In some cases the units to be revamped, such as the chlorine and caustic plants, are the mainstay for the entire site production. The idea that we could just demolish the units would mean that all downstream production suffers. There are also other plants that are supplied. Any interruption in supply means they would suffer as well.
Why not build new? In this case, the chlorine and caustic plants cover a lot of ground and are also centrally located to their various "customer" plants. So building in another location would involve rebuilding a whole network of pipelines and electric lines making the "build new" option much more expensive. Another factor is the environmental and permitting problems a new site would involve. In our case, 100 foot deep piling would be required and the State would be unlikely to approve anything deeper than 60 foot. This means that we had to take advantage of the existing piles. These and similar issues make the revamp project unique because they are not normally associated with a new site.
The main reason to revamp is that there are generally some key elements that are in good condition or there is a technology improvement that will extend the life of the revamped equipment. Thus, with some effort there is an economic advantage to revamping over building new. The question is always: By how much will the life of the revamped plant be extended?
Another reason to revamp is that the plant to be revamped is already ideally located to its suppliers and users. To move the plant to a new location would mean higher infrastructure costs for longer pipe runs, new racks, and so on. Also if you do elect to build in a new location, you still have the problem of disposing of the old. Any replacement facility could still have the same infrastructure problem.
Obviously, the return on the revamp investment must be sufficient to cover the cost and the temporary loss of any production and be advantageous compared to the alternative of building new.