Jeremy in the belly of the wind turbine, with Nili supervising

Editor’s note:  This is the second blog in a series related to the testing of the wind turbine at the NASA-Ames wind tunnel in Mountain View, CA.  Check out the first blog in the series for some context!

Working squeezed/scrunched up in the belly of the Army Aeroflightdynamics Directorate 7’x10’ wind tunnel gives one lots of time to think about the world and our place within it, and how these “tiny” wind generators (the term “micro” has already been claimed for systems up to 5000 watts) can help contribute to an improved quality of life for some.  Remember the target market is people who either have zero access to electricity, or who perhaps depend on charging worn out car batteries in distant grid connected towns – you pay bus drivers to transport your battery back and forth – to get a trickle of power.  How people use that first few watt-hours of high quality energy they have access to fascinates me, since while we sometimes have an impression that everyone else wastes scarce resources too, in reality people with scarcity tend to know the value of conservation and wise use best – especially when their costs are high.

Extending their day by a few hours with an efficient light is usually the first use – most unconnected places seem to be close to the equator where the days are always short, and recurring costs for candle/kerosene lighting are cumbersome/prohibitive – allowing people to read, do homework, and maybe even earn extra income.  Charging batteries – for flashlights, the radios all campesinos carry to the fields, and cell phones – is another priority, hopefully reducing the number of discarded disposable ones that litter the ground.  Both of these applications require very little energy – for us it would be worth just pennies worth a day, but for people who all year around are used to calling 6 pm bedtime… priceless!  And yes, one of the first appliances to appear is the ubiquitous television, often for soap operas and soccer matches, but also news and education.

Charlie investigating Chinese instream picohydro generators (~30 watts) in a Vietnamese market

Just a hundred watt-hours a day will do all kinds of things when the appliances are efficient, and in a breezy location it shouldn’t take an expensive turbine to provide this.  As a slightly technical aside, it is best to remember that people use energy to do things while we have a tendency to express the output of wind generators (and photovoltaic panels, and microhydro installations, and nuclear power plants) in units of power (watts).  The wind tunnel tells us how many watts we might generate at a given wind speed, but winds fluctuate so we can’t count on getting that much all of the time.  Commercial turbines are almost invariably rated just in watts, and you always have to ask “At what wind speed?” – and you’ll quickly find that they choose to rate at some phenomenal (and usually unrealistic) value, like 25 miles/hour (~11 meters/sec).  Southwest Windpower has now started doing the right thing by helping you estimate how much energy (in watt-hours… each one of these helping to perform a useful task, such as a one watt LED lamp aiding a kid do homework for one hour) you might expect to generate from their products, after making some assumptions about your local wind speed distribution.

This brings us to the question “How do we extract power (and energy) from the wind – which comes originally from the sun?”  The maximum power available from the wind, per square meter of turbine swept area, can be easily calculated from the equation

Power = ½ rAV³

where r is the density of air, A is the swept area of the turbine, and we see that the power increases as the cube of the windspeed (doubling the windspeed gives 8 times as much power), so that while there is lots of power produced at high wind speeds there is almost none available at very low speeds.  Our Lenz blades sweep out an area of .75 m² (the Savonius configuration we tested is .45 m²) and we know that we can only realistically have a fraction of the energy the wind contains – Albert Benz said that 59% is the maximum, but more like 30-40% is typical for small tubines like ours.  So the amount of power you can tap into depends on how much the wind blows, and with like so many other things (like per capita income) the averages provided to us by the government don’t always do us enough good – some days it doesn’t blow, some days it blows too much, and luckily some days it blows just enough for your turbine to fill up your batteries for the

Weibull distribution (wind speed vs. probability) for an average wind speed of 6.6 m/s and a shape factor of 2.

coming week.  That’s the concept of the distribution (vs. and average), and luckily the wind speed variability tends to follow a Weibull distribution, a statistical function, where just two variables describe the distribution.  These are the average wind speed and a number related to the general amount of time with no or low winds (the shape parameter), and this site does a much better job of explaining it than I can here – and they allow you to type in your power vs. wind speed data (such as from wind tunnel testing), plug in a shape factor, and get the anticipated energy output (say in watt-hours/day) at your target location.  Now you can buy the right number of storage batteries to get you through the wind-less doldrums, and compare the cost of your tiny wind system with your other electricity alternatives – including continuing to charge your car battery for the equivalent of $3/kW-hr, and waiting a long time for the grid to arrive.

Maximum power vs. wind speed of the Lenz 2 turbine configuration

Taking the raw wind tunnel data Tyler showed (torque and power vs. RPM) we can determine the maximum amount of power a given blade set or configuration can extract from the wind at each speed and plot it – that upward curved shape is very important because it tells us that not much power is available to us at low wind speeds (say, <10 mph) and this is an unfortunate fact of life.  Our experimental method did not include a generator to turn the winds power into the electrical power we need to run appliances, and there will be losses in this conversion process – we expect it to be ~75% efficient – so we have to take this into account, giving us the ability to get about 25% of the energy embodied in the wind – not bad if the resource is free.

As mentioned, a single “power rating” for a turbine is not very useful (and only meaningful if the wind speed it was measured at is associated with it), but people are used to hearing just one number so we may need one.  Catapult Design will tend to rate these turbines (a set of blades plus the associated generator) at more realistic wind speed values, like 15 mph (7 m/sec), and then we’ll do our best to try and characterize the wind resource at a specific locale.  If we choose to rate at 15 mph, for example, then the real power output of the Lenz blades is ~30 watts, and the wind will need to blow at that particular speed for ~3.5 hours/day to provide 100 watt-hours of energy per day to a family or small business.  Blowing at half that speed for twice as many hours does not do us much good, since the blades of VAWTs often don’t start turning until 8 mph, and at 10 mph we might have to rate these tiny turbines at only a watt or 3.  For estimation purposes, Weibull wind speed distributions with very low shape parameter values would be an example where it blows very little, much of the time.

Its unfortunate that life is never as simple as it needs to be – it seems like that if a family wanted to consider buying a tiny turbine at X dollars, to decide whether it is worth it they need that power performance curve for it, decent information on their local wind conditions, and some idea how much electricity is worth to them (for example based on how much they are presently using and the cost for charging that car battery, or how much more they want to use – say if their neighbors pay them for charging cell phones).  Now if we just knew the probable lifetime and annual maintenance costs we could start to understand the cost of each future watt-hour… what an exercise, and don’t forget that investing in all forms of renewable energy is tantamount to buying at one time all the electricity you will use for the rest of your life, which is not an easy decision to make.

The end of September saw a major milestone completed on the wind turbine project – we finished our wind tunnel testing at the NASA-Ames research center! The experience was fantastic and so far we have been blown away by the quality and quantity of data we collected. While it is still a bit too soon to publish our complete test results here (we still have a lot of analysis to do), I nevertheless wanted to get started by explaining our test set-up and posting a few photos from the tunnel.

Test Set-up

Since the beginning of this project we have focused on two vertical-axis turbine designs: Savonius and Lenz (see photos below). The major purpose behind our wind tunnel testing was to characterize the performance of these turbines, allowing us to select the most promising design and move forward with developing the alternator that will generate the electricity. In order to characterize the turbines, we needed to collect data on how they perform in different wind speeds while under different loads.

The end result of this testing will be a series of power curves (see chart below) describing the mechanical energy generated by the turbine as a function of wind speed and turbine speed (rpm). We also explored three different angles of attack for the turbine blades on both turbine designs to see how they impacted turbine performance.

The theory behind our turbine test is this simple equation:

Power = Torque x Rotational Speed

When the wind blows it hits the turbine, causing it to spin. In spinning, the turbine converts the energy contained in the wind into mechanical power. In order for us to figure out just how much wind energy is converted into mechanical power we needed to place the turbine in a flow of wind with a known speed and then apply a known torque load to the turbine while measuring the effect of the load on the turbine’s rotational speed. Referring to the equation above, we would supply a known torque and measure the rotational speed, thereby allowing us to calculate power.

The way we mechanically applied the torque load was relatively straightforward. We coupled a DC motor to the bottom of the turbine shaft and tried to turn the motor opposite the direction of the turbine rotation (see image below). Since the torque produced by a DC motor is a function of current, we could apply a known load to the turbine by applying a known current to the motor.

This test set-up, while elegant, nevertheless proved problematic as the motor we sized proved unable to resist the amount of torque the turbine was generating and we quickly overheated the motor. Another way of applying torque to the turbine was needed, fast. Thankfully, the NASA-Ames facility is full of advanced testing labs, many of them working on rotating equipment. The day after we identified our problem we had a torque cell in-hand and connected to the turbine shaft (see image below).

This torque cell contained a strain gage mounted on a shaft that output a voltage based on how much stress was applied to the shaft. One end of the shaft was attached to the turbine and the other end had a mountain bike disc brake attached to it that could apply a drag load that would slow the turbine and provide the strain that the gage would measure. This set-up allowed us to apply our known torque load as desired and the testing went on as planned. The only downside was that we were unable to maintain a steady torque and were forced to take our readings dynamically as the torque applied and turbine speed varied. This scenario was less than ideal, but we still managed to collect a copious amount of data that should allow us to compensate for any dynamic and inertial effects.

Ultimately we collected all the data we wanted and, at first blush, the results look great! My next blog post will focus on that data and our resulting analysis. Many thanks to Malcolm Knapp, Jeremy Kimmel, Sarah Felix, and Charlie Sellers who all devoted many days to the wind tunnel testing. Other Engineers Without Borders volunteers that played an important role are Jerry Pugh, Matt McLean, and Ann Torres. Finally, none of this would have happened without the help of Jose Navarette, Nili Gold, and Farid (all of whom work at the NASA-Ames facility), the technicians running the tunnel, and the generous donation of the facility by the US Army Aeroflightdynamics Directorate (which leases the tunnel from NASA).

To see more images of our wind tunnel tests, check out our Wind Tunnel album and watch the video below:

In truth, I tried not to create too many expectations for this trip – because I figured most things would be out of my control and that I would have to roll with the punches.  But there were still a few unexpected things.

1.  First and foremost, I expected more cell phone ubiquity.  It’s not that they didn’t exist, but many people just didn’t have them.  At a minimum of $10, they’re still pricey for many rural communities.  500RWF ($.87) is the minimum to make a few calls.  If that doesn’t sound like a lot of money, know that the average rural Rwandan earns less than $1 a day and has an average family size of five.

2.  I‘d read the New York Times article that states hospital equipment donated to developing countries is rarely used.  We even found some collecting dust in the Ruli Hospital – an infant incubator, an operating table, and two ultrasound machines.  But the unexpectedness came from what was being used.  Like, for example, the equipment from World War II powering the lighting in the operating room.  Ironic that the latest and greatest technology is in storage while 60-yr-old technology is still being relied upon.

3.  I’m a vegetarian and anticipated a nutrition struggle.   But the Ruli parish we stayed had a talented cook who performed daily miracles with plantains, cassava, beans, eggs, mystery greens, and potatoes.   (My favorite dish was “spaghetti omelet” – eggs cooked with spaghetti noodles and onions.)  All of our meals were prepared with fresh ingredients – I’ve never had a more healthy and balanced diet in my life than during my 2.5-week stay in the Rwandan countryside.  To deepen the irony, the only time I fell ill during this trip was after eating at a popular, American-style restaurant in Kigali.

4.  The people surrounding me on a day-to-day basis had little to no income, lost parents in the genocide, or were infected with HIV.  They had everyone reason to be angry at the world for their situation, yet they were some of the most generous and forgiving people I’ve ever met.  Some examples:  On my last day in Ruli we visited a women’s basket weaving cooperative to pick up some gifts for friends back home.  The women surprised us with a feast and dances prepared in our honor, and gifted us each with a traditional basket full of fruit.   Foster, a 56-year-old army veteran who lives off of the generosity of the community, acted as my translator during trips to the hardware stores and shared stories of his life in the military and in exile.  Yvonna, a young orphan infected with HIV, corrected my Kenyarwanda translations and taught me new words.  Ngarambe arranged our meetings, sent text messages (that I only ever half understood) and phone calls, arranged transport to Kigali, and picked us up at the airport.

These are just a few of the wonderful, unexpected things that you can only experience from in-country travel.  More importantly, these are the things that quietly influence design decisions.  They shape your perspective and remind you of the human face behind technology endeavors.  Ultimately, it breathes new life into our work with The Ihangane Project.  Stay tuned on our future efforts with the people of Ruli, the health clinics, and The Ihangane Project.

Photos from the trip available on our Picasa.

And check out our YouTube channel for interviews with SELF, Manna Energy, and FrontlineSMS Medic.

When I think of East Africa I will think fondly of a woman in a green blazer. After 22 hours in the flight from San Francisco, I landed in Nairobi right in the middle of an airline strike. I expected a two-hour layover before jumping on my final flight to Kigali, Rwanda. Instead, I ended up sleeping on a cold cement floor surrounded by other stranded commuters without access to a functional bathroom, telephone, food, or water for 17 hours. Yet I had it easy – I met several other people who had been stranded for three days. Needless to say, the mob of angry, desperate people grew by the hour. The moment I had given up hope, two key things happened: I met another group with my plight and a Rwanda Air employee. This wonderful woman, in a green blazer, was the only employee trying to help people get on flights on other airlines. She somehow managed to get our group on a flight to Kigali within one hour of meeting her. I don’t remember taking off or landing in Kigali – I was asleep the second I latched my seatbelt.

Garambe, the nephew of my host, Dr. Jean de Dieux of the Ruli Hospital, greeted me at the airport. Dr Wendy Leonard of The Ihangane Project, Catapult’s client who had the smarts to avoid Nairobi, arrived shortly after. After a night’s rest in a Kigali parish, we departed for Ruli, a small community 40km west of Kigali. Despite the relatively short distance between all the health clinics surrounding Ruli, the underdeveloped dirt roads make for long and bumpy journeys. They remind me of rural reservation roads, making me slightly nostalgic for home. Outside my window immense numbers of banana trees cover the hillside, a lazy river curves through the valley, and small rusty brick homes housing large families pop up as far as the eye can see.

Ruli’s the kind of town where there are no paved roads and everyone knows everyone else’s business. Infrequent visits from outside means only four rooms are available for travelers at the parish. Everyone is up with the 6am sun and the daily tasks of water and wood collection begin. Kids, whose families can afford the fees, attend school in dusty blue uniforms and puffy neon green sandals. Women cultivate and harvest the hills. Men make bricks, do construction, or operate stores. And many more have no employment and spend the day walking the streets and socializing. At 5pm the dirt throughway is a flutter of activity – a mass of children run home from school, bicycles and wheelbarrows roll past with jerry cans full of water, bananas, or sugar cane. By 7pm the sun is down and the streets are quiet. The complete blackness is a good reminder of the complete lack of electricity. The air is saturated with the smell of cooking fire. And by 9pm, most have turned in for the night.

Wendy and I start every day at the Ruli Hospital in morning rounds. I shadow Wendy visiting patients in the HIV treatment ward, where the staff is well trained, attentive, and caring, but lack the tools and medication to treat some of the sickest patients. Cement floors, brick walls, flickering fluorescent lighting, crowded rooms, metal beds with chipped paint and squeaky springs, and a sweaty stew lingers in the air drawing flies to patients too weak to move.

On Wednesday, working through the hospital staff, we develop questionnaires designed to educate us on how community members spend their time, treat dirty water (or not), and grow food. The hospital staff conducts the interviews in the local language. We’ll take this knowledge to help guide design and technology decision-making further down the line.

The following day we travelled an hour west to the Nyange Health Clinic, which serves 20,000 people and has 18 patient rooms. Four solar panels power 24 CFL bulbs to keep the clinic operating at night. When there is enough power left over, they operate a laptop or desktop computer. The clinic has more electronics and lab equipment they would like to operate and we spend the day assessing the energy loads in order to expand the solar system.

Friday we drove two hours north to the Minazi Health Post, which presently has no electricity, to install a WE CARE Solar prototype. It’s a plug-and-play “solar suitcase” designed to provide immediate, small-scale power for a structure such as a small health clinic. Minazi’s health post was built entirely by community initiative when the Ruli Hospital proved too far to reach. The post is now supported by the Ruli Hospital system and thanks to WE CARE Solar, has enough electricity to power lights, cell phones, and charge batteries. Even this seemingly small effort will make a tremendous difference in their treatment of patients and communication efforts with the rest of the health clinic network.

With the remainder of my time in Rwanda, Wendy and I will travel back to Kigali to visit solar vendors, meet with several other NGOs focused on health/nutrition/solar in Rwanda, re-visit and assess the WE CARE Solar prototype at the Minazni Health Post, and continue a water assessment study in Ruli in the hopes of providing formula for infants of HIV mothers.

In early September I travel on to Kenya to visit Tevis Howard of Komaza. My only hesitation is that I have to fly back through Nairobi. I still suffer flashbacks of the horrible state and smell of the non-functional bathrooms serving hundreds of people. I’m assured that what I experienced is not the norm – but I’m still keeping my eyes peeled for the woman in the green blazer. Just in case.

Tomorrow morning, I board a KLM flight to Kigali to join The Ihangane Project’s Dr. Wendy Leonard for a three-week stay in Rwanda. Our first destination is Ruli, just north of Kigali and the base for the seven health clinics served by Ihanane. From Ruli, we travel on to Nyange for several days stay around the health clinic.

The clinic serves approximately 20,000 community members and presently has four panels on their roof – not nearly enough for their needs. Nearby, a new lab is under construction that will run off of a diesel generator until solar panels can be provided.  While in Nyange, Dr. Leonard and I will review the energy needs of the clinic with health clinicians, as well as explore options for a water treatment system targeting HIV mothers.

.

The trip also includes meeting with a few solar vendors in Kigali, meeting with the Rwandan staff member from SELF to get a sense of they handle remote monitoring of their solar installations, and doing a field test with a WE CARE Solar “solar suitcase.” I met Dr. Laura Stechal of WE CARE Solar a few weeks ago and was impressed with the solar suitcase, her passion project that emerged from her experience living in Nigeria. The Solar Suitcase is in prototype stage and I’m curious to see how the health technicians respond to the idea. The prototype they built for me has a 20W panel and 12V sealed lead acid battery for energy storage. Any 12V DC electronic can plug into the system – and the panel size can scale as high as 600W. I’ll be leaving the prototype behind when I leave Rwanda to see if it truly survives a walk away test.

.

A special thanks to the team at SELF (Solar Electric Light Fund) for their willingness to share their methodology and to Manna Energy in Kigali for their input on Rwandan water systems. We’ll be touring some of their sites while in-country as well as visiting with Gardens for Health International, Nicholas Evans of FrontlineSMS, and Tevis Howard of Komaza. More to come and more project info here!  

Since our last post the wind turbine team has been single-mindedly focused on our wind tunnel tests at the NASA-Ames research center. This testing opportunity will be a godsend for us as we will finally be able to get some clean, quality data concerning the performance of the two turbine styles we are evaluating: Savonius and Lenz. Good bye chasing wind around San Francisco!

What we are most excited about is the controllable nature of the wind tunnel and the ability we will have to set our test parameters. As a result of this unprecedented level of control, we will finally be able to apply a degree of rigor to our tests that has previously been unattainable. This more scientific approach will allow us to accurately predict not only the performance characteristics of the turbines, but also determine the appropriate alternator specifications for maximum power extraction from the system.

The tunnel in which we will perform the tests is a 7’x10’ wind tunnel owned by NASA, but leased to, and operated by, the US Army – specifically the Aeroflightdynamics Directorate (AFDD). We are extremely grateful to both the US Army and the folks at the test facility for their generosity and assistance. Without this unique opportunity, we would be hard pressed to continue making substantive progress with our turbine design.

Over the past handful of weeks Jeremy Kimmel, our wind turbine intern, and Malcolm Knapp, an EWB/Catapult volunteer, have been diligently working out the bugs in our set-up. Malcolm has spent countless hours programming the Data Acquisition System (DAQ) and they have both made a number of trips down to the test facility in order to ensure that we’ll be ready once the wind tunnel is ready for our appearance. Speaking of which, we’ve been told that we will finally get into the tunnel on August 17th! There have been a number of delays, but it is finally going to happen.

Below is a photo from our data collection training session just prior to shipping the turbine down to the NASA-Ames facility. We are fortunate enough to have an entire two weeks of dedicated wind tunnel time! During that period, Jeremy, Sarah, and Charlie will be leading our testing efforts and maintaining a constant presence at the tunnel. A number of other volunteers will cycle through to lend a hand and participate in this major testing milestone. Stay tuned for a post about our wind tunnel experience…

Left to right: Sarah Felix, Malcolm Knapp, Charlie Sellers, and Jeremy Kimmel

Watch the video on GOOD or on our Press page.

That said, it is still informative to thoroughly evaluate a technology to see how it fits into the panoply of options. This is exactly what we are attempting to do with our wind turbine and the testing we intend to conduct at the NASA-Ames research center. Our goal is to establish just how much mechanical power one can reasonably attempt to collect from the wind with a small, family-sized, vertical-axis wind turbine using the most promising blade designs we’ve encountered. Using the data we collect we should be able to estimate just how much it will cost per watt to simply capture the energy before turning it into electricity (this is based on the costs associated with building the turbine: blades, bearing, shaft, etc.). From there we will be able to predict just how much money we can afford to spend on the generator and circuitry and still achieve the price point our partner, AIDG, will help us establish (is is truly $1/watt or a more achievable $10/watt?). The amount of money available for the generator and circuitry based upon this price point will directly inform just how feasible this product may be for rural Guatemalan villages.

It could well be that the raw materials alone for the alternator simply price the whole concept above our acceptable price target; it is, after all, notoriously difficult to scale wind generation down and keep it cost-effective. As a rule of thumb, the larger a turbine is the less the electricity it generates costs per watt. We hope our tests prove that small, vertical-axis turbines are a promising avenue for additional design research (we’re fans of wind energy, after all), but the numbers won’t lie and should allow us to provide AIDG with a realistic evaluation as to how good an investment their continued development appears to be.

In addition to making certain the turbine is mechanically ready for the wind tunnel we have also been busy jumping through the hoops that NASA and the Army (which manages the the 7′x10′ wind tunnel we’ll be using) sets up for all prospective wind tunnel users. One of those hoops is a “failure analysis” of the turbine and its component parts to ensure that the turbine won’t fail during testing and harm the tunnel or its operators.

At the suggestion of the extremely helpful tunnel technicians we’ve been working with we decided to forgo the analytical approach to failure analysis and instead go with real-world “proof” testing. This involves showing the turbine can withstand substantially larger loads than it is expected to see in the tunnel by calculating the maximum load it is should see in the tunnel, multiplying that of a factor (say four), and then subjecting the turbine to that load outside the tunnel. If it survives it should pose no threat to the tunnel at the loads it will see.

In this photo we are applying a 40 lb load horizontal to the middle of the turbine to simulate approximately four times the 11 lb force the turbine is expected to see if we test it to our proposed maximum speed of 25mph. I think it’s going to be ok.