Week of January 18, 2009

I don't usually focus a weekly report on one individual company, particularly not on a company that isn't making products on a commercial basis and therefore isn't having a real impact. The reason I'm making an exception here is that this particular company is implementing an idea I conceived on my own, though, as it happens, they got there first, at least in terms of devising a working prototype.

The company is Fellowes Research Group in Austin, Texas. They are possibly the world's oldest startup, dating back to the late seventies.

So what are they doing that's so special? First I'll explain my own concept which will serve as means of introducing the Fellowes approach.

Long time readers of this publication will recall that we've run a couple of pieces on thermoelectric and thermionic devices, transducers used to convert thermal energy directly into electrical energy. The trouble with all such devices to date, at least those subject to third party validation, is that they exhibit only single digit conversion efficiencies, a shortcoming which has severely restricted their acceptance in the marketplace.

It must be pointed out, however, that a less direct means exists for converting heat into electrical energy that is much more efficient, up to 50% efficient or more, depending on the amount of thermal energy present in the environment where the device is to operate. This is the venerable Stirling cycle engine, invented in 1809 and patented in 1816, and since subject to multiple largely unsatisfactory commercialization attempts.

The basic design of the Stirling, named after its inventor, one Robert Stirling, is quite simple. Stirlings are reciprocating piston engines with two cylinders and two pistons. The first piston is called a power piston, and the second a displacer piston.

Unlike a steam engine or internal combustion engine, the Stirling is a closed system. The working fluid, which is a gas, circulates between the two cylinders, but is not released through an exhaust valve. That tends to keep thermal energy within the system where it can perform useful work.

The Stirling is an external combustion engine, or, in some cases, it will run on solar energy or on industrial waste heat. In any case, the heat source is external to the cylinders.

At the beginning of the cycle, the power piston rests close to the hot end of the power cylinder which is exposed to the heat of combustion. The gas within the cylinder, generally helium in modern examples, expands as it is heated and causes the piston to rise to the cold end of the power cylinder. As the piston is rising it actuates a cam that causes the displacer piston to rise as well. The displacer cylinder communicates with the power cylinder, but does not communicate with the heat source, rather it functions as a heat sink radiating residual heat out into the external environment.

The displacer cylinder pulls in the heated gas from the power piston causing it to expand further and cooling it in the process. The pressure and temperature drops in both cylinders and the power piston is pushed down compressing the gas in front of it. The gas in front of the power piston is heated by the heat source and the cycle begins anew.

Stirlings are among the most efficient heat engines, and can achieve conversion efficiencies exceeding 50%, though the precise efficiency is a function of the temperature differential between hot end of the power cylinder and cold end of the displacer cylinder. A Stirling with a hot fire operating in subzero ambient temperatures can be very efficient indeed.

Used with a black solar collector without optical concentrators, a Stirling engine will better 20% efficiency on a hot day. That's about double what you'll get with commercial solar panels.

So one day I had a brainstorm. I was aware of attempts to make tiny internal combustion engines out of silicon with the same wafer fabrication techniques used to produce microchips. Why not take a large wafer and etch a bunch of miniature Stirlings into it?

I actually began investigating the economics of doing so and soliciting grant money for a development project, but eventually my determination flagged. I just didn't have the time, energy, or patience to pursue funding, and besides I thought it would be a hard idea to sell even though it had merit on the face of it.

The Fellowes team had roughly the same idea, but they took a more sophisticated approach.

There happens to be a variant of the Stirling known as the thermo-acoustic engine where instead of moving pistons one uses the heat to generate acoustic traveling waves within resonant enclosures tuned to precise wavelengths. The energy of the sound wave at resonance moves a compliant device called a passive radiator which interacts with an electrical circuit to produce power.

The almost total lack of moving parts makes the thermo-acoustic engine much easier to fabricate. Essentially, it is just a bunch of cavities. The Fellowes approach is simply to etch a multitude of engines and cavities within a single large wafer. It is a type of MEMS (microelectromechanical system) device.

The company claims that panels could be sold for $0.50 per watt as opposed five times that amount for photovoltaic panels. If what they're saying true, then you'd have the basis for a truly cost effective renewable energy generator for the home and for small businesses.

What makes this development especially interesting is that renewable energy from large generators such as concentrating solar installations and giant wind farms has come under increasing attack from some the same environmental advocates who initially supported its development, particularly in California. Their support is shifting toward small scale photovoltaic, but the latter is not particularly cost effective and moreover is a low voltage DC source that requires considerable power conditioning to be useful.

The Fellowes Research Group has been unable to secure funding to proceed beyond the prototype stage, and it may be that this left field approach will never the hearing it deserves. As I suspected with my own similar idea, it promises to be a hard sell.