My school's science fair just required the teacher to sign off on it that it wasn't dangerous. My experiment involved running electricity through water (to turn rust back into metal), and no one cared.
After the informational meeting, I told my kid, point blank, "No people. No animals. Plants. How about a nice experiment involving plants?"
All that without paperwork, and I'm not even very old -- I'm only in my early 20s.
Is this primarily the case in the US, or did I just manage to avoid it during my school days? I performed experiments that were probably no safer than those described here[0], and the only hurdle I had to pass was satisfying my teacher that it was safe enough (which boiled down to an informal a) you've thought about the risks, and b) aren't reckless; no actual paperwork)
[0] e.g. giving 16/17 year old unsupervised access to a 5kV power supply. The current used wasn't dangerous (pretty unpleasant to be shocked by though!), but it would destroy equipment easily (we went through 10+ multimetres, even though they were allegedly rated for that voltage) and the spark would set fire things pretty easily (paper, leaves, etc.)
In fact, it has given rise to a cottage industry of science fair mentors-for-hire. If you can find a highly enough motivated high school teacher to manage the paperwork, and can find a lab that has the correct rating for whatever you are doing, then great for you. Otherwise, there are a few places that can provide lab space and a mentor at an hourly fee, not to mention experience in getting the mountain of paperwork correct. It is good to have money to rub on the problem -- I'm sure a lot of kids get left behind because of that.
My daughter did an electronics project -- and the paperwork was still daunting, even though we could check "N/A" to all the biohazzard management sections, lab-has-eyewash, mentor qualified to handle animal tissues, etc. We still had to submit schematics and test plan material -- which is OK I guess because it is probably worthwhile to see if the project involves a direct connection to 220V AC mains before things get out of hand.
In the end though, one of the biggest problems with science fairs is getting enough qualified judges. The kids pour countless hours into the project, and hope for a 5-10 minute visit from 4 judges. If you are lucky, the judge might actually have a clue about your area of science or technology. Then the judging team tries to figure out some winners. One of the best ways anyone here on HN could help is to volunteer an afternoon per year to be a judge. Fairs are always short of qualified judges. Heck, they are short of judges that can fog a mirror. The best thing you can do for these kids is give them your time to encourage them and hopefully nudge the system towards a rational result.