Hi Bravo,
What do you mean by fixed L & W. If they are fixed then how will one design ??
Meaning of 180nm technology is that the minimum possible length that you can use is 180nm. But you are free to use higher values (there is also a high L range but that is very large generally, may be 10um). So only the min L is fixed. So is the min W.
Again in an given technology there may be 1.2/1.5V devices, 1.8V devices, 2.5V device or 3 or 5V even. The min length for them can be different i.e. min length of 1.2V device < min length of 3V device.
Hope this will help ..
I dont have TSMC180 info with me but I will explain with TSMC40
In this technology there are devices that can be used only up to 1.2V supply, there are a few that can be used up to 1.8V, few up to 2.5V and few up to 3.3V supply.
This means in this technology you can go for high voltage designs. Say you need to design an amplifier in 3V supply and you only had 1.2V MOSFETs then there will be reliability issues (discussed later) with the devices.
So you have have to use 3.3V MOSFETs for such design.
Now 40nm means you can go as low as 40nm in length. But with that small dimensions in higher voltages there can be reliability issues. So from TSMC it is made fixed that 40nm can only be used as min length in 1.2V MOSFETs, while 250nm in 1.8V MOSFETs, 300 or 360nm in 3.6V MOSFETs (this I don't remember). So you can see that Min L in 1.2V < Min L in 1.8 < Min L in 3.3V devices. But you are free to use 1um or 5um L in any of these devices.
Reliability: In general it is a huge topic to discuss. In short you must have seen max voltage & current specifications on capacitors and transistors when you go and buy them from the market. They are specified because their performance is guaranteed within that range. Above it they may burn out / fail to operate / or give wrong results. Same is for the MOSFETs.
Hope this will help ....