Sunday, March 29, 2009

Bench tests, destructive testing, and design failures

We had a fun-filled hour or two in the shop the other day bench testing the new wind turbine for the Eco-Cottage. This is a small, four bedroom, student dorm equipped with a stand-alone solar/wind system to charge batteries.

The system, intended as a demonstrator, is essentially what you would build if you had a small homestead, cottage, or camp, and were able to use some other form of power than electricity for cooking, refrigeration, and water pumping. These three needs constitute by far the largest drawers of electricity in most houses, and so if you, for example, cook and refrigerate with propane, and install a gravity fed water system that pressurizes with one pump and pumps with another 'trickle pump" to a header tank, then you can reduce the needed KWH by about two thirds. This doesn't save much energy or climate emissions, as you are still cooking and refrigerating with a fossil fuel, but it does save a lot of money in setting up an off-grid home, because solar photovoltaic panels remain very expensive, despite major recent improvements in production efficiency.

A household scale solar PV system capable of cooking, providing refrigeration, pumping water with a regular jet or submersible pump, and the rest of regular household electrical needs of about 600-700 KHW/month, will cost you over $30,000 in the state of Maine. We have one of these larger systems too, also intended as a demonstrator, in the shape of the Unity House. Indeed, the Unity House is so well-insulated and designed, the solar electrical system also heats the home when the passive solar heating system does not, on cold, cloudy days in winter. The house has a connection to the grid, which it uses as a kind of virtual storage battery, but over the course of the year it draws no net power from that grid. The power it borrows in winter, it gives back in summer.

Both houses, and both types of system, the $3,000 minimal battery-powered system, and the $30,000 grid-tied system, would be yet more effective with wind turbines also attached. The nature of household scale wind in Maine is that it tends to be available when solar power is not, on cold stormy days in winter, and so you can downsize a solar array considerably if you have a wind power system too. The economics of this are such that if you have a windy site, you might save $5,000 or $10,000 of photovoltaic resource by fitting $1,000 or $2,000 of small wind turbine.

So we built our own small turbine for the Eco-Cottage, which we ran off-and-on due to various mechanical failures, for about five years. It finally blew itself apart in a gale a few months ago, and I've been trying to find the time to replace it.

To replace the home-built device, we purchased a Southwest Windpower Air X, one of several small wind turbines on the market, although likely the most popular American-made brand. This device is regulated in high winds, up to 120 mph, by a combination of blade flutter and variable generator resistance. It can aslo be remotely shut down by throwing a "stop switch." None of these protections were built in to our homemade device.

The Air X has some standard bench tests detailed by the manufacturer, and recommended before installation. (It's a big pain to fly a wind turbine on a tall, difficult to raise tower, only to find it doesn't want to work!)

I used the bench test opportunity for a lesson on aerodynamics and electricity for some of our first and second year students. I also got a little help from our senior residents of the Eco-Cottage, who have had the same lessons from time to time with the older home-built turbine.

The remnants of the older turbine were also available for inspection, and from these we were able to determine forensically how it died. The blades and hub, which were purchased as a kit from an wind engineering parts store, showed damage, but this was secondary. The primary cause was the failure of the connection between the hub and the generator. This had loosened, and the mild steel hub had worn itself a bigger hole, also wearing out the connecting nut and slightly damaging the generator shaft, and eventually falling off, damaging the blades.

The usual correction for such a problem would be some kind of keyed or slotted connection between the hub and the generator shaft, and indeed the purpose-built Air X has a kind of locking system like this built into the hub. But the home built turbine used a special permanent magnet rotor fitted to the stator from a car alternator (actually a circa 1969 AC Delco model). Car alternators typically do not use keyed shafts.

So the Achilles heel of our Envirothon model turbine is in these off-the-shelf engineering parts, and some more positive connection between the shaft and hub is required for long life. A spot weld, some epoxy glue, split washers, or a machined slot and Woodruff key, any of these or a combination would likely suffice. None would do much to help the cut-in speed of the turbine though, which we found by direct measurement some years ago to be 15 mph, far too high for our site. This home-built device really only produces significant power above 20 mph, at which point it can make quite a bit. But 20 mph winds only occur around 3 per cent of the time on this particular site, on top of this particular tower.

While the generator itself, which we also tested, was undamaged. It seems that the permanent magnet alternators created by modifying this particular car alternator are quite rugged, and might be used in many helpful developing world applications.

I believe the Harris mini-hydroelectric turbine uses a similar concept.

But not for wind turbines. Not without some kind of expensive reduction gear, and probably much larger blades to overcome the increased resistance that would result. An engineering dead end.

However, using the standard anemometry models, the Power Law and the Weibull distribution, to produce this <3% estimate is also a good lesson. So is the forensic work to determine how the old experimental model failed.

I'd like to have better facilities and bigger, better equipment for these students to work with in this fashion, and am working on a scheme to get access to an industrial scale turbine in the locality. We already have pretty good demonstrators for most renewable technologies, scattered between the Unity Campus and the MOFGA campus, where my colleagues CJ, Verne, and John Mac allow us regular access with students and visitors. We have in total two wind installations, four solar PV, one solar air, and one passive solar house between us, all up and running and working well. The Unity House, at 5KW capacity, with all its design lessons built-in, may be one of the 5 or 10 largest PV installations in the state, and certainly is the best combined technology demonstrator in the state. Not to mention various local sites for household retrofit and weatherization. We can also look at the Beaver Ridge turbines from afar, and learn some very interesting lessons from the planning issues and local concerns they raised.

In the meantime we continue our efforts to grow the program. We're looking for bright, dedicated, engaged students interested in solving environmental problems by working with energy, energy policy, engineering, technology, economics, analysis, and management. Our current group of first years include some of the best students Unity College has ever attracted.

The last picture above is me hamming it up for the camera at the Admissions Open House this last Saturday. I displayed the Air X.

Funnily enough, I always spend more time answering questions from parents than from students.

Not that surprising. Parents read the news. 18 year-olds generally don't.

No comments: