Thermal Conversion: Stirling and Beyond


The conversion of energy from heat to free energy (usually torque or electricity) is one of the most fundamental, far reaching topics both in basic science and in global energy economics. The challenge itself is very old, but new opportunities have emerged for R&D to make radical improvements in the attainable efficiencies for the two main approaches to thermal conversion – heat engines, which convert temperature differences to torque, and thermoelectrics, electrical devices which convert it directly to electricity. If we can take advantage of these unmet fundamental opportunities, we can enable a whole series of fundamental transformations in the world energy system, of immediate benefit to energy security and sustainability.

            The most near term and urgent opportunity concerns small-scale heat engines, which can be used to produce torque (or, with minor extension, electricity). Down at the level of 200 kilowatts or below, the world mainly depends today on internal combustion engines, which really convert specific fuels (gasoline or diesel) to torque. From basic thermodynamics, it has long been known that “external combustion” or Stirling engines, which convert heat from any source to torque, would have many, many advantages over internal combustion.  The theoretical advantages would include fuel flexibility (i.e. using heat from any source), efficiency and ability to operate flexibly at varying levels of torque. Because of these theoretical advantages, many large efforts were mounted by many groups in past decades to develop workable Stirling engines; however, almost all of these efforts failed very badly, because of tricky challenges in design and in materials.

            This past year, two new events have occurred to reopen this door.

First, in Business Week (September 12, 2005) it was announced that a new company, Stirling Energy Systems (SES) signed a contract with Southern California Edison (SCE) to build a 500 megawatt solar farm in Southern California, using reflectors to focus light/heat onto a Stirling engine, to make electricity. SCE stated that the agreed price was “well under the 11.3 cents per kwh that we pay today for daytime peak power.” SES certainly had access to a working, affordable Stirling engine – a licensed version of the first generation engine developed under the uniquely creative Swedish inventor, Lennart Johansson, for use in Swedish submarines (where it is still in use today).  Even in this century, the creativity of individual human inventors is a crucial ingredient in obtaining real breakthroughs.

Second, the original inventor himself submitted (with a group at Kettering and Oak Ridge) have put together a proposal, to try to prove that it is possible to design a credible third-generation Stirling engine, doubling the efficiency (electricity output) without measurably increasing the cost, say, of a solar farm. The benefit of doubling the efficiency is that it cuts the cost of solar electricity in half, allows solar to beat natural gas in the market for peaking power, and revolutionizes the world energy market. Whatever the risks and uniqueness of this effort, the potential benefits to humanity and national security are a matter of life or death.

In an ideal world, mechanisms would be found not only to fund their technical proposal, but also to allow universities such as Kettering to compete for a new university-based center in external combustion technology designed to fill the critical hole here – the development of a new talent base of people creative and knowledgeable enough to follow up on the potential of the 40% 40 kilowatt engine and to push ahead to 55% in the future.

            If such a base can be developed, it would enable further research and transformations,  and new partnerships, such as: (1) combining the new engine with new reflector designs, new technologies combining optimal (ADP) control and power electronics to hook these systems up more efficiently and cheaply to power grids, aimed at developing a credible option for a 100 megawatt minimum-cost  solar farm demonstration – the basic R&D needed as a prerequisite before someone else actually pays for the demonstration; (2) developing variations for use in highway vehicles, where efficiency would be comparable to fuel cells but fuel flexibility would be automatic and production could be ramped up quickly in existing US engine factories; (3) developing new processes to convert CO2 or other gas streams in “Clean Coal” plants into methanol fuel, by using an input of solar energy; (4) developing cheaper systems for remote power and remote power use in poor remote areas of the earth, using local heat sources from biomass or other sources.

            For the minimum cost solar Stirling, the best guess is that costs of 4 cents per kwh or less can be achieved. The significance of this is that solar power could suddenly beat everything else on cost, for daytime power. It could be scaled up much faster than coal or nuclear power, because the engines and reflectors could be mass-produced in existing auto industry factpories, of which there is a surplus. For the vehicle option, the significance is higher efficiency (and thus long-term sustainability) than the presently “Holy Grail” of PEM fuel cells – without the huge costs, rapid depreciation and fuel inflexibility of the PEM. The importance of the other two potential uses should be fairly obvious at face value.

            Thermoelectric conversion is perhaps similar to what Stirling was 20 or 30 years ago – an unproven technology with enormous long-term potential, where risks and novelty dictate that we focus on niche markets, basic principles, and diverse high-risk approaches for now.