As I recall, studies in this area indicate that approx 80% of project hard capital costs are determined in front end engineering. Hence, it is at this up-front stage where the greatest opportunity for improvements in capital costs lie (not to mention later operating costs, etc).
Review of process simulation software vendors web sites, or contacting them directly, may provide the information you are looking for. It is this type of information that demonstrates pay-back in use of their products (eg: saving on surge tank cost by improving control through simulation exercises to minimize tank size and cost, or, etc, etc). I would check with Hyprotech or AspenTech and the like... They may have info on particular case studies or examples that may help ("proper" vs "improper" front end engineering? or rather, justifications for higher resource level application at the front end! Do it right the first time).
As DaveBridger said, above, that's why front end engineering is done. ...looking forward to any examples our fellow tipsters may have!
PS: One example I experienced was in a petrochemical plant expansion. As I understood it, it was decided to "simply" double up on utilities equipment as there was "no time" for front end engineering in the brownfield expansion (typical of utilities, work starts last, when process design is complete, yet typically needs to be completed first). Approx 1MM$ was spent on equipment that wasn't needed if time was taken to integrate existing and new processes. Applying a higher level of resources (utilities process skills), for a short period, would have provided the "time" required to take advantage of process integration opportunities.