原创 Unique issues for long-life batteries

The increased interest in long-life embedded applications such as remote data logging or utility meters brings up the eternal engineering question: how are we going to power these devices? These devices should operate on a small battery for 10, 20, or more years without any attention, often under difficult environmental conditions.

For shorter-life installations on the order of a few years, the battery decision analysis begins with some basic analysis of the current drain on the battery under various duty cycles and operational modes compared to the energy capacity of the battery (mA-hour). This can get fairly involved for applications with complicated operational cycles, but it's not too hard to at least put an upper-bound, worst-case number on requirements. However, when you need to run for a decade or more, the basic electrical analysis of load current and power versus capacity is only a small factor. Issues such as self-discharge, chemical deterioration and enclosure corrosion become major concerns.

Which is why I was intrigued when I read “Designing and Fabricating a Multiple-Decade Battery in Aerospace & Defense Technology. The article detailed a thermoelectric generator (TEG) based on radioactive decay that could run for 150 years, in theory. The architecture used a two-step process which I had not read about before, where the decay generates light, and then that light generates power via photovoltaic cells. The authors allude to the low efficiency, but unfortunately give no numbers, although I suspect it is in the <5% range.

 

Behind this deceptively simple symbol is a complex world of electrochemical and even radioactive activity.

 

TEGs powered by radioactive decay have been used successfully for decades, especially for deep-space vehicles where solar radiation is minimal. These TEGs use a single-stage conversion process based on heat of radioactive decay rather than the photons of the two-step process, with Seebeck-junction thermocouples to generate the electrical power from the decay's heat.

This approach powers the Voyager 1 and 2 craft launched in 1977, which are still traveling and sending back data even as they have crossed the vague border of our solar system into exospace. (For a fascinating book on these spacecraft, see "Voyager: seeking newer worlds in the third great age of discovery" by Stephen J. Pyne; and note that their extraordinary journey was made conceivable by the space-mechanics insight of a graduate student working on his own.) There's some also work being done to use thermocouples to capture waste heat from engines, but how practical they will be (cost, reliability, size, efficiency) is still unclear.

Of course, you can theoretically make a decay-based battery last as long as needed by using more of the core material. The question is how long the rest of the assembly and electronics will last under the application's operating conditions before it deteriorates and falls apart, unrelated to the radioactive-decay mechanism itself. But, hey, if the battery doesn’t meet last 100+ years, no one from today will around from today to criticize the work?

I also saw two other articles on long-life batteries, albeit on "only" a few decades. “Choosing the Right Batteries for High-Tech Batteries" from NASA Tech Briefs looked at the attributes of various chemistries, and especially the many interesting sub-varieties of the lithium-battery family. In a word: it’s complicated. When you need a few decades of use, even at very low current levels or low-rate pulsed duty cycles, there are many factors which come into the analysis such as self-discharge and temperature ratings. The mA-Hr capacity becomes only one of many parameters to consider.

While the author of this article is from a leading vendor of such batteries (Tadiran) and perhaps has some bias, I'd rather hear from someone who has real units out in the field and a track record, and who has dealt with subtle manufacturing and production issues, rather than just an academic expert. The same vendor also had a piece "Power Your Wireless Sensors for 40 Years" which had some overlap with the previous piece, but added new information as well.

Are you involved in decisions for long-life battery selection? How do you assess basic capacity needed with complex operational cycles? How do you decide on the long-life chemistry and form factor that will work?