Microbial Fuel Cells

[molly_6]

I am participating in the Google National Science Fair and my project is related to microbial fuel cells. My project is to test the effects that a natural compost booster will have on the energy output of the fuel cell. I am also testing the efficiency of various soil samples from different locations. As part of the project, we are to interview a leader in the field of research we have chosen. i am conducing my research at this point and the beginning of the experiment is imminent. I am interested in any advice that an expert can give me that will help me to succeed and conduct accurate experiments. For example, I would love to know the characteristics you should be looking at for the soil to have a successful experiment and how many trials would be appropriate to preform. It would also be helpful to know what the average energy output is, or in general what the field of microbiology consists of.

[chrisshen]

I did a small experiment on microbial fuel cells when I was in 9th grade. As for how many trials, the more the merrier. But seriously, just look at your time constraints to determine how many trials you want to perform. In addition, you may make changes and modifications along the way as you learn more. Just remember, your project doesn't have to be fully completed by the time of your fair, but you should have sufficient data and a clear path for further research. I'm not sure how big your cell is, but in most cases output will be very small; much less than a watt. for reference, the max output achieved is around 1.5 watts/m^2 of electrode surface are, meaning you would need either really big electrodes or crunched up carbon nanotubes made out of graphene to obtain a large surface area. <-- this is really cool you should check it out

Look for soil that is very rich; the microorganisms feed on nutrients in the soil - the top layer of the soil called humus and is very rich in nutrients from the decay of leaves and dead stuff.

For accurate experiments and data, look for any variables and minimize them as much as possible i.e. same time of day, temperature. The goal is to have maximum control over the environment. In addition, make sure everything is sterile first before some sort on transfer of materials. Other than that, just be very perceptive at all times and carefully record your results.

Microbiology is a very broad field covering all tiny life forms including unicellular organisms, multicellular organisms, and viruses.
There are many subtopics under this field such as taxonomy, fungi, protists, immune system, parasites, viruses, algae, etc. The main application of this field is in medicine and pharmaceutical microbio.

Good luck! If you have any more questions, please ask!

[bradleyshanrock-solberg]

Statistical significance is way beyond the scope of middle school math, but you don't need all of the theory to grasp a few important concepts about trials. Let me give it a try.

Any time you test something, there will be some variability in results. If you bounce a ball, it might bounce a bit higher or lower each time, even if you try to do it from the same height, same surface below, same ball. You can't completely control your ability to hold it steady and release it consistently.

If you only test the bounce once, you don't know where in that range your results fall. You won't even know how big the range is.

If you change the ball, or the surface, or the height, you can't know if the difference you see is just part of your variability in "dropping procedure" or caused by the variable change.

For science fair projects, as a judge, I like to see a minimum of 3 trials per test (that means, test it exactly the same way 3 times, then change one variable and test that configuration 3 times, repeat for each variation you wish to test). With three trials you get a pretty good shot at knowing what the normal variance is, and whether or not your change was "significant".

If you change a variable and all of your next 3 results are about the same as the first 3 trials, the change probably did not matter. In my example above, ball color might not matter if the ball is otherwise identical (this seems obvious, but in grade K-3 kids are just learning that sort of thing) and you could be pretty confident ball color didn't matter with 3 trials of a blue ball, three trials of a red ball.

Now in real life, the more tiny a change you are trying to detect, the more important it is to increase the number of trials. So if you are pretty sure a variable matters, but your results seem similar, a few more trials with each test might show a difference, as you get more and more results clustered around "normal" for each scenario. So there is no perfect answer for "how many trials" except that "1 is probably not enough, 2 is barely adequate, 3 is usually enough given time constraints of a science fair project). The harder it is to actually measure your subject, the more trials you need (which can be a problem, as if it's difficult to measure, it's also probably taking you longer to measure)