Firstly, the '103' (= 10nF ) capacitors will be ok, but the pulses fed to the probe will have a frequency about 5 times lower than if C1 were 1.8 nF. You don't say how you checked the supply, but I'm assuming you have a multimeter. If I understand you correctly you have replaced the optocoupler with an LED. That is ok.
I have run a simulation of your circuit, and when the probe resistance is less than ~ 500k ohms the LED lights.
Check you have all the diodes the right way round and that the CD4093 pins are correctly connected. You can test your circuit by temporarily shorting the probe contacts to each other. The LED should light. If it doesn't, measure the voltages at pins 2 and 3 of the CD4093 and let us know what they are. Be aware that this chip can be damaged by static electricity.
Water can have a surprisingly high resistance, depending on its purity, so it's possible (though unlikely) that the immersed probe reistance is > 500k, which would prevent the circuit working.