I know you said you don't want to run outlets everywhere, but I think a pile of wall warts is exactly what you want. Jameco has a large selection. 9 V sounds pretty good and that is one of the standard voltages, but you might want to consider 5 V (see below). Get something in the 500 mA to 1 A range.
These supplies are inherently isolated from the line, are usually short-circuit protected (check to make sure, you definitely want that), and draw so little AC power that you can string a bunch of outlet strips together without harm.
In the end, I think this will cost less and provide a better experience for the kids. I remember when I was a kid tinkering with this stuff how frustrating it was to have batteries run down, especially when you're not aware of when you are asking a lot from them. You also feel a lot less guilty abusing a power supply than running down consumable batteries.
You can start a fire with almost anything. If you do just the right thing, even a 9 V 500 mA supply can catch something on fire, but no more so than a 9 V battery and without the chance of the battery itself doing something bad and causing chemical burns.
If you are worried about LEDs getting damaged by them getting hooked up backwards, maybe you should get 5 V or 6 V supplies. Most LEDs can handle 5 V in reverse, and some 6 V. With 5 V supplies, you can eventually run logic cicuits directly without having to use a linear regulator. Lots of stuff will run well from 5 V. If you get 5 V supplies, get at least 1 A capability. That will be useful for running small motors. 5 W total power really isn't all that much.
It is always case dependent. Not only with IEC 60079. For general, there are my seven "NOs":
1) no thermal chokes (enclosure/surface temp growth up/down too high)
2) no thermal differential (enclosure/surface temp inequality)
3) no electrical sparks (while connecting power, on, off, etc)
4) no vibration (with freq/mags sensible by the environment)
5) no electric/magnetic fields (with params sensible by the environment)
6) no focused irradiation (ir, uv, others sensible by the environment)
7) no other stimuli you find sensible by the environment
How to design such devices?
When you are facing safety problem, you need to analyze
- the environment you need to operate in,
- the scenario(s) how your device interacts (within) the environment
- the factor(s) and direction(s) your devices impacts the environment with
- the limits the impact is insensible (safe) for the environment
When you know this, next you harmonize your (initial) design step-by-step, trial-and-error (trial-and-error-and-error-and-error :-] depending on the experience your have/can access) to the state in which:
- its impact to the environment does not exceed the safety limits and/or
- its behavior does not predictably run out of safety enabled scenarios
My notice for you is the following:
1) read all the parts of the standard series, beginning from the first, to get better understanding of the safety regulation approach and details
2) read some books related to the standard (generally standards authors like to write related books because they give them money and additional points in the science world :-)
3) read information about similar devices, buy one or two and examine, but have a respect for third-party's intellectual property, of course.
Best Answer
They're probably not calculated because they include strong safety margins for maintaining intrinsic safety thus, the current won't produce an excessive over-temperature (for a given track width) that might ignite a contacting gas. It's highly likely (that along with spark ignition testing on various gases), these currents were deemed "infallible" in that a particular ignitable gas couldn't ignite it AND, this testing was quite likely to have been done in a small ignition chamber hence no calculations. I've seen the chambers used for spark testing and I suspect a similar method was used after all, why not.