This couldn't happen unless the lenses were powered and controlled. Here are the obstacles you'd run into:
Unless the lenses were actually monolithic crystal spheres, the lenses would rotate. You can't tide-lock an object on a lesser axis, so it would tend to rotate such that an edge was pointing at the planet. You could probably compensate for this by giving them counterweights that extended forward and back, but I'll leave that for you to do the math.
I'm presuming that you are thinking that the lenses would be on orbits that line up between the star and the planet on a schedule. For instance, if there were three of them, lens A and B would line up every third month, but lens C only aligned with them every ninth year or some such.
I suggest you look at the paths that the moon's shadow traverses over the Earth on a solar eclipse. As you can see, just getting ONE satellite to pass over a single point reliably would be nigh impossible.
Even if you manage to get everything lined up the way you want it to, natural satellites would mess things up for you by pulling the lenses out of their perfect orbits. Even without natural satellites, the other planets would mess things up for you. If there are no other planets, then the solar wind will push them around enough to make the alignment fail.
In conclusion, this would have to either be done through magic or technology advanced enough that it's basically magic. It could work if you had an active community maintaining engines that keep the lenses in the right place, but that makes them far less mysterious.
Also, if there are people on the space stations, you can accomplish the same thing with a single mirror set, or even with an orbital laser cannon.
How would it work?
Let's say you have an order of mystic astronauts that keep the lenses in the right place. At this point, you would have the "when single shines the triple sun" situation. The nearest one would be moving the fastest, so let's presume that it performs a final culmination of an otherwise wide beam. Let's also assume that all lenses are either convex transparent or concave mirrors.
For the big lens case:
The larger lenses would look like dark spots in the sky most of the time, getting brighter when they passed directly overhead. The furthest out one would have to be really huge to be worth the effort. When they were on the "full moon" side of the planet, they would either be invisible or they would have a tiny reflective spot that looked like a star moving slightly faster than itself. You would probably be able to see a ring of the outer edge most of the time.
For the mirror case:
For a pass-through reflection, you'd need a big outer ring that reflected light to a smaller inner circle. This would be a circular shadow most of the time.
Regardless of what they looked like the rest of the time, when aligned, the device would probably a line in the ground, tracking the nearest mirror's trajectory, instead of just hitting a single spot. Even if you had all of the orbits offset from each other, that would only make the line less intense and more wavy leading up to the final burn. If you absolutely NEEDED it to just hit one spot, you might consider having an opaque shutter object that kept light off of the nearest lens except for when it hit the critical point.