How to perform Estimation with entirely new Scrum Team when we have backlog of say 20 Stories?

new scrum team may face problems as they are new to the project.

What problems do you anticipate, with this particular team?

@Ian - What problems do you anticipate, with this particular team?

Familiarity with the project and applications involved. 

As a SM how should I handle this situation so that estimation goes smoothly?

What do you consider a smooth estimation session? It sounds like there's bound to be a good bit of uncertainty in the estimates based on having little familiarity with the project and applications. 

My suggestion to you is to make sure that the team knows that their first few refinement sessions and sprint will be full of unknowns.  Encourage them to focus every discussion and decision based on what they know at the time and not what they are afraid that they don't know.  Explain to them that as they start working on the unknown, it will become clearer. Have them do their best to use past experience of working with unknowns to frame their behavior. You won't be able to do much to make the first refinement session great.  Just help them make it the best that they can do based on what they do know.

Another thing that you should help them with is that 20 items of complete unknowns is too many to complete in a single refinement activity.  Honoring that the Product Owner is doing their job of ordering the Product Backlog have them focus on the item at the top of the list and work down only after they feel that everything they know has been discussed and that they know the most they can based on the information available to them now.

new scrum team may face problems as they are new to the project.

May be they wont face. Isn't it better to start with your first refinement session and then see how team does it ? 

I echo many of the comments above, mainly Daniel's comment that you cannot estimate all of these in a single session. for your first session, plan to focus on 2 or 3 items. Prior to the session, get with your PO to identify which should be the primary focus. 

Make sure to let your team know that it's not an exact science. They will estimate something at a 2 and come to find that it should have been a 13; that is expected with a new team. Encourage and empower them to try. With a new team, estimating at an early stage like you've described is akin to throwing spaghetti against the wall to see what sticks.