“Do you think you could catch lightning in a bottle?”
That was how the meeting began. What does it even mean to “catch” lightning? Lightning is plasma. It is a transfer of charge. It is a transfer of energy. It is fleeting. It dissipates itself. I had no idea how to literally or physically catch lightning, but I sure wanted to see where this journey was going. Fortunately, one teammate said, “Yes, I think there are some things we can do.”
Our first concept was to capture an image of lightning to demonstrate that lightning had safely passed through the bottle. Lichtenberg figures are created when excess charge is loaded into a solid and rapidly released due to an external impulse, like a lightning bolt. The concept was to create a Lichtenberg figure in a bottle: lightning in the air; lightning in the bottle. Cool. But, capturing a lightning-like image is not the same as literally capturing lightning.
If we were to catch lightning, what would we catch? Catching some of the plasma would be very difficult, and it would extinguish itself quickly as it lost energy. Even if we could do it, it might not be very impressive to see. We could probably capture the charge from the electrical current in a capacitor. This is not trivial – the capacitor would have to be fast enough to receive energy from the lightning in much less than a second, and it would need to handle high voltages. Without proper protection, the voltage in a lightning bolt can go up to several million volts. But with today’s technology, this we can do! We have experience with metalized film capacitors, and they might work to capture the energy from a lightning bolt.
We now faced two new challenges: 1) It is the middle of winter, and there isn’t a lot of lightning. Where do we get lightning? 2) How do we prove to the world (on film) that we captured the lightning in the bottle?
We solved the lightning availability problem by turning to a lightning testing lab. Lightning testing labs evaluate aircraft components or scale models as well as wind turbine blades and other similar items. Depending on the testing needs, the test lab can provide either high voltage or high current pulses. Our new problem is that we needed both high voltage and high current to make the experiment work.
To prove we had captured lightning, we decided that we could start a car. What kind of car? How much energy would it take to start a car? We determined very quickly that supercapacitors would be best for starting the car; they can store a large amount of energy in a small volume. Indeed, we used the supercapacitors in the final experiment. How could we move the energy from the high-voltage capacitors that captured the lightning to the low-voltage supercapacitors used to start the car?
Our first several attempts to start a car with supercapacitors were failures. Looking online, we were able to find that a starter for a Honda Civic draws about 1.2kW. We figured that at worst, it might take three seconds to start a car, so the capacitor must hold about 3,600 Joules of energy between 14.5V and 11V. This number was important because it determined how much energy we must get from the lightning bolt in the first place. The less energy needed, the easier this would be.
From our failures, we learned some basic things, like the importance of optimizing everything to minimize electrical resistance. We also learned about equivalent series resistance of super capacitors and their charge/discharge cycles. While we could store the energy we needed in a single assembly, it would not deliver the required current. We needed an additional capacitor assembly in series. Nature was working against us; we would have to get more energy from the lightning to charge an additional capacitor. Eventually, not only were we able to start a Civic, but we were able to start a Toyota Camry!
We considered a flyback converter to move the energy from the high-voltage capacitor to the low-voltage capacitor. We would need relatively high energy efficiency. We setup a prototype in the lab. It was difficult obtaining efficiencies greater than 50%. Would we require multiple lightning flashes to capture enough energy? We were also limited by the number of high-voltage capacitors we could purchase in time for the filming.
Solutions began falling into place when we began setting up at the lightning test facility. First, was the car. It was a circa 1960 Fiat 500. It is as cute as a button, and more importantly it has a very small engine. The variables impacting the ease of starting the car are the size of the engine, the compression of the pistons, and the ignition system. The old, small engine of the car was much more important than not having a modern, computer-controlled ignition system. We could regularly start the car with less than 1,000 Joules of energy, and we needed only one supercapacitor assembly.
Our next break came in setting up the lightning. That’s right; we were doing this experiment in a lab. We weren’t dealing with actual lightning. The lightning is simulated by power supplies and spark gaps. The “lightning” must be set up. Working with the testing lab experts, we combined the high-voltage and high-current impulses. A consequence was that we were able to deliver the current directly to the supercapacitors, bypassing the high-voltage capacitors and the flyback converter. We significantly simplified the experiment and further reduced the amount of energy needed from the lightning. This was going to work!
The GE Beliefs describe the set of values that unite us all together and identify the core of how we work. One belief is to Learn and Adapt to Win. To catch lightning in a bottle, we knew where we were and where we wanted to be. At each step in our journey, we did not know what challenges we would face next. We identified our problem, we learned how to overcome it, and we repeated.
Ultimately, this was the result…
This experiment was one of three Unimpossible Missions that Global Research teams embarked on. Watch videos of the other experiments here, meet more members of the teams here, and read another behind-the-scenes account from the leader of the Talking to a Wall team here.