> > Of course the second (which is what is actually usually done) the issue being that in real world after 50/60/70/80 years the bridge won't be replaced/repaired.
Assuming the bridge would actually be replaced at its actual end of life (25/50 years) and assuming a horrible return on spending (double money for only double the life), I'd say the 25 year option would be better in many (not all) cases.
From a cash flow perspective, it'd be helpful for a local government to avoid immediately spending on infrastructure costs that could be delayed for 25 years (when the lower cost bridge would be due to be rebuilt).
From the unknown future perspective, the city/town population could possibly change (bigger/smaller) 25 years into the future.
If the population grows, replacement bridge would need to be bigger and they could use the saved money to build a bigger one. Of course, if the population shrinks, they could spend less on a smaller replacement bridge.
Again, this assumes a poor return on increased spending. Doubling the costs to double the life. If you could double the costs and triple the life or increase the cost by only 50% to double the life, the equation favors going with the longer term bridge.