This Tree Map visualizes greenhouse gas (GHG) emissions by buildings on campus. It enables quick comparison and shows, for instance, that some individual laboratory buildings (colored gold) produce as much GHG emissions as all administrative buildings (colored dark blue).
On the road to climate neutrality
Researcher assesses a university’s carbon footprint using JMP® statistical discovery software from SAS
|Challenge||To assess, monitor and help reduce greenhouse gas emissions at a university as large and complex as a midsize city.|
|Solution||JMP visualizes the university’s greenhouse gas inventory, helps track energy use and helps identify opportunities for energy efficiency.|
|Results||The university demonstrates its commitment to become climate neutral by 2050 and saves tens of thousands of dollars in energy costs annually.|
When Chancellor James Moeser signed the American College and University Presidents Climate Commitment (ACUPCC) in 2006, he pledged that the University of North Carolina at Chapel Hill would become climate neutral – meaning no net greenhouse gas emissions – by 2050.
The first step of any improvement methodology is to find out where you currently stand, and for the university that means a comprehensive inventory of the greenhouse gases it was emitting – no small feat for an institution with the demands and dimensions of a vibrant midsize city. As of the beginning of the 2008-09 school year, the UNC-Chapel Hill community comprises 28,567 students, 3,450 faculty and 8,632 staff. The campus is sited on 729 acres with 17.5 million square feet of built environment and expansion in its immediate future.
Having just received his PhD from UNC in physiology, Daniel Arneman was recruited by the university’s Energy Services Department to lead the effort to achieve climate neutrality. Although moving from physiology to carbon footprint reduction may not seem logical, for Arneman it was. He’d become intrigued in graduate school with biomimicry, a field that looks to nature for inspiration for technological innovations. A great example of that, says Arneman, is Zimbabwe’s Eastgate Building, the design of which was inspired by termite mounds. By borrowing ideas from these natural structures, the Eastgate Building uses less than half the energy of a traditional office building and contains no air-conditioning equipment.
UNC hired Arneman to conduct the university’s inventory and then begin identifying, implementing and brokering the strategies to help reduce emissions.
The task attracted Arneman because it allowed him to “rediscover my passion for environmental stewardship and sustainability.” It also gave him a chance to creatively seek greater efficiencies through process improvements – an interest spurred by his study of biomimicry.
The university had an important task and the right person for the job. The next step was finding the right tool for the project.
Arneman soon found that tool: JMP statistical discovery software from SAS.
Comprehensively studying emissions data
Along with his desire to improve processes, Arneman wanted to more effectively visualize the data collected in his research – to display information so that it’s instructive and useful without many distractions.
“As you might imagine, I was working with a huge body of data,” Arneman says of the greenhouse gas inventory, “and I needed a way to manipulate it with ease to be able to ask the questions that I wanted to ask.”
He evaluated a few different products. “We had a campus license for JMP, so I started to play with it, and here I am with it eight months later,” Arneman says.
The ACUPCC pledge requires an inventory of the use of fossil fuels in the university’s facilities and the use of gasoline and diesel in its vehicle fleet. But that’s not all. It also requires tallying the impact of wastewater treatment, solid waste, refrigerant leaks from chillers and air conditioners, as well as commuter activity and airline travel.
“It really is a very comprehensive look at the campus’ emissions,” says Arneman, “and at quite a number of stages I used JMP to look at that data.”
Arneman has extensive records from the university’s cogeneration plant that track fuel use, the amount of electricity and steam produced, and various conditions within the plant.
“With JMP, I am able to look at five or six years of data all at once,” he says. “That is helpful in assessing how the plant was performing. We also have survey data to help us understand how commuters are getting to and from campus, and I am able to look at that data and, with JMP, very easily pull out the pieces that are useful for my work.”
Exploring data that pops visually
Arneman explains that he built a database in which to store his greenhouse gas data and that JMP seamlessly connects to it.
“So then when I was writing the report, if I wanted to show emissions by source or demand or scope or location, I could very easily just pull those numbers into JMP from that database, tabulate them and then have the data I needed.
“By looking at the data in a different way – not just at a timeline, but at a scatterplot of consumption by temperature, for example – more obvious patterns emerged. It was helpful for us to say that this building is consuming a lot more or a lot less than it normally should at that temperature.”
He noted that comparing numbers in a table can take a long time, but with JMP “the differences just visually pop out.”
Arneman regularly uses distributions, bivariate plots and most graphing functions in JMP.
The feature that initially attracted him to JMP, though, was the ability to click on a data point in one chart and have it highlighted in other relevant charts. So, for example, if he wants to look at a building’s energy use, he can simultaneously see how it used steam, chilled water and electricity in a given month and compare that to the same month in other years.
“The connectedness of JMP is really appealing to me,” he says.
Achieving major savings already
A major finding of UNC’s inventory was that almost 90 percent of greenhouse gas emissions on campus are related to the energy used in buildings.
“So our first focus has been on the buildings, understanding how they use energy so we can peel back the layers and find reductions and strategies for efficiencies,” Arneman says.
In fact, with a critical assist from JMP, Arneman and his team have already saved the university a lot of money.
Some months back, Arneman noticed that one building on campus had shown a dramatic improvement in its energy performance.
“The change would’ve been lost in the numbers on a spreadsheet,” he says, “but it was obvious in the visual display.”
He contacted the operators of the building and learned that they had done a retro-commissioning project that resulted in improved energy efficiency. But after about eight months of the improved performance, the building suddenly reverted to its old energy pattern.
“I needed some way to show this to the building operators, so I used JMP’s ability to turn Bubble Plots into an animated Flash file so they could see the performance changes. Within two days, they found a defective valve in the HVAC system and corrected the problem.”
By Arneman’s estimates, this change in one building will save the university about $30,000 a year. “Who knows how much we can save by analyzing the other 300 or so buildings on campus?” he asks.
Tailoring data visualization to different audiences
JMP has delivered on Arneman’s goal of visualizing data more clearly and concisely, both for himself and other audiences.
“It allows me to really explore the data, and if I’m curious about a certain pattern or question, I can ask that question of myself pretty quickly. Things just become more immediately obvious,” he says.
JMP helps Arneman communicate with others who need to understand the data: “When I develop a report, I ask myself what the person I’m sending it to is going to use the information for and how I can show it to them in a way that will be immediately obvious. There’s a lot of flexibility in JMP in how to display the data that allows me to tailor reports to their needs.”
Arneman says he’s always discovering new analyses in JMP. And each time he does, he also discovers a use for them. Just a few weeks ago, for example, he discovered the variability/gauge chart.
“It allowed me to make a plot that I’d just been dying to make, but didn’t know how to do it,” he says.
He had wanted to be able to look at the changes in the energy use of particular buildings and to quickly scan to see which ones were improving in efficiency and which ones weren’t.
“I found that plot and did a couple of test cases, and it really did describe the data in a way that I could very quickly look at and get three or four pieces of information specific to each building. And with 300 or so buildings, it’s nice to have the ability to do that quickly,” Arneman says.
Continuing to make data-driven decisions
The university is also looking into renewable energy sources. The Energy Services Department is helping to develop a project to capture and use methane from the local landfill. It also has brought together a group of engineers, consultants, planners and administrators to assess the feasibility of using such renewable sources as wind power and plasma gasification of solid waste in a comprehensive renewable energy plan.
UNC is one of the largest universities to sign the ACUPCC pledge, and it has a great deal of work to do to achieve climate neutrality.
But Arneman is optimistic: “We’re headed in the right direction, and this is certainly a priority of the administration. There’s been great support in both developing the inventory and also in finding solutions for reducing our carbon footprint.”
Arneman says the university’s objective is to make data-based decisions, and he’s confident that he and his colleagues have the data they need to move forward.
JMP has played a key role in interpreting that data, he says.
“Without JMP,” he says, “I think I would be wrestling with spreadsheets and repeating my efforts over and over. I’d probably be pulling my hair out.”
[JMP] allows me to really explore the data, and if I’m curious about a certain pattern or question I can ask that question of myself pretty quickly. Things just become more immediately obvious.
Daniel Arneman, PhD
Energy Services Department
University of North Carolina at Chapel Hill