https://www.nycurbanism.com/map-store/penn-station-map
At the turn of the 20th century, two large public buildings were built above the train lines in New York. Those were Penn Station and the James A Farley post office.
These two Beaux-Arts buildings sit adjacent to each other on 8th in Midtown West Both were designed by the firm McKim, Mead, and White
What started as an open-air station was capped by Penn Station and the Farley building. In 1960 Penn Station was demolished by the Pennsylvania Railroad Company to make way for Madison Square Garden.
Images courtesy New York Times
The demolition was a very controversial decision and became a turning point for how the US thought about architectural preservation.
Enter Senator Daniel Moynihan
Moynihan held a position at the Harvard MIT Joint Center for Urban Studies. and in the early 90s, he spearheaded a plan to adapt the Farley Post Office building into an extension for Penn Station because Moynihan was nostalgic for the old Penn Station and his time commuting through it.
SOM won the project with a design where the old mail sorting room used to be is now the new passenger sorting room above which are four large cable braced steel grid shells that use the existing steel trusses.
Losing proposals from a range of firms used a similar strategy, some including the existing trusses, and some replacing them. This led to the question:
What are the structural, spatial, and embodied carbon consequences of keeping the historic trusses?
For the study, a workflow of form-finding, structural analysis, and embodied carbon analyses were used.
Two cases were devised to compare against the existing design, the "Base Case".
Interestingly, Case 2 which uses normalized loads across the grid shell nodes for form-finding performed more efficiently with much lower maximum compressive force.
A possible explanation for this is the reduced reactions forces that are out of plane with the truss at its peak as can be seen in Case 1.
Conversely, a shell form found with normalized node forces produced much higher planarity across the shell (seen in green).
Altogether, Case 1 and 2 embody a similar amount of carbon equivalent. Case 3, which required additional structure, having removed the existing trusses, unsurprisingly embodies more CO2e.
I feel that it is only fair to include the embodied carbon from the discarded trusses in the embodied carbon calculation of Case 3 which means it embodies ~2 kilotons of CO2e more than Case 2.