you have gotten some good advice so far. I wonder if the piezo is resonant at the frequency you want. If you have an o'scope you would know.
I got currious about this circuit - I've not seen multivibrators for high frequency (10K) applications before so put it on a simulator (switcherCAD). It does not oscillate with a 103 cap for the piezo. I am not sure what the problem is. I can get it to oscillate at a much lower frequency (50 hz or so) by fiddling with the resistors and caps. I'm not sure the simulator is handling it correctly as I get some rather odd results with the values from your circuit.
I agree that a 555 would be good way to drive the piezo. I KNOW that can be made to oscillate at 10kHz.
Phil