How far bench marking is used as successful in case of similar type of structures once cost analysis is carried out. what would be difficulties in tendering in this approach?

If I think what you are saying is how to compare the two buildings then we must use this comparison in a flexible way. Remember the whole  purpose is to try to estimate the likely cost of a new project by reference to an existing one. Using cost analysis we are examining previous projects and looking at the the elemental breakdown and then, maybe, reduce to a GFA cost. If the the new project is likely to have extensive and innovative piled foundation then clearly there will be no direct comparison so we need to subtract cost wise from the previous project and add in the new likely costs of the piled foundation for the new project. It isn't an exact science and may assumptions have to be made. It is only as the design develops that the cost becomes more accurate.

In the second instance of variations then understanding what the variation were for or more accurately why they occurred can help in estimating the new project. If on the existing project the designs were incomplete then that would be the reason for the variations. If on the other hand the variations were due to unforeseen ground conditions then an examination of the site investigation would be needed. Was it adequate? If not what not. If it was adequate then is the new project in a similar strata to the existing one? If so then you could expect similar 'unforeseen'  conditions.

You have to have a flexible mind set when doing these exercises. It can be difficult to move away from prescriptive analytical procedures but if by doing so you are able to be more accurate then this is what has to happen.

once cost analysis is completed, bench marking is carried out