Why Does My [-24V to 24V] to [0 to 1.8V] Converter Need Vee = -Vdd in Simulation?

Hi
I use the circuit in the picture below to convert a DC signal from [-24V=>24V] range to [0=>1.8V] range
I have two questions:
1)When I simulated it in ISIS I found that it gives me the right results only if Vee=-Vdd that is mean if I connect Vee to GND it gives me wrong results I can't understand why can anyone give me an explanation?
2) I need to use an adjustable resistance (timmer resistors) to make the output in the right range after realisation and it is generally the one which has the lowest value which means 1K but in this case I have 4 resistances unless of 2 so did I need to replace both of 1k resistances with adjustable ones or replacing just one is sufficient and if it is just one what is better replacing R8 or R10 and why?
Thank you. 

You have -ve volts on the input of the op amp(worst case at -24V), to keep the input stages biased, they need a negative supply. There are a few op amps that will operate a few hundred mV below ground with a grounded Vee.
The same problem occurs at the output, op-amps will not go to ground with a grounded supply, a rail to rail output op-amp will get close, but only at very low output currents.
So in both instances, you generally need to have the negative supply.
Note that SPICE will not pick these problems up unless the op amp used has a model designed to handle this issue
Re the resistors, you have both offset and "gain" to contend with. Offset to get the output to 0V, and gain to get the correct span of 0-1.8V, so you will need two variable resistors. In choosing pots and location, always use a value that results in the smallest span that is needed, and wire it to minimize change if the wiper goes open circuit,
cheers,
Richard