Please be aware that the speed of sound also depends on the humidity and inside a well this is probably higher than at the top. It might even vary with depth.
My SRF05 library is not suited to solve your problem but has information about the effect of humidity and might help you to tackle this effect.
How about "a long rope with knots", throwing it in the well and count the knots?
These days throwing can be replaced by a rope between two gears to measure the length of the rope. When it touches water tension changes, or maybe better use an electric cable. As long as it doesn't touch water infinity Ohm, touching water means a drop in resistance.
In my home town we have a water meter for ground water and it is a transparant tube about 10" (25 cm) in diameter in which a styrofoam cylinder is placed. This styrofoam floats on the water (defined depth) and if the water rises the foam rises. Its height above the ground is almost 7 foot (2 meters) its accuracy is better than 2 inch (5 cm). See picture
