A journey from Lake Bonney to Bhutan

15 February 2016

Whilst researching a paper on the use of siphons, Stephen Hughes from Queensland University of Technology in Australia realised that siphoning water from glacial lakes in countries such as Bhutan could help prevent dam failures and catastrophic inland tsunamis.

This article is a brief account of a journey, both literal and metaphorical, from a lake in drought stricken South Australia to the Alpine regions of the Himalayan Kingdom of Bhutan. The story begins in January 2009 as I stood beside Lake Bonney in South Australia watching water pour through 18 siphons replenishing the lake with water from the River Murray.

Drought stricken Lake Bonney had been cut off from the river to reduce evaporative loss but as a result the water level had fallen by 1.3m and the fish started dying. The population of the township of Barmera on the southern shores of the lake got very upset which resulted in the South Australian authorities allocating 10GL to increase the depth by 0.6m, halting the decline in the lake's health.

When I saw this siphon in operation I thought it would form a great paper about siphons operating on a much larger scale. In the course of writing the paper I discovered that there was controversy about how siphons worked. One paper in particular mentioned that many textbooks and most dictionaries contained erroneous explanations and definitions on how siphons operated. Since the paper was published in 1971, and I was reading it in 2009, I presumed that the problem was now resolved and wasn't going to explore it any further.

However one day at home, on the spur of the moment, I reached out and plucked our copy of the 1911 edition of the Oxford English Dictionary from the shelf. I let out a cry of disbelief when I read the entry for siphon: "a pipe or tube shaped like an inverted V or U with unequal legs to convey a liquid from a container to a lower level by atmospheric pressure."

The next day, back at Queensland University of Technology (QUT), I raced into the library and checked every dictionary I could lay my hands on. Every one asserted that atmospheric pressure was an essential ingredient in the operation of a siphon. I then checked every online dictionary I could find - the same story. When visiting people's homes I would look through their dictionaries - all atmospheric pressure.

However, as a trawled through all the dictionaries I could find, I came across this definition of the siphon in Encyclopaedia Britannica - an "instrument, usually in the form of a tube bent to form two legs of unequal length, for conveying liquid over the edge of a vessel and delivering it at a lower level. Siphons may be of any size; they are used in civil engineering to transfer water or other fluids over elevations. The action depends upon the influence of gravity (not, as sometimes thought, on the difference in atmospheric pressure-a siphon will work in a vacuum) and upon the cohesive forces that prevent the columns of liquid in the legs of the siphon from breaking under their own weight. Water has been lifted more than 35 feet (11 m) by a siphon."

A few weeks later I was in a bookshop in Brisbane when I came across a copy of the Oxford Dictionary of English that had the following definition of the siphon which is essentially correct (at least in my book) - "a tube used to convey fluid upwards from a reservoir and then down to a lower level by gravity. Once the fluid has been forced into the tube, typically by suction or immersion, flow is maintained by the different fluid pressures at the tube openings."

As a result of the controversy triggered by my paper the definition of the siphon was changed by the Oxford English Dictionary and the current definition is: "a tube used to convey liquid upwards from a reservoir and then down to a lower level of its own accord. Once the liquid has been forced into the tube, typically by suction or immersion, flow continues unaided."

This definition makes no mention of gravity or atmospheric, presumably to reflect the current doubt over how siphons work. However, in my opinion, terms like "of its own accord" and "unaided" makes the siphon seem like a device out of Harry Potter that works through magic.

Although this may seem strange, I have come to the conclusion that the original atmospheric definition of the siphon in the OED is correct. Not scientifically correct, but correct in the sense that it correctly reflected the prevailing view on how siphons work. The remit of the OED is not to define how the English language should be used but how it is used.

How siphons work

There are two main views about how siphons work. One is that as liquid flows out of the siphon negative pressure created at the crown enables atmospheric pressure to push liquid into the inlet. A different view is that as liquid flows out of the siphon there is in effect a chain link between the two ends of the siphon. A key experiment demonstrating that this is so was performed by Adrian Boatwright at the University of Nottingham in the UK who got an ionic liquid siphon going in a vacuum.

An argument in favour of the atmospheric mechanism is that the maximum height of the siphon is 10.3m at sea level, which is the height of the column of water that can be supported by sea level atmospheric pressure. If we were to construct a water barometer, at sea level the height of the column would be 10.3m high making barometric pressure a bit difficult to read -a water barometer would have to be placed in a star well. Mercury barometers are much more convenient since they only need to be about 760mm high to read atmospheric pressure as mercury is 13.5 times denser than water.

However, the reason that siphons fail above 10.3m is not because the column cannot be supported by atmospheric pressure, but because water boils in the low pressure generated by the tension of the water pulling down on either side of the crown. This happens when the pressure at the crown of a siphon goes below the vapour pressure of water at a given temperature. In this case the siphon reverts to two back-to-back barometers.

Journey to Bhutan

The siphon controversy ultimately led me to Bhutan. I have been working with my colleagues Professor Les Dawes from QUT and Som Gurung of the Royal University of Bhutan to investigate the use of siphons to drain water from glacial lakes.

Bhutan, like the rest of the world has a major problem. Climate change. The Himalayas, sometime referred to as the third pole, have warmed significantly over the last five decades and this has led to the contraction of the glaciers. A study of 103 glaciers in Bhutan between 1963 and 1993 revealed that 87% of them had shrunk by an average of 6m per year.

The glaciers are no longer in balance - in other words the amount of snowfall is less than the melt rate. There is extra water and some of this is increasing the size of lakes at the end of the glaciers. As the lakes get deeper the hydrostatic pressure at the base of the dam walls made from loose rock and stone gauged from the mountains is increasing, elevating the risk of breaching the walls. When glacial walls breach the water has a long way to fall. In the Himalayas, glacial lakes are more than 4km above sea level, and that is a lot of gravitational potential energy. A wall breach is called a Glacial Lake Outburst Flood or GLOF and can cause catastrophic flooding of communities downstream. They are also sometimes called inland tsunamis because of their destructive nature.

Bhutan has had two tsunamis over the past few decades. In 1994 a tsunami in Punakha Valley killed 21 people and caused massive destruction, while an earlier incident involving Jichu Drake glacial lake in 1968, destroyed much of the city of Paro killing an unknown number of people.

Lowering lake levels

Over the last few years attempts have been made to lower the water level in Bhutan's Lake Thorthormi by getting hundreds of people to haul away rocks each summer to create drainage channels. This approach is unsustainable. At QUT we are investigating the use of siphons to keep the water level in the glacial lakes a few metres below the top of the dam walls.

In September 2013 I went on an expedition to Jichu Drake Lake with my colleagues Les Dawes and Som Gurung, plus a local support crew.

We spent three days hiking 35km from our base at 3000m up to a glacial lake at an elevation of 4200m. All our equipment was transported on the backs of ten ponies. It was a really physically challenging experiment as working at this altitude meant our oxygen level was down to 60% of what it would have been at sea level. After several hours we managed to get a siphon going with a one inch diameter tube at a small lake actually within the moraine wall. The drop of our siphon was only about 3.3m and the length 75m.

We measured a flow of 0.38 litres per second. This doesn't sound like much but over several months this flow amounts to a substantial volume of water. A major advantage of using siphons in remote regions is that they run 24/7 and no external power supply is required. According to measurements I made using Google Earth the area of Jichu Drake Lake is roughly 50,000m2.

Over a six month period our siphon would drain about 6M litres, which would drop the level of the lake by 12cm. The flow through a siphon with a drop of more than 3.3m would have a greater flow. So, using siphons of only one inch dimeter could be used to prevent glacial lakes from overflowing. Extra siphons could be added in situations where the level needs to be reduced quickly to avoid catastrophe. The number of siphons could then be reduced to maintain the lake level below a certain height.

Our experiment was just a proof of concept to see if we could transport all the equipment necessary to get a siphon going at 4200m. At the end of the experiment we dismantled the siphon and just managed to get down to base camp before the darkness of night descended. A local Yak farmer helped us with the experiment and we thanked him with 100m of polypipe.

Next steps

The next step in the project is to perform experiments at Australia's Lake Manchester which used to be the main water supply of Brisbane. The plan is to measure the flow for one and two inch bore siphons measuring 500m in length. Some of the moraine walls of the glacial lakes in Bhutan are several hundred metres in depth and so the siphon tubes need to traverse this distance before descending the wall. We will use the information provided by these experiments to plan the next expedition to a glacial lake. This expedition will be in conjunction with a team from the Bhutan Department of Geology and Mines under the leadership of the chief glaciologist Karma Toeb.

The ultimate aim of this project is to train the local Yak farmers in Bhutan to deploy siphons in the summer months to reduce the risk of breaching of the moraine walls. Once ice in the lakes begins to thaw at the end of the winter, they could keep an eye on water levels and empty siphons as necessary. This would be an extremely cheap and effective way of preventing potential disasters.

We have some money to perform experiments at Lake Manchester but need some more funding for the next expedition to Bhutan. Siphon tubing is relatively cheap and so we hope that Bhutan as a country will have the resources to maintain a network of siphons to reduce the risk of catastrophic flooding downstream from the lakes.


Dr Stephen Hughes is a senior lecturer in Physics at the Department of Chemistry, Physics and Mechanical Engineering in Queensland University of Technology in Australia. Anyone interested in supporting the next expedition to Bhutan or to find out more information about the research should contact him via email at: [email protected]

Lake Bonney siphons Lake Bonney siphons
The team The team
Moraine At the moraine wall in Bhutan
Experiment Experiment in progress

Privacy Policy
We have updated our privacy policy. In the latest update it explains what cookies are and how we use them on our site. To learn more about cookies and their benefits, please view our privacy policy. Please be aware that parts of this site will not function correctly if you disable cookies. By continuing to use this site, you consent to our use of cookies in accordance with our privacy policy unless you have disabled them.