Riding with Savonius

A university friend from the Caribbean first told me about Savonius turbines, at the time I saw a future in gas turbines and only a little interest.  This post is a belated thank-you note and an apology.  Savonius turbines are often used to power irrigation pumps.  The method of construction is beautifully simple, first a 40 gallon oil drum is cut in half lengthways with a gas torch, this is then welded back together with the two halves offset around a steel bar.  The steel bar forms the shaft which is drives a pump, if necessary some automotive shafts and gears are used to make the connection.  Googling for images of Savonius machines will show a wide range of variations on the theme and many degrees of constructional sophistication, but the basic design is simple and practical.  The sketch below shows a cross section:

The downside of the design is that the coefficient of performance is lower than more sophisticated designs such as the Darius and many horizontal axis machines.  However, in my mind there is a well developed relationship between performance and cost:

As a general rule, it is always worth investigating the low cost options before seeking out exotic solutions to a problem.  Apart from simplicity, two features of the Savonius design attracted me, first it is a vertical axis machine which can respond quickly to changes in wind direction.  At 10 metres or less, the wind in an urban environment often consists of gusts from varying directions.  Often a horizontal axis machine in an urban setting will "hunt" the wind and by the time the blades are aligned with the wind, the gust has subsided.  Secondly, and this remark should be treated with caution, the Savonius has some inherent protection against overspeeding as it's speed is proportional to the wind speed and not directly related to load.  The concept of mounting any form of wind turbine on a Victorian chimney stack is something no sane man can face with equanimity.  Safety is important and I'm uncertain about the wisdom of rotating machinery on the roof or in the garden.  Having said that, there seems to be an increasing number of rotating chimney cowls appearing which are in effect small wind turbines.

To explore the design I built a small model using two lengths of guttering to form the blades.

Sadly, to the amusement of my neighbors, builders working on adjacent houses and my son, I never managed to get the turbine to turn continuously in my own backyard.  However, within cycling distance, there is a an old Second World War pillbox close to the coast and near a weather station.  This is effectively an open air wind tunnel with some basic instrumentation.  The roof of the pillbox is clear of obstructions and at a height of approx. 10m higher than the surrounding land.  Over several visits, I found that the turbine would turn easily when the wind velocity was greater than 3 m/s and run reasonably smoothly up to 10 m/s in gusty conditions.  The "load" was finger pressure applied to the base of the rotor.

The pillbox is an "ideal" location and not representative of most accessible locations (e.g. my backyard), to see how it would behave in other locations, I spent a few mornings cycling around the town stopping off at the seafront, public parks and the roofs of multi-story car parks. Apart from on the seafront, periods of continuous rotation were rare.  It is worth noting, that this activity attracted little or no interest apart from that of a large alsatian dog but fortunately we became friends after a couple of visits.

Sticking one's finger into a rotating machine does not provide a useful measure of performance.  In an attempt to quantify the performance of the turbine, I attempted to build a small dynamometer, the fluid clutch and torsion bar of which are shown in the photo below:

Whilst this was a fun and absorbing activity, I came to the conclusion that utilising urban wind was a project beyond my resources and moved off in another direction.  At some point in the future, I may return and complete this device.  Small processors such as the Raspberry Pi and Arduino make projects like this feasible within a small budget.    The successor to the Savonius project is something called "Doris" (the goddess of wind) which will be the subject of a future post.