Moving On - The Case of the Adyar

By shash

7 February 2016

On February 1st, a report appeared in the Economic Times with the title "Tamil Nadu response to flood was slow, says central analysis". Among other things, this report claims,

reports suggesting that rains and the excess water released from the dam at Chembarambakkam resulted in the flooding of many parts in the city, are being looked at.

It appears now that this central analysis mysteriously doesn't exist, but this particular charge - that the main reason for the flooding was the release of waters from the Chembarambakkam into the Adyar. Further, according to this story, PWD (or other government officials - the details vary with the teller of the tale) had to wait for instructions from the CM's office to release water from Chembarambakkam, and that, had the water been released earlier, the city would have been saved. Apparently, the BBC weather service had predicted heavy rains, and if their warnings had been heeded in time, everything would have been fine, but the bureaucratic delay in opening the sluice gates at the tank is what caused so much of the city to inundate.

To anyone who knows the hydrology of the city and its history, this narrative should immediately sound suspicious. But to really make sense of it, we need to look into the structure of the Adyar River and the Chembarambakkam Tank.


A key concept I want to introduce here is the "watershed basin". The US Geological Survey defines[1] a watershed as

the area of land where all of the water that falls in it and drains off of it goes to a common outlet.

A watershed is an area that drains into a single water body - this could be a stream or river, a lake, or even the sea. When thinking of a particular water body, a watershed is all the bits of land from which water flows into that body.

The map below shows the major watersheds in the city. I will keep returning to this map - which I made using satellite images from ISRO[2]. Each of the named areas feeds into that particular water body. Adyar's watershed is in green.


Adyar the crowd-sourced river

The Adyar is probably the second most misunderstood part of the hydrology of the city and its surroundings; it is not a typical river with an upper, middle and a lower course. Under the usual classification, Adyar is all lower course, with everything happening on the flood plains of Chennai itself.

The other two rivers of Chennai are both linked to the Palar - the Koshastalaiyar rises from Palar near Kaveripakkam, and the Cooum rises from Koshastalaiyar at Keshavaram[3]. The Adyar on the other hand, is linked to this system only peripherally, through the Chembarambakkam tank. However, Chembarambakkam is not the only source of water in the river. It is fed by something like 450 different eris,[4] discharging into various channels which eventually flow into each other, and then at a point just west of Guduvanchery, they become a recognisable river.

In this sense, the Adyar doesn't have a single source - the entire watershed region is its source. The map below shows a more detailed view of the Adyar's watershed along with some of its eris and streams.


Chembarambakkam and Adyar

Of these, undoubtedly the biggest single source is one of the biggest Eris around, Chembarambakkam. It used to be fed by two different sources. First, a large number of other Eris around Sriperumbuthur drain into it, and second, it is also fed from the channels leading into it from the Cooum, most importantly the one at Korattur and the one that flows through the Nemam-Kuthambakkam tank. Recently, a third source was added, the Telugu Ganga canal, bringing Krishna water to the city. In the case of the December rains, Telugu Ganga did not really play a significant role.

Once the water is stored in Chembarambakkam, it can be drained through the sluice gates into a large channel that leads into the Adyar near Thiruneermalai, passing through Sikkarayapuram, Kundrathur, Thirumudivakkam and Thandalam, before finally joining the Adyar south of the Airport, at Thiruneermalai.


Look at any old map of Chennai, going back even up to the 1950s, and you'll notice that the southern boundary of the city was the Adyar River. It was only after the 60s that areas south of Adyar, starting with Kotturpuram and the "Adyar" locality itself were brought inside city limits.

The channel connecting Chembarambakkam to the Adyar has existed for longer than we have a map of this area for; in the Jaghir Lands map it's the L-shaped channel that seems to stop just short of the river.


Until the 1990s, Chembarambakkam was still an irrigation tank, used to feed the fields of the areas I just mentioned, as well as all the agricultural fields of places like Porur, Poonamallee, Mangadu, even areas we would now consider deep inside the city, like Valasaravakkam and Mugalivakkam. Probably, it drained partly into Porur Eri, into the Long Tank of Mylapore and into the Adyar river.

Now, however, there is no Long Tank, and Porur Eri is less than a third of its original size. In a topographical map made in the 1960s, there are several canals leading out of Chembarambakkam, mostly for agricultural use. Now, most of them have disappeared.


Areas along its banks were not urbanised, much of it being agricultural. When the hydrological structure of the city was tweaked in the 1940s[6], the Adyar, along with Koshastaliayar, was apparently viewed as a place to dump the excess floodwaters. This was done by linking Cooum to Chembarambakkam through the Korattur Anicut. In an agricultural setting, a bit of surplus water is not a bad thing. In the context of paddy cultivation, it can be a really good thing. The changes made in the 1940s were made to protect the city, while allowing the agricultural lands outside it to flourish.

However, when it was decided in the 1980s to repurpose Chembarambakkam as a water supply source for the city, and to start urbanising the areas that it fed (these were linked decisions, by the way), we suddenly face a situation where surplus water is a calamity, not a boon.

How was this dealt with? Actually, it wasn't! The disaster was just waiting to happen.

Reconstructing the release of water

Now, let's try to reconstruct what happened on the day that the city went under water.

Some statistics about Chembarambakkam are now in order. It is 15.38 square kilometres in area, the bund is 85.4 feet high, and it has a storage capacity of 3645 million cubic feet (mcft).

On the 30th of November, the day before the massive rainfall hit the city, the level of water in the Eri was 2 feet below the maximum level at 83.42 ft and the storage was 3396 mcft.

Chennai Metrowater maintains a daily statistic of the major reservoirs around the city, showing their inflow, outflow and storage levels at which we can use to see what happened.When the rain actually happened on December 1st, there was a massive inflow into the Eri, of 26000 cubic feet every second, from the other tanks that feed it, apart from the 475mm of rain that fell directly on the eri. When water was released, it flowed out at an equally massive 29000 cubic feet per second.

First, to drain 3645 million cubic feet at the rate of 29000 cubic feet per second, it would take about 1.5 days. When you consider that there's an inflow of 26000 cubic feet of water would fill it back up in about 1.6 days, you can see that it would have made very little difference to the city whether the release happened a few hours earlier or not.

Along with the 475mm of rain, whatever water was coming into the eri was flowing out without affecting the water level as such. In fact, the level on Dec. 2nd was 84.47 ft, higher than where it had started.

The second point to be made is that the peak outflow at 29000 cubic feet proved to be more than what the city could handle. The only thing that would have been achieved by an earlier release is that there would have actually been more flooding, not less.

Finally, consider that weather predictions in the tropics are inherently unreliable. At best, they're valid for about a day or so. Releasing a large amount of water from the tank several days before, on the basis of an unreliable weather prediction, would possibly leave the city water-starved in the coming summer (or at best, two summers down if the monsoons of 2016 are not good). The mandate of the PWD and Metrowater is both to avoid flooding and to ensure the continued water supply to the city. They had to tread a fine line in this situation.

I personally think the authorities acted in the only possible way, given the enormity of the problem facing them at that instant.

A failure of planning

Does that mean that there was no failure of government? Assuredly not. Let me quote a small bit from the CMDA's master plan that almost seems like a cruel joke to us today. In Chapter X of Volume 1 on Disaster Management, it is stated that

From the flood hazard map of India (mapped by meteorological department, New Delhi), it is seen that no area in Tamil Nadu falls in the risk zone.

Of course, the Master Plan then goes on to talk about "a few areas along the rivers and canals and low-lying areas, which are susceptible to flooding/inundation during heavy storms.". From our experience in 2015, about a third of the city consists of "a few low-lying areas"

There was indeed a failure - one that goes back to the decision in the 1980s to urbanise the floodplains of Chembarambakkam that I referred to earlier. Most of these places were paddy fields, which, as anyone who has seen a paddy field can tell you, is basically a swamp with reeds growing on it. These places will be inundated when the rain comes - it is their hydrological function, after all. To permit high-density residential and commercial construction in these places, without adequate thought (actually, without any thought) to their drainage in times of flooding, was the first clear failure of planning.

What the planners failed to consider was things like the width of the flood-bank of the river and the probability of flooding. Many of the places along the Adyar where there was massive inundation, such as River View Road in Kotturpuram, West Saidapet, Jafferkhanpet, or Defence Colony and the offices of two major TV channels in Guindy, had no business actually being built upon. The reason I say this is that these areas are built on the part of the landscape that is called the flood bank. Though the river, during the lean season, will flow in a small channel at the centre of its valley, during the rains, it will swell up into the areas immediately around the channel. The level to which it swells constitutes the flood bank. Any area on the flood bank is technically a part of the river itself. Typically, no construction is allowed here.

However, it is incorrect to legally call these constructions "encroachments". They all have approvals from every possible authority for their plans, they hold their titles in clear, and they are being served by the city infrastructure as though they were always part of the plan. We can't even blame the people who live there - most of them would have moved in from far away places, after the last floods of 2005.

Nobody really told the river this...

A century of flooding

A defence that was offered by the administration, for the unprecedented levels of flooding, was that the rains themselves were unprecedented; "floods never seen in history" (in Tamil, வரலாற்றில் காணாத மழை) was the phrase used.

This is partially true; the rains are pretty much at the outer end of everything we've experienced for about a hundred years. But is it right to say that the floods themselves are unprecedented? Well... Not really...

Here is a list of the different flooding events that took place in the city[7] since the 1970s:

1976:Heavy Flood Submergence in Adayar-Kotturpuram TNHB Qtrs. Flood could not enter into sea due to High tide. Chembarambakkam Tank surplused into Adayar – 28,000 C/s
1985:Floods in Adayar - 63,000 c/s submergence of encroached flood plains
1996:Floods in Adayar, Cooum and Kosasthalaiyar Rivers Poondi Dam surplused around- 80,000 c/s Karanodai Bridge collapsed Chembarambakkam Tank surplused into Adayar – 20,000 C/s
1998:3 persons Marooned in Thanikachalam Nagar - a residential colony in the flood plains of Kodungaiyur drain
2005:100 year RF 40 cm in a day, Flood in Cooum 19,000 C/S, Adayar 40,000 C/S, Otteri Nullah, Cooum, Adayar, B'Canal, Virugambakkam- Arumbakkam Drain over flown, 50,000 people evacuated.
2008:Chembarambakkam tank surplused into Adyar - 15,000 cusecs

The Adyar, which took a nice 28000 cubic feet per second from Chembarambakkam in 1976, and which took a total of 63000 cusecs in 1985, could not handle the 29000 cusecs we saw in December. The reason for this is what I pointed to in the previous section - the thoughtless conversion of agricultural lands with minimal obstruction into high-density urban areas starting in the 1990s.

The larger takeaway from this table is that it's not like there was no warning. In the past, the amount of water in the rivers during the rain has been comparable to what it was in this monsoon, and the river has, in the past, inundated many of the same places that were flooded in 2015.

As an aside, note that the floods have come, quite regularly, about once every ten years. This is a point I will come back to in another article.

Now, notice that there was a flow of 63000 cusecs in the Adyar in 1985. But we just saw that Chembarambakkam has a maximum outflow into the river of just about 29000. Where did the rest come from?

Not just Chembarambakkam

Remember how I said that Adyar is a crowd-sourced river? That it's not just Chembarambakkam? Let's go back to that point.

The from where I took the map of the Adyar's watershed, "Assessment of Flood Potential Ranking of Subwatersheds: Adayar Watershed a Case Study" by S. Suriya and B. V. Mudgal, is, well, what it says - a study of the Adyar watershed. Thanks to them, I don't have to redo the analysis!

They divide the whole watershed into four sub-parts: Trisulam, Manimangalam, Orathur, Dasarikuppam. Chembarambakkam is part of the Trisulam watershed. Further down in the paper, they model the contribution of these four sub-watersheds to the river's flow. I personally think that the rainfall model they've used needs to be re-worked for the new data from the 2015 monsoons. Nevertheless, the percentage contribution of each watershed will still stand.

It turns out that the Trisulam watershed is the largest, but it contributes just about 36% to the river's flow. The rest comes from the other three. Thus, even without the 29000 cusecs from Chembarambakkam, about 60% of the flow in the river would still have been there. This, though probably not enough to inundate as much of the city, would still have had a massive impact!

Actually, it did.

Consider areas like Mudichur, Mannivakkam, and other places like Potheri, which were among the worst affected during the floods. All of these areas are upstream of the discharge channel from Chembarambakkam. The map below, from ISRO's Bhuvan, made on 3rd December from satellite images, shows that many areas upstream of Chembarambakkam's discharge channel were very badly affected:


Though I have not personally verified this, there are many stories abounding of Eri bunds in this region having breached and their waters flowing out into the Adyar. This isn't even strictly necessary; the Eris, like Chembarambakkam itself, were already full from the rains in November, and would only have overflowed in December.

Again, like the areas we were discussing down-stream of Chembarambakkam, places like Mudichur and Manimangalam are haphazardly urbanised, and the town panchayats in this area are probably not even aware of the complications with regard to hydrology and drainage channels when they approve these constructions.


All of this points to the real problems that caused the flooding of areas along the Adyar. A failure to appreciate the hydrological structure during planning, a failure to prevent construction in areas that were unsuited for dense urbanisation during the approval process, and a failure to plan adequately for a disaster that we knew was coming.


[2] Specifically, the CartoDEMv3 series of Digital Elevation Maps, created from CartoSAT data.

[3] There's a major controversy about this - some say it rises from the Cooum tank, others say it's from Keshavaram, but right now, the main channel comes from Keshavaram.

[4] Statistic from CMDA's Second Master Plan


[6] I will take this up in a future article.



1 Shobana Ramkumar says...

will wait for more

Posted at 10:11 a.m. on February 7, 2016

2 GRamchandran says...

Excellent. But the mute point is who will bell the cat?

Posted at 2:14 a.m. on February 9, 2016

3 RG Chandramogan says...

Interesting article in that Adayar had 29000 curses/second from Chembarambakkam as 40% [HTML_REMOVED] another 60% came from other Eris. If so 74000 causes flowed through the river. Each cusec is equal to 27 liters of water. In that case water released was equal to 19.98 lakh liters per second or close to 2 million liters per second.

Assuming to fill an acre of land with water with one foot height quantum required is 1.17 million liters only. So per second 1.7 acres of land will get a load of 1 feet of water to be absorbed. Assuming water went on this rampage for two days ( 172800 seconds) subsequent to heavy rain [HTML_REMOVED] opening of flood gates @ this rate of filling of 1.7 acres of land with 1 feet water it can fill 293760 acres of land with 1 feet deep water. At this speed it can flood 455 sq. miles within two days with 1 foot water assuming sea resists water [HTML_REMOVED] it gets flooded on the plains.

Among the plain there are high attitude area [HTML_REMOVED] low area. Leaving places of high attitude like ECR rest of the areas got more than 10 feet water. Unimaginable crime. It is nothing but Greed led suicide.

Posted at 11:17 a.m. on February 12, 2016

4 K.S.Balasubramanian says...

Excellent Article. There is no use blaming the Govt/authorities for the situation. We have to blame ourselves for moving out to the areas and constructing houses knowing well it was once a lake/tank. In the recent floods at Mudichur, Tambaram west which was submerged is a typical example. The width of the river is narrowed by about 100 mts thanks to encroachments. When the Kancheepuram collector ordered eviction/ demolition of of about 110 houses in the area water receded in a day and drained to the Thirunindravur/ Thiruvalluvar lakes. If the encroached areas are cleared this calamity could be avoided in future. The area I reside now, West Tambaram, Kadaperi a big lake called Kadaperi is encroached upon to a larger extent which flooded nearby areas. Though authorities have visited the lake, no concrete action has been taken to evict the encroached structures (Even a church with a big shed is built upon recently in the lake area. Unless our Political class wakes up, this will continue to happen once in a decade (after 2005 it was in 2015). It is time that citizens wake up from their deep slumber and take upon the task of removing the encroachments from the water bodies.

Posted at 5 p.m. on February 12, 2016

5 Mangalam Balasubramanian says...


Very good analysis. Yes our Urban Planning lack many scientific facts as you had rightly brought out in your article. Still, we have not learnt .Constructions are in full swing in many flood banks/watershed areas. Good luck Mangalam

Posted at 7:10 a.m. on February 13, 2016

Comments are closed.

Comments have been closed for this post.