Doing a greenhouse gas inventory for a college or university involves crunching a lot of numbers. Activities which generate emissions, factors for calculating the emissions generated, forcings to translate other gases into CO2 based on global warming effect, lots of stuff. The tool Greenback used to prepare its baseline inventory was the defacto standard of such things -- Clear Air - Cool Planet 's Campus Carbon Calculator. A large majority of ACUPCC  signatories used exactly the same toolkit.
But the version we used for our inventory was version 5. There's a new toolkit in town -- version 6. While v5 was strictly a spreadsheet for inventorying the past, v6 includes functionality to project the future -- both likely trends if your school continues with business as usual, and any number of "what if" scenarios based on combinations of emission reduction projects you might decide to undertake.
Actually, the way the thing is set up, if you want to really compare a number of different scenarios (combinations of projects), you have to embody each scenario in its own copy of the spreadsheet, but how hard is that? And each scenario comes complete with both financial calculations (IRR, NPV, discounted payback period) and graphics to show the effective emission reduction over time, coded to the particular project which produced the result (what's often called a "wedges chart").
The graphics are a little ugly (my opinion), but the thing's an Excel spreadsheet -- what do you want? For presentation purposes, you'll probably want to clean them up some. But most of the heavy lifting, as in the previous version -- so September!! -- is done for you. And given that most of my work is indoors already, avoiding heavy lifting rates high on my list.
BTW, if your school is still completing its inventory, there's no rush to convert. On the other hand, there's no need to hold back converting, either. Version 6 will import whatever data you've already entered in V5. So long as you played by the rules in the older model, the import function seems to work just fine.