Positive Energy

Monday, March 09, 2009

Entropy - The Key to Stability

Energy flows. It moves from one region to another. It bumps into constraints as it flows, and causes these constraints to change: perhaps to heat up, or perhaps to move. The specific effect on the constraint depends on how the energy flow interacts with it. The key point here is that these interactions are repeatable - they happen in the same way every time. This means that people can study the characteristics of these energy streams in a scientific manner. Using the knowledge gained from their observations they can then engineer various kinds of constraints so that the end effect of an energy flow produces something of value. For example, the flow of heat from a hot gas in one container to a cooler gas in another can result in a light bulb glowing brightly enough for someone to read a math book. That is a valuable result.

We use the word "entropy" to indicate how much a unit of energy has dissipated and therefore how reliably it will flow when we open the gate. If it has not spread out very much, ie if it is densely packed in a small volume, we say that the entropy of the energy is low. On the other hand, if the unit of energy occupies a large region and has flow characteristics that we cannot predict and control we say its entropy is high. The entropy of a unit of energy increases when that energy flows from a densely packed configuration (ie low entropy) out to a loosely packed situation (ie high entropy).

The phenomena of energy flowing from a dense state to a loose state has been studied over and over again using scientific methods. No exceptions have ever been noted - energy always flows spontaneously from low entropy to high entropy regions. You can depend on it.

We see then that a container of low entropy energy is a useful thing. It can be plugged into a machine, and energy can flow out from it in a predictable and manageable manner, creating benefits for human beings. Sources of low entropy energy are highly valuable. Over time humans have sought them out and learned how to work with them. From food to wood to coal to oil to gas - we have steadily moved to lower and lower entropy sources, and improved our living conditions by working with entropy to generate increased benefits. And now we are making the transition to a new low entropy source that is millions times better than gas, and so clean it is hard to believe - namely fissionable metals such as uranium and thorium. It will take a while, a few decades, for us to comprehend just how significant this giant leap forward really is. We have never seen anything like this before.

We build a useful energy supply system by putting a large amount of energy into a small space, and providing a gate that we can open or close easily to let out specific amounts of energy in accordance with current demand. Precision is vital if we want this system to be useful. Both the amount and the timing of energy releases are critical. Too much or too little energy in response to demand can damage equipment at the application end, while releasing too early or too late can disrupt schedules and cause chaos.

This sounds like a difficult problem. How can we possibly guarantee that the right amount of energy will flow at the right time? Well, thank goodness for entropy, that dependable characteristic of energy. Entropy increases spontaneously whenever energy is packed into a dense region and then given a chance to expand into a bigger region. So, every time we open the gate the energy flows in a predictable and controllable manner - as long as the entropy is low enough in the energy storage device. This makes the design of a good energy system reasonably simple - start with a really low entropy source and let the energy flow from there through various gates, shunts, loops, and meters to applications such as light bulbs that allow us to read math books. Clearly, the source with the lowest entropy and the longest life span will be our best choice for our energy system. Such a source reduces costs and improves reliability.

The fuels available for a low entropy energy supply system offer different degrees of low entropy combined with long duration. Longer duration leads to less work for refueling and therefore less cost. Wood is good - it releases heat at a steady rate for a few minutes before more fuel has to be added. Coal is better than wood - more heat and longer duration. Oil and gas are better than coal. Then we meet the millions of times better stuff - uranium and thorium. These elements provide clean, low entropy energy for eighteen month fuel cycles in modern reactors and decade long cycles in the reactors being tested for future use. This is millions of times better than coal and gas. Clearly, uranium and thorium provide our best alternatives for building better energy supply systems.

High entropy energy sources such as wind turbines and solar panels are less useful than low entropy energy sources such as nuclear reactors. We do not get as much valuable book reading light from them as we do from nuclear for all the work involved in their construction and operation. Wind is a high entropy source due to its turbulence. Sometimes the wind is strong and its energy density is high at a specific location, and sometimes it is weak and its energy density is low. This is why we have to build wind farms as expansive fields of turbines - we hope that at any one time at least some of the turbines will be experiencing energy dense wind gusts. The result is that the total energy from wind farms varies from nothing on calm days up to more energy than we need on windy days when all the turbines are catching wind. The changes in energy output are unpredictable and fast, making the total energy stream difficult to use. This is why we refer to it as a high entropy source. Solar panel arrays have similar deficiencies with their variability caused by night, clouds, and weather that covers the panels with snow and ice.

If the energy source is high in entropy (ie chaotic, in that sometimes it flows forcefully and sometimes it doesnt), then a lot of conditioning is needed to get the energy to flow to the application device in the smooth and controlled manner necessary for it to be useful. This conditioning consumes energy and requires operational management. If the management demand is challenging then the likelihood of mistakes increases. A tipping point point is reached when the chance of an error becomes significant and it happens with an unacceptable frequency, causing system failures. In our electricity grid such failures trip safety devices and cause a large scale shut down. This is why we can operate the grid with a few high entropy sources feeding into it - the tipping point has not been reached. When the amount of high entropy input increases its management complexity increases and the probability of a failure goes up. We eventually get into a situation where safety devices are being set off too often and the value of the grid has decreased. The tipping point seems to show up when ten to fifteen percent of the energy input comes from high entropy (ie variable in unpredictable ways) sources. In other words, a few wind turbines feeding into the grid doesnt do any harm, but too many cause it to fail too often.

To reduce the risk of system failures a high entropy energy stream can be modified to make it less disruptive before it reaches the grid. When the stream is low it can be supplemented with additional energy, making it a smoother source in total. The supplemental energy has to come from a very low entropy system since it has to vary quickly and forcefully as commanded by its managers to counter-match the unpredictable high entropy source. The range of this supplementary power extends from zero to full load when the high entropy source varies from full load generation down to nothing. This leads to an interesting situation - the use of highly variable sources such as wind and solar forces us to complement them with a low entropy generation system to back them up, and this low entropy system has to be robust enough to do the whole job when necessary. Note that the backup system has to run a lot. With wind turbines some additional energy is required seventy percent of the time. The backup system has to run all the time, however, so it is always ready to quickly jump in when needed. For example it may have to provide a five percent increase for a while, followed by a short interval where it provides the full load, followed by a period when it provides a fifteen percent boost. As you gaze at those turbines lazily turning in an expansive wind complex you should understand that somewhere a gas fueled generator is running twenty-four hours per day, releasing carbon dioxide, to back them up.

The practical observer asks why dont we just build the low entropy system and use it and forget about the high entropy complications caused by wind farms and solar arrays. Well, the high entropy system might be less polluting so using it makes sense when cleanliness is important. Wind and solar panels backed up by natural gas makes sense. But if we have a backup system that is clean, safe, and inexpensive and we use it to back up the high entropy inputs, this combination is not reasonable. Switching back and forth between the high and low entropy systems does not reduce pollution while it does increase the chance of grid failure. It makes more sense to just run the clean backup system all the time without the high entropy complications and have a more stable grid as a result. In other words, by adding nuclear reactors to the grid we eliminate requirements for wind and solar use. Nuclear reactors are just as clean as wind and solar sources, and have the advantage of being inexpensive low entropy sources that make the grid more stable. I think we should improve our electricity system by adding new nuclear reactors to it and forget about the unnecessary and costly exercise of plugging in wind turbines and solar panels until we discover the tipping point that makes grid instability intolerable.