This is the second in a three part series. Read the first part here.
The Koshi Project required cooperation with Nepal because embankments, in a sense, necessitate more embankments. If one builds an embankment on the right bank of a river, it will likely shift to the left, and so that side will also need protection. Embank downstream, and one must embank upstream too, lest the river meander and bypass entirely the entrance to the embankments. Thus, it was proposed that India build roughly 60 kilometres of embankments on either side of the river in Nepal, from the border to near the base of the hills.
To sweeten the deal for the Nepalis, it was proposed to build a barrage – a low dam designed to reduce water velocity – on Nepali territory, just before the Koshi crosses the border. The barrage would divert some of the Koshi’s water to canal systems irrigating several hundred square kilometres in Nepal and over ten thousand square kilometres in India. Because the barrage would have little storage capacity, its purpose was not to control floods, as the cancelled Barahakshetra dam would have done, although it would generate a small amount of electricity. India would pay for the project and operate the barrage under a 199-year lease.
In 1954, the two countries signed the Koshi Agreement, and for the first time in Nepali history, a major river was embanked. Five years later, Prime Minister Nehru, dressed in characteristic khadi with Gandhi cap, and Nepal’s King Mahendra Bir Bikram Shah Dev, wearing sunglasses and a wide-brimmed military hat, attended a foundation-laying ceremony for the barrage.
After the Koshi leapt from its embankments at Kusaha in 2008, it followed what geomorphologists call a “paleo-channel” that the river had not used in 120 years, over 100 kilometres to the east through Bihar. The Indian government was not able to fully repair the embankment until more than half a year later, when the water level had subsided. In the meantime, millions of people in Bihar and Nepal were forced to stay in tented camps on whatever high ground was available. Many survived on relief provided by the Nepali and Indian governments and international aid groups.
Surprisingly, the breach happened when the river was relatively calm: the discharge was 4,320 cubic meters per second, or cumecs, only about 15% of 28,500 cumecs the embankments were designed to withstand. Poor maintenance of the structure was clearly a large part of the problem; local people claimed that the section of the embankment in question had not been maintained in seven or eight years.
On the Indian side, some argued that Nepal’s political instability had contributed to the neglect. A general strike as part of an ongoing Madhes political movement had ended just four days before the breach, limiting access of Indian engineers to the site. The Times of India ran a headline: “Did Nepal ‘neglect’ lead to Kosi flood?”
Nepalis were understandably irate. The 1954 agreement with India had clearly placed India in charge of maintenance. Upendra Yadav, then Nepal’s Foreign Minister, blamed Indian corruption, telling the Kathmandu Post, “India should understand that the strong nexus between the contractor, engineers and Bihar Government officials was responsible for the Koshi Disaster, regardless of how India explains it.”
Pushpa Kamal Dahal, who had just become Prime Minister of Nepal, called the Koshi Agreement “a historic blunder.”
By this time, water relations between Nepal and India had already experienced some vicissitudes. Following the Koshi Agreement in 1954, two other agreements were signed to develop major shared rivers: the Gandak (known as the Narayani in Nepal) in 1959, and the Sharda River (known as the Mahakali in Nepal) in the 1990s. On both rivers, India constructed barrages on, or near, the border for shared irrigation purposes, but Nepal was entitled to a relatively small portion of the water. Critics argued India had taken advantage of Nepal, and popular sentiment turned strongly against the agreements after they were signed. Similar issues were at play with the Koshi Agreement. The barrage irrigated far more land in India than Nepal, and India’s use of Nepali territory was seen by some as an affront to national sovereignty. And while fewer people lived within the embankments on the Nepali side than in India (part of the Nepali side was declared a national wildlife reserve in the 1970s), there had been long-standing disputes over the Koshi Project’s compensation to Nepali flood victims living in the vicinity.
While the blame for the Koshi breach may have rested with India, by 2008 the Nepali government had also constructed river embankments of its own across the country. Many of these breached on a routine basis.
An example could be seen just 10 kilometres away on the Sundari River, which flows through northern Saptari District. The Sundari is relatively small, with a catchment area of just a couple dozen square kilometres in the Siwalik range, the lowest range of hills that abut Nepal’s plains. Like many Siwalik-fed rivers – of which there are hundreds across the Nepali plains – the Sundari is dry for much of the year, but runs in full spate during the monsoon. Because of the friable nature of Siwalik rock, the river also carries a relatively high sediment load.
Sitting in his home’s shady veranda near the right bank of the Sundari, Surendra Prasad Chaudhary, 59, a bespectacled school teacher with silver hair, vents his frustration with the embankments, which he says have existed for at least the past four decades.
“This embankment is always a problem. They’ve repaired it many times. They repair it, and then it breaks again one or two years later.” Chaudhary says that between his home and the highway, 4 kilometres away, the embankment breaches at some point every year.
A kilometre or so up-river from Chaudhary’s home lies the village of Raghunathpur, where the embankment has breached several times, including just last year. The village has a mix of castes and ethnic groups, including Brahmins, middle-caste Yadavs, and the most marginalised caste – the Dalits. The Dalits occupy the low-lying land closest to the river, while the higher castes inhabit plots at a safer distance.
Shardaa Sardar, a Dalit woman clad in a red and green sari, stands in a bamboo grove on the outskirts of Raghunathpur, keeping an eye on a group of goats she is tending. She guesses she is between 50 and 60, and recounts the difficulties embankment breaches have given her family. One year, a torrent of water dumped sand on top of her fields, rendering them infertile for years. When the embankment breached last year, the family had to sleep outside for days; her grandchildren were troubled by fever and diarrhoea. She worries about losing her few assets – goats, a water buffalo, and pigeons – if the embankment breaches again.
“It’s hard to sleep at night,” she says.
Metres away, a team of labourers working for a contractor – Siraha-based Nice Construction – are reconstructing a 200 metre stretch of the embankment where it breached last year. Using a back hoe and tractor, they dig sand from the river bed and dump it at its edge, behind a bamboo revetment. The embankment they are constructing is about 3 metres wide and 1.5 metres high. According to the government tender, the cost is NPR 15,000 (USD 137) per metre – on the low-end for embankments in Nepal. Soil embankments girded with boulders in gabion cages NPR cost 30,000 to 35,000 rupees per metre (USD 266-310) while similarly-sized reinforced cement concrete retaining walls can run over NPR 50,000 (USD 443) per metre to construct.
Standing on top of the embankment, Mohammed Samir Rain, the foreman, notes that the work is divided up between two contractors: one for the bamboo revetment, another for the earthwork. Meanwhile, two separate authorities – the Chandra Canal Office and a local user’s group – are constructing other sections above and below Nice’s section. Coordination has proved difficult. Pointing to a gap of about 10 metres between Nice’s embankment and an upstream spur, Rain says, “If this area is left out, then it will be bad.”
Nearby, at a chowk in Raghunathpur, locals sit in front of a blue tuck-shop built on wooden stilts, listening to an old Bollywood tune on a scratchy radio.
Shaha Dev Ram, 61, a local resident, is sceptical of the quality of the embankment work. “If they brought soil [for the embankment] from elsewhere, it would sit well and be strong – it wouldn’t get taken away by the flood. This sand is light, and it will be carried away, just like before.”
Others complain that repair work is usually done at the 11th hour, just as the fiscal year comes to a close and the government scrambles to spend its remaining budget. Unfortunately, they explain, this coincides with the onset of the monsoon, when construction work is most difficult.
People also cite corruption as a problem. RK Dev, 78, says he has lodged a legal case against engineers who oversaw the construction of the embankment that broke last year, whom he thinks connived with a previous contractor to pocket some of the money.
“Where is the weakness? It’s really with the government,” he says, pulling on a cigarette. He adds, in halting English, “People should perform their duty honestly. God helps those who help themselves. Honesty is the best policy.”
Although some are older, most of Nepal’s embankments date from the 1990s or after. Embankment construction expanded greatly after several devastating floods in the late 1980s and a cloudburst in 1993 in which almost 2,000 people were killed.
The vast majority of recent embankment construction has been funded by the Nepali government itself, although some donors are also involved. India followed up the Koshi embankments with smaller projects on other rivers: the Bagmati, Lalbakeya, Kamla, and Mahakali. The Asian Development Bank has also funded some embankment works associated with large irrigation projects on the Karnali river.
Likewise, the Japan International Cooperation Agency (JICA) provided extensive institutional support to the Disaster Prevention Technical Centre, which was created within the Ministry of Water Resources to work on flood-control in 1991. Early projects with Japanese support focused on “green embankments” planted with trees and grasses on medium and small rivers to control bank erosion, as well as measures to control erosion in upper catchments. But the Japanese also recognized the perils of attempting too much river control. A 1999 JICA report to the government recommended only short embankments to protect specific locations, noting that continuous embankments would be dangerous because of rivers’ high sediment yield and the need for constant maintenance to prevent breaches.
Nonetheless the government has undertaken increasingly ambitious embankment projects. It initiated the Janatako Tatabanda Ayojana or “People’s Embankment Project”, which is building embankments on 28 of the country’s largest rivers, including those fed by the high mountains and middle-hills. Efforts are underway to embank long stretches of Nepal’s second-largest river, the Karnali, where over 15 kilometres of embankments are complete using various budgets. Unlike the embankments on the Sundari Khola, embankments on larger rivers are usually built using extensive spurs and gabion boxes, and are sometimes combined with roads built on their surface.
Because a variety of authorities have built embankments – including district-level offices, the Department for Soil Conservation and Watershed Management, and the Disaster Prevention Technical Centre, which was renamed the Department for Water-Induced Disaster Management (DWIDM) – it is difficult to determine exactly how much the government has spent. But data provided by the DWIDM – which is the largest player – shows that it has a rapidly growing budget. It spent NPR 34,50,04,98,000, or over USD 300 million in the past decade, while its budget for 2017-18 alone was NPR 9,71,00,00,000, or about USD 94 million. The Department says it spends 85-90% of its budget on river training and nearly 900 kilometres of embankments have been constructed to date.
Although the DWIDM and other authorities sometimes work with user groups formed by local residents for small projects, all works over NPR 10 million (USD 91,000), as well as many less-expensive projects, are given to private construction companies. Such public infrastructure contract work has gained notoriety in Nepal as a field rife with corruption. A 2017 article by Nepal’s Centre for Investigative Journalism details how politically well-connected contractors routinely neglect to complete work on roads, but face few sanctions from politicians who may be receiving kick-backs. In addition, parties across the political spectrum field contractors to run for office, or candidates who take large donations from contractors, creating conflict-of-interest issues. Such corruption is facilitated by Nepal’s lack of a strong legal system to enforce contracts and codes.
The government’s increasing investment in embankments has certainly provided more opportunities for corruption and political patronage. But embankments also receive widespread popular support from communities that live along rivers.
Back on the banks of the Sundari River, Surendra Chaudhary, the schoolteacher, says that despite the frequent breaches of the embankments, he does not oppose the construction of embankments in general. “Without embankments, this whole village would be swept away,” he says, adding that he only wishes to see more permanent structures built, using boulders and gabions rather than merely sand and bamboo revetments.
This sentiment is widespread. Even in upriver sections that lack embankments, many people favour their construction. In Dharampur, 5 kilometres or so upriver from Chaudhary’s home, Prem Bahadur Biswokarma and Madhu Rai, two local residents who live near the river, both say they would like an embankment to be built.
“They need to build an embankment here so that the river won’t come this way,” says Rai. “People are becoming fed up.”
Niru Pokharel, the vice mayor of Kanchanrup Municipality, which includes most of the Sundari River, lives in Dharampur in a stately home on a small hill surrounded by a mango orchard. Sitting in a gazebo in the garden, she says her constituents all want more embankments, although there is sometimes opposition from those whose land is used for the construction. (As a rule, the DWIDM does not provide compensation to landowners.)
“We need a permanent process for dealing with this river,” she says. “Currently, it is all temporary. They build a small embankment in one place, and build a bamboo revetment. I think it would be better to build in a permanent way, adding a little piece every year.”
Most embankment work is currently carried out by the federal government through the DWIDM, but local governments across the country seem poised for an expanded role in embankment construction.
Nepal’s new 2015 constitution, which the country is slowly implementing, lists “disaster management” and “water use” as concurrent powers shared by the federal, provincial, and local governments. And according to the DWIDM, the federal government plans to devolve responsibility for embankment construction projects less than 10 kilometres long to the provincial level, and smaller sections to municipalities. Many municipalities, including Kanchanrup, have already purchased heavy earth-moving equipment, which are in frequent use for other infrastructure construction like roads.
Read Part III: Sediment and solutions
This is the second in a three part series on the history of embankments in India and Nepal along the Koshi, a major tributary of the Ganga. Read the first part here.