power rails for layout design

hi i am doing the layout for my circuit and i need to include the power rails. Vdd, Vss and gnd. Now when i saw on example the vdd and vss r just metal1 pin in the LSW ( a small box ) and in another example i saw they used metal1 drawing and made a large rectangle. I understand that if there are lot of connections to power rails a small pin won't be do much good so we have broad metal 1 rectangle. But suppose say i have just one connection to vss so should i create a metal or pin. Does the size of the metal in anyway affect the connection or design. Besides how to determine how much width of metal i am supposed to use. The power consumed by my design is 0.3uW. And is there a certain distance of how i can place Vdd,vss and gnd. So far i have just made it compact by trying to build my design as close as possible(following the design rules). Can someone pls help... Thank you!!!

With the supply and ground routing length, you need to keep in mind two things. The resistance of the routing which eats into the total room available for the main circuit and the routing inductance which impedes the high frequency current (Wider the metal lower the parasitic inductance as well as parasitic resistance). While the inductance effect depends on the operating frequency of the circuit, the resistive loss is definitely to be taken care of.

Oh thank you for the answer!!!! so it is good to have wider metal as it reduces parasitic resistance but the concern is it will take up some room that is allocated for the main circuit. am i right??? Also normally i have seen they use same metals (say for eg metal 1) for VDD and GND. but can we use metal 3 for vdd and metal 1 for gnd. I am not sure if it is wrong after all its just metal to just connect the circuit right... Will it affect the circuit performance in any way???

It is not the average power but peak power that counts then the impedance that represents compared to pin inductance (nH) and step rise time which represents an apparent frequency of interest usually up to 2/Tr.
For uW peak power , that is very high impedance.
For peak power in Watts then series inductance can be significant as well as ESR of the connection.

In large circuits the lower level power bussing is probably
overstrapped and the current dropped down from higher,
thicker, wider metal. In bottom-up design, if you don't
know the strategy you should at least understand the
limitations to what you -do- control, and assert the
needs that will address those (e.g. your reliability
current density limited distance between feed-points,
your signal integrity noise margin limited distance
(Ipeak*Rbus making a voltage perturbation locally
that must not take you past VILmax / VIHmin as
rated) and so on)). If you can't control it, at least
document that it's Somebody Else's Problem and
help them do their part.

Simple rule is "mobetta" but there are always other
interests.

The parasitic resistance of the routing reduces the total voltage room available for the circuit causing a I*Rpar drop. Yes, wider routing also affects the total area of the circuit. It is important to find an optimum point between the area taken by the supply, ground routings and the voltage headroom loss (head room loss calculated when the I is maximum).

You can use different metals for supply and ground routings as long as the other requirements are met.

Generally, lower metals are used for internal connections (tappings from source, drain and gates are easier if you are using lower metals).
Higher metals are used for supply connections (can create a mesh like structure allowing to keep the routings as wide as possible without affecting any other internal connections. Also, higher metals are thicker and thus have lower resistivity. An hence require lower width compared to the lower metals).