Probably the most commonly recognized task that campus sustainability staff perform is the undertaking of a Greenhouse Gas Inventory. (BTW, when I say "commonly," figure maybe 35-40% recognition.) The term "inventory" is something of a misnomer. If any store inventoried its retail stock as haphazardly as Greenback U inventoried its greenhouse gas emissions, it would likely go out of business. And we did a more complete job than many, many other campuses.
Doing a campus GHG inventory boils down to identifying the major things that you know are happening on campus, and which result in GHG emissions. Then you look for ways to estimate how much of each of those things your school is doing. In general, the arithmetic to figure out how much GHG is emitted per unit activity is pretty simple -- the challenge is knowing how many units of which different activities your school actually engaged in. Some activities you know are going on, but you can't find a way to quantify. The emissions from those activities don't typically get inventoried.
You might think that a large, established enterprise like a university would have good administrative record-keeping in place. For activities ( e.g., buying electricity) which are direct and cost money, you'd almost certainly be right. For other activities (like local or long-distance travel) where the costs are incurred indirectly and/or in a decentralized manner, you'd probably be wrong.
So, the trick to doing a GHG inventory is to find ways to count certain basic types of activity across campus. That often means contacting administrative support staff in a wide range of campus departments -- administrative, academic and auxiliary. You have to explain what you're looking for (and why) before they can give you useful information. Once you tell them what you're up to, cooperation usually isn't a problem. But once the academics know what you're about, ...
Increasingly, grant-making organizations (some of the for-profit firms, and a lot of the not-for-profits) are writing into their requirements that some or all activities (research travel, conferences, etc.) must be carbon-offset. Folks writing grant proposals learn that you know how to estimate emissions and the costs involved in offsetting them, so you get a call or an email shortly before the proposal submission deadline. While doing the GHG inventory is a long-term project which is amenable to real planning, doing an estimate for the global warming impact of an academic conference is pretty much ad hoc. The estimating algorithms are basically the same, but the overall shape of the work task is entirely different. I used to get one or two requests for ad hoc estimates and prices, each semester. Lately, I'm getting that number each month. If things keep up at this rate, I may start getting one or two every week, which would be a significant drain on my time, especially since ...
One of the services I provide to departments across campus is to perform sustainability audits. I try not to use the term "audit" out loud, because nobody likes to get audited (I know I don't). So I come up with some euphemistic way to describe the task. In a nutshell, what I do is to review the administrative and other practices of departments, and make suggestions about how they can save energy, save money, and decrease their GHG emissions all at the same time. It's a "low-hanging-fruit identification" exercise, but it gets me in the door and (usually) allows me to establish some early credibility. There's no stick involved -- nobody has to follow the suggestions, there's no embarrassment or penalty if I'm ignored -- but if I can tie GHG reduction to cost savings, what's not to like? In truth, the sustainability audit service is a bit Machiavellian in its intent: down the road, we're going to have to look again at these same departments, and some of these same activities, and get more rigorous in our management expectations. But for now, everything's on a friendly basis -- no sticks, just carrots.
So there's three different applications of the same basic emissions estimating algorithms. Do a long-term inventory project, looking at major campus processes. Do some quick-and-dirty estimations, usually limited to a specific and localized (if somewhat hypothetical, at the proposal stage) scenario. Do a set of middle-sized projects, somewhat repetitive but each with its own unique features, looking at the continuing operations of a fairly small group of people on campus.
The expertise, I guess, comes from knowing how to crank the numbers. The work lies in establishing the network of contacts which will let you get the information (in whatever form it's available) to derive the numbers on which to turn the crank.