Low-Tech Wind Power for Small Farms
The Savonius Rotor
The idea for the savonius rotor began with our cheapness, not wanting to pay for the electricity it takes
our Austrian flour mill to run for the long stretches of time required to mill flour for the bakery. I was
surprised to find out that although milling is one of the oldest and most proven application for wind
power (along with pumping water), nobody does this anymore, at least not outside of farming museums.
Why not? Small scale decentralized grain growing and milling are making a comeback, why shouldn't
the same be true for wind-powered mills?
One of the few examples of a wind-powered grain mill I was available to find was the Savonius rotor
built by Job Ebenezer as part of the project Technology for the Poor in 1978, at the height of the oil
embargo. This project was carried out through the UIniversity of New Mexico. I subsequently made
contact with a retired turbine engineer living in Charlotte, Victor Gardy. Victor has had a long-term side
interest in the Savonius rotor.
A little bit about the Savonius. As a design, it doesn't get a lot of respect in the engineering community,
because it's an inefficient device. Wind devices are all pretty inefficient (a 100 % efficient wind machine
would capture 100% of the energy in a breeze, leaving dead-calm air downwind of it. This is
impossible.) The maximum efficiency theoretically possible is around 58%. Most devices are in the 20
% range. The Savonius is around 15%, and can't be improved much higher than that no matter what
materials are used or how they are configured into a design. This is because the Savonius is a vertical
axis wind turbine (or VAWT) and as such has inherent loss of efficiency that a horizontal
axis wind turbine would not have. This is because the wind is collected in cups, which then fall away from the wind, allowing the
next cup to be filled. The cups return with their backsides facing into the wind, which creates drag. The cup catching the wind is
enough to overcome the drag and allow the rotor to turn, but energy is still being wasted, no matter what you do. This is very
annoying to most professional engineers and leads most to write off the VAWT concept entirely.
So I grant you, the Savonius is not-so-efficient. So what's good about it? Well, lots of things. Let's assume we are going to build
our own rotor out of everyday materials, and that the resulting device will be somewhat inefficient. So what? You can
compensate for that inefficiency by building two, or by making it bigger, until you reach the production level you want. Let's
compare the Savonius with the typical wind power solutions available on the market today. I'm assuming that in both cases we are
aiming to power a home, small business, or a farm, and not a town, city ward, or a factory. The table below is very general, but it
should give you the general idea that the efficiency of the device is far from the whole story.
||A few thousand dollars
||Thirty or fourty thousand dollars
height needed to
|Low. Possibly even
|As high as possible. 100 feet or more.
Such a tower is also very costly and
troublesome to permit.
speed needed to
(11 mph or less)
|Ordinary materials available
|Highly-engineered materials only produced
at specialized facilities
needed to make
or repair it
|Level of Danger
working on it
(unit is slow moving and close to
So, you can see that despite
the inefficiency, the
home-built Savonius has some
good things going for it. If
you are a handy person with a
moderately windy spot who is
up for an alternative energy
investment of a few thousand
dollars but not for a second
mortgage, maybe the
Savonius is for you! Also if
you're not so excited about
climbing a 100 foot tower or
having to winch the whole
tower up and down, or paying
specialists to service your
windmill (the warranty of
which would be nullified if
you worked on it yourself).
At any rate, it seemed to me
that this is a serious option for
a small farm like ours.
At first I was interested in a Savonius that would transmit power directly
to a flour mill without being converted to electricity. Victor Gardy
convinced me that electricity is a sensible approach which would avoid
some of the pitfalls of mechanical transmission, which would only allow
milling to take place when the wind was available and within a limited
range of wind speeds. A net-metered or battery-bank system allows for
more flexibility and would allow the device to contribute towards all farm
energy usage, not just the flour mill.
As it seemed like we had a good approach that was innovative enough to
be of real interest to like-minded folks but not so innovative as to be a
total shot in the dark, we applied for a SARE (Sustainable Agriculture
Research and Education) Farmer Grant to build a prototype device and
subject it to extensive testing.
We received funding in 2010 and
are still in the midst of the
project. By late winter or early
spring 2011 we will have the
device collecting data. Once we
have collected three months of
voltage-production data we
will be ready to publish a complete report, including a construction manual, and to go on
tour presenting our findings. Exactly how well a low-cost low-height Savonius will
compare to commercially available wind and solar alternatives is yet to be seen. Once we
have the production data that should speak for itself.
However if I may hazard a guess (with the support of Victor Gardy, who is serving as a
consultant and advisor) I would estimate that our prototype will produce enough
electricity to offset the cost of its materials as well as a living wage for the builder in
about three or four years. Compare this to a 15-year payback for a typical commercial
option. One Bergey 10kw windmill we looked at had a payback period of about 23 years
at our current utility rates.
|Installation concept on top
of a modified (shortened)
farm silo. We're most
likely not using this
approach due to the
difficulty of modifying our
steel farm silos.
The Mark II
Now in progress is our Mark II design, building on the earlier phase of research. One of our conclusions is that
the wooden design, though very easy to execute, would have durability issues that might prevent it from lasting
the entire payback period. Since the first phase of the project came in under budget, SARE approved a second
prototype building off the results of the first project. This second one, rather than lumber, uses scrap 275
gallon fuel oil tanks and iron pipe. It can be put together for a material cost comparable to the wooden unit but
will last much longer. It also requires welding for assembly rather than carpentry, or course.
The tower is now much simplified, consisting of a timber support beam, two vertical posts (made of 2"
galvanized pipe) and top and bottom crossbraces with central bearings for the rotor shaft. The tower is
secured with guywires.
Working with Gund Institute engineer Sam Gorton, we are hoping to install this unit in our developing rice
system to power an archimedes screw. The windmill will have sufficient power to lift hundreds of thousands
of gallons of water to irrigate the rice. We will install and test this second prototype before concluding the
project in February 2012. The results and a revised design for both the rotor itself and the archimedes screw
will be added to our project report and construction manual.