bihar—the build-up of silt, sand, gravel and stone in the riverbeds, increased height of the mean water level of the rivers, and scouring of embankments leading to catastrophic breaches. He explained:

Prior to the construction of the bandhs, when the rivers were in spate, they would gently sweep across the landscape, depositing their burden of silt. As the rivers were channelised, their velocity increased, thus allowing the waters to carry, in addition to silt, sand and gravel and rocks. Initially, this burden was deposited in the river channels, as defined by the levees … the mean level of the rivers rose, as the original river bed was filled with debris. Soon, the mean levels of the rivers were higher than the original floodplains. In some areas, the river levels increased to heights of three to four metres above their floodplains, requiring that the levees be further strengthened and elevated.
… Before the embankments, the sand, gravel and stone that the rivers carried was released soon after the rivers entered Bihar. Farther downstream, the waters carried only silt. Now, sand, gravel, stone and silt are carried much farther downstream. The sand, gravel and stone, in addition to filling the old river beds, scour the face of the embankments. In time, the scouring action eats away at the embankments and creates holes or breaches in the levees. 

He further explained that floods caused by embankment breaches were more severe than the regular flooding of the rivers:

The breaches, in and of themselves, would not be catastrophic but the increased height of the mean river levels causes them to become catastrophic. Instead of simply spreading out over the flood plain, the rivers now cascade/crash down onto the floodplains below. These cascades carry with them sand, gravel and stone. These are then deposited on the surrounding landscape. In many areas, the accumulation of sand and gravel has made cultivation impossible.
… Some villagers insist that they are prepared to deal with natural floods. “It was man-made floods that wreaked havoc on them [farmers] all these years” [Karunakaran 2004:3]. The floods that come through the unrepaired breaches rise slowly, not like bulldozers smashing everything in their path, as happened when the levees breach suddenly and without warning.

Another adverse effect of the embankments in Bihar is the loss of agricultural productivity of the soil in the region:

Between waterlogging and lands lost to sand and gravel deposits, cultivators have lost well over 8,00,000 hectares of previously cultivable land. At every turn, the land base of the cultivators of Bihar is shrinking, all in the name of flood protection. The loss of land means a loss of agricultural production.

Kumud Bhushan Ray (1954) warned against embankments as a flood management intervention, citing the example of the Mississippi river in the United States:

It will thus be seen that the flood embankments along the Mississippi continued to be breached in spite of their heights having been raised, and it became necessary to carry out channel improvement works at a considerable cost annually… Thus even ordinary flood, not to speak of major floods, are likely to cause breaches in flood embankments in China or India in some year or other, as had happened in the Mississippi embankments before 1927. Moreover, the fact that embankments along the rivers in China have now been strengthened and that they are being well maintained will give inhabitants of river-side lands a sense of security, so that they will cease to be on the alert. Hence when flood embankments arc breached and lands are suddenly submerged, there is the danger of greater flood damage along the embanked reaches of the yellow, Huai, and Yangtze rivers in China and the Kosi and Brahmaputra in India.

Politics of Flood Management

Despite the availability of research and literature on the drawbacks of infrastructural approaches to regulating floods, why do dams and embankments continue to be constructed on rivers?

Rahman (2020) placed the blame of such policies on the technological fix mindset and extractive nature of the water bureaucracies or “hydrocracies”:

The abilities of the hydrocracies in promoting technological fixes, lock-ins and regimes through their technical expertise, material capabilities, discursive power, knowledge production, building of narratives, agenda-setting capabilities, securitisation and political support to influence and gain from river development projects make for a heady mix of power and hubris. (Molle et al 2009; Mirumachi 2015)

In the context of floods in the Brahmaputra, he wrote about the interrelated motives of the state and private dam construction companies in contrast with the risks for riverine communities:

The geological and hydro-climatic hazardscape of the Brahmaputra river basin, overlapping with the eastern Himalayas, is under a perpetual shadow of unequal risk, constructed by aggressive state-led hydropower development policies (Huber 2019). The deployment of the narrative of hydropower dams in the hands of public and private dam construction companies, with scant regard to cumulative environmental impact assessments or public consultations (Vagholikar and Das 2010) brings such extractive policies of the hydrocracies as a site of popular public resistance by riverine communities against the dispossession of their lives and livelihoods (Baruah 2012). The cumulative layers of jeopardy are materialised for communities living in the Brahmaputra river basin through such embankment-induced and dam-induced floods.

These contrasting political interests are embedded in the “development” projects in the region and are often tacked on with “national security” dimensions:

The hubris of large river-engineering projects in a geologically fragile and high-seismic region such as the eastern Himalayas comes from strong political support, and an overarching national security frame, which supersedes any traditional community world views of the river. The strong political backing to the hydrocracies is seen in India (Piper 2017) as well as in China (Ball 2017). Large river-infrastructure projects, such as hydropower dams in Arunachal Pradesh, for instance, are pushed in the name of national security, the logic being if China is building dams upstream, so must India. This has led to a sense of democratic deficit among riverine communities in North East India in their abilities to participate in decision-making on environmental governance of rivers, ecologies that are sacred to their socio-cultural world views. (Rahman 2019)

While such technological interventions have become commonplace, they continue to have a real effect on increased flood damage in the region.

The art of infrastructuring floods is not something new, and history is replete with examples of flood control engineering projects causing massive floods, both within India (D’Souza 2006), and in neighbouring China, where the rural countryside were deliberately flooded to protect important cities, during the 1954 Yangtze floods (Courtney 2018). The catastrophic floods seen in the Brahmaputra river basin as well as in China’s Yangtze river basin in 2020, are stark reminders of how the respective hydrocracies of India and China imitate each other in many ways, infrastructuring floods and inflicting annual suffering upon riverine communities in the longue durée.

Adding another dimension to the problem of politicisation of floods, Rorabacher (2008) argued that one reason why embankments continue to remain a part of the flood management strategy in Bihar is the vested interests of politicians and businesses: 

The cost of the embankments is not just an issue of upkeep or maintenance, which amounts to millions of rupees annually. The maintenance of the embankments is big business for engineers, contractors and politicians. Each group has a vested interest in keeping the levees in place. None wants to cut the throat of its own personal cash cow.

He went on to contend that flood disasters themselves could be beneficial to certain interests:

Added to the annual maintenance costs are the relief costs. Again, contractors and politicians have no reluctance accepting the money or praise their relief efforts engender among those being assisted. Entire humanitarian aid organisations would lose their mandate without the annual floods in northern Bihar. They, too, have no reason to resolve the problems of the people they serve. If the reports are to be believed, government-sponsored relief reaches no more than 10 to 15 per cent of the affected people. Where do all the monies and supplies go?

Benefits from flood disasters could also be in terms of gaining political mileage. In the context of recurring floods in coastal regions of Odisha, Kishor C Samal (2011) highlighted:

Politicians belonging to the state’s ruling party were seen openly indulging in the “politics of relief”. Even as the state government was conducting relief operations, these leaders would visit the affected areas almost daily to distribute relief materials they had collected from the private sector. In the worst affected areas the scene resembled that of the pre-poll days. Luxury vehicles carrying these politicians flocked to these areas and they did not forget to ensure that the media followed them everywhere (The Hindu, 22 September). It has been alleged that almost all the ruling party politicians badgered the district administration to keep their constituencies in mind during relief distribution. Their concern was no doubt also due to the massive compensation funds that were to be distributed in the flood-affected areas and the equally huge sums that would be spent for reconstruction and rehabilitation works.

Even when vested political and financial interests are not imputed, bureaucratic inefficiencies of the state contribute to mismanagement of floods. On the floods in parts of West Bengal in 1998, Nileen Putatunda (2001) wrote:

It was not so much the torrential rain in northern and western India that caused the devastating floods in Malda in 1998. The blame must lie with the state government and Farakka Barrage authorities who failed to take proper preventive measures. The real reasons for the flood were breaches in the 1.5 km embankment along Farraka Barrage, weakened by theft of mud from its walls; failure to erect spurs – long poles on the banks to control direction of flow – along the Ganga at Farakka and failure to dredge the Ganga. There might have been some respite for Malda had the 27th and 28th spurs at Manikchak, along the Ganga, been put up to deflect the water current; but the Farakka Barrage authorities, who were in charge of the Rs 80 crore project, delayed it for months. When they finally decided to issue tenders, it was a case of too little, too late. Floodwater had submerged towns, villages and highways and left thousands homeless.

Rethinking Development to Manage the Impact of Floods 

The Fifth Assessment Report of the Inter Government Panel on Climate Change (IPCC) released in February-March 2015 suggests an increasing vulnerability to natural disasters like floods in the coming years.
… The National Flood Commission of India states that 40 million hectares of land in India is flood prone which makes India one of the highly flood prone nations of the world. 

—noted Aprajita Singh (2015), emphasising the “urgent need for the government to rethink its flood management approach and practices.”

The approach to floods needs to look beyond infrastructural interventions and adopt inclusive policies that take into account stakeholders from riverine communities. What needs to be understood is that flooding of rivers is a natural phenomenon that need not devolve into a disaster. Calling the river a “living reality that needs space,” Putatunda (2001) wrote:

Inundation is a natural phenomenon, but when it affects people and damages property, we hype it as ‘the ravages caused by floods’. What we forget is that the entire area in which homes and cultivation fields have developed – the Bengal delta – is the creation of annual inundation over millions of years. It is only greed that has forced settlements right on the banks, reclaiming land prematurely, against the wishes of nature. Now, the river has been forced to rebel and hence the floods. The river is a living reality that needs space.

In this regard, Dinesh Kumar Mishra (2001) highlighted the futility of technological methods of “flood management:”

Modern flood control technologies have neither been very successful nor are they people-friendly. This is because their focus is on trying to control the waters of rivers in spate and not of making use of the flood waters in the best possible way, while ensuring the least damage.

Accordingly, an EPW editorial (2017) proposed:

What is required is an important normative shift that sees floods as a natural phenomenon, and a change in the discourse from flood protection to flood governance. Flood protection necessarily starts and ends with structural intervention and provision of relief. Flood governance would require the innovative combination of initiatives undertaken at various levels. At one level, it is important to conduct “strategic environment assessment” of all development activities in the ecologically pristine locations of the Eastern Himalayas and aim for river basin management. This might include some structural interventions. At the institutional level, strengthening the moribund Brahmaputra Board in Assam, and staffing it with scientists from a broad range of disciplines is essential. But the most important shift would be to plan a comprehensive initiative to build resilience within the riverine population through an integrated set of interventions which should be based on three pillars: reducing vulnerability, enhancing access to developmental services that flood-prone populations are deprived of, and creating conditions that enable the optimal use of people’s resources.

Pathak (2020) also argued: 

Our approach to conservation, equity and climactic changes and their impact on everyday life, and on calamities should be relooked at in order to evolve new ways of sustainability and eco-sensibility, which are needed in both the regions [the Himalayas and the Western Ghats]. Natural calamities will be there, but we will be able to minimise their fury and the loss of human lives if our approach takes into account the changing realities of climate change.

Flood management can also not be viewed in isolation and has to be integrated as part of a holistic environmental policy that rethinks “development” altogether. An EPW editorial (2020) wrote:

Environmentalists have argued that an approach that integrates water management with land use planning, agriculture, and ecology is needed to manage floods. For a long-term sustainable solution, a “basin-wide approach” that would bring together all the basin-sharing countries and states has been advocated to address the problem at the source. Additionally, it needs to be recognised that the political and economic systems that prioritise overconsumption and growth have also led to the destruction of nature. This being the root cause of recurring natural disasters, there is an urgent need to pursue alternative models of development that also accord importance to nature and its conservation.

Read More

India’s Relief and Rehabilitation Efforts Need to Be Revised for a More Inclusive Approach to Disaster Management | EPW Engage, 2019

Dams and Development | Anant Phadke, 2002

Floods in Uttarakhand: A New Deal Relief | Hui Chi Goh, Mehul Pandya and Mihir R Bhatt, 2013

Flood Disasters and Forest Villagers in Sub-Himalayan Bengal | Bidhan Kanti Das, 2009

Flood Prevention in the Rivers of Bihar, North Bengal and Assam | Kumud Bhushan Ray, 1954