What follows is a tale of two Dukes, one power plant and a bunch of hot air.

As in steam. Steam produced by a combined heat and power plant that would burn natural gas on the campus of Duke University, but be owned and operated by Duke Energy.

Duke the utility proposed the project for Duke the university (they share a common ancestor) in May as a way to install local generation while helping the host institution heat its classrooms, labs and hospital facilities. If the plant is approved, it could cut the university’s carbon emissions by reducing the need to burn natural gas in its campus steam plants.

Clean energy advocates on campus and in the region had other ideas. They questioned the accounting Duke used to assess the benefits of the project, and warned that the project would create a long-term addition of fossil fuel infrastructure on a campus committed to carbon neutrality by 2024.

Duke Energy, meanwhile, is hoping to gain experience in the up-and-coming CHP technology for development elsewhere in its territory. CHP works particularly well for college campuses, which have the electrical load and heating needs to justify such projects. Many of those campuses also hold strong public commitments to climate-change mitigation.

"We definitely see this as a growing business within Duke Energy," Duke Energy spokesperson Randy Wheeless said of CHP. "There’s nothing like going to one that's already in existence and operating, and we hope the Duke University one will be that one for us. [...] We want to make sure this one goes correctly."

This proposed 21-megawatt turbine, quite small by power plant standards, has shown the challenges of winning over public support for a new energy asset that's cleaner than traditional gas generation, but not absolutely clean.

More gas for less gas

The proposal developed by the university administration and Duke Energy promises several big benefits.

The gas plant will capture waste heat from electricity generation and pump it onto campus, making it a more efficient use of gas than traditional generators. That incidental heat source would allow the university to pull back on the natural gas it currently burns in two campus steam plants. A presentation on the topic says getting steam from the CHP project will cut campus natural-gas consumption by a whopping 50 percent, and reduce emissions counted under the Climate Action Plan by 18 percent.

This tackles a major source of greenhouse gas emissions on campus, as well as avoiding $2.5 million in investments that would otherwise be needed for hot water infrastructure. The university expects a two- to three-year payback.

Additionally, having an on-campus power generator serves the resilience of the university’s labs and hospital in the face of grid interruptions like hurricanes or thunderstorms. Diesel backup generators currently serve that role, but gas turbines are generally more reliable machines. Duke Energy would need to perform "minor relaying additions and programming" to enable this microgrid application, said Michael Schoenfeld, vice president for public affairs and government relations.

A clean alternative like a solar array cannot guarantee backup at night, and pairing energy storage with solar would cost a lot more for this service than the gas turbine will at this time.

The administration views resiliency improvement as the top priority in the project, Schoenfeld noted, followed by emissions reductions and cost savings.

These benefits come at the expense of a new facility built to burn natural gas, and that on its face conflicts with the institutional drive toward a smaller carbon footprint.

Drawing the fence

After the administration announced the plan in May, campus activism forced a pause for additional review. Duke Energy asked the North Carolina Utilities Commission to postpone a hearing on the project from Jan. 24 until May, campus newspaper The Chronicle reported.

Some of the concerns centered on the internal processes that led to the proposal, and whether or not the campus sustainability bodies were adequately included. The local response also posed a deeper energy question: Should an institution committed to reducing its climate impact embrace a fossil fuel technology, even a relatively clean one?

In this case, Duke aims to host a gas plant on campus and claim the carbon reductions it allows versus existing campus emissions. Since the new plant's emissions come from Duke Energy's business in generating electricity for the grid, the university would count them as part of the emissions factor for purchased electricity, not as emissions originating from campus.

A pair of professors from Duke’s Nicholas School of the Environment argued in The Chronicle that the university should account for the plant’s carbon emissions when assessing the benefits of the proposal.

“Under this alternative assumption, the net greenhouse gas emission benefits to Duke University are dramatically reduced, since the carbon intensity of electricity produced at the new CHP plant is considerably higher than the carbon intensity of the current Duke Energy electricity generation fleet,” they wrote in a letter to The Chronicle.

When factoring in those emissions, plus transmission losses and upstream leakages associated with the additional gas consumption, the professors calculated a net Climate Action Plan emissions reduction of less than 4 percent.

This debate comes down to a matter of where to draw the boundaries of responsibility. If Duke Energy built the plant nearby on land it acquired, and the university bought the steam and burned less fuel, there would be a stronger case for the 18 percent emissions reduction.

But that’s not the case here.

“This plant would not be built without Duke University support," said sophomore Claire Wang, who runs the Duke Climate Coalition. “We tout ourselves as a climate leader, and part of that is being responsible for the emissions that we cause.”

The administration has asked the Campus Sustainability Committee to evaluate this and other questions about the proposal, and to make a recommendation about how to proceed. They are open to a change in the carbon accounting, Schoenfeld said. The hope is to get an answer by May, which would guide the administration's approach to a board of trustees meeting that month. If all goes well for the plan, the utility could then move ahead at the PUC.

Who pays?

The questions facing utility regulators touch on a different set of concerns, specifically the equity of who pays for what.

As Duke Energy's first CHP project in the state, this would set a broader precedent that the regulated utility can leverage ratepayer funds for a grid-serving generator that also delivers a specific, localized benefit to a particular customer.

Ratepayers will get a deal from the Duke project that's as cost-competitive as more conventional, large-scale gas plants, Wheeless said. The utility would save on cost thanks to "a very attractive lease price" for the campus land, and the university's steam payments will go back to ratepayers.

Ratepayers, though, would have to pay $55 million for the ability to create that steam in the first place. The ratepayer dollars going to create a resource for a particular host customer is what sets this apart from typical plant construction.

There are some similar cases in the world of microgrids, with an important distinction in payment structure.

Arizona Public Service is building two microgrids, one at a military base and one at a new data center. The host customer contributes for the resilience benefits in the event of an outage. The microgrids will provide grid power and other services most of the time, so APS asked regulators for permission to rate-base the costs for those services.

A microgrid collaboration in Denver between utility Xcel Energy and Panasonic includes solar panels and a big battery on the latter's new facility there. Panasonic contributes some of the assets and gets backup power for its operations center; Xcel gets to rate-base its expenses for the assets that will improve grid reliability the rest of the time.

Duke, on the other hand, has been piloting a “utility-controlled, single-customer microgrid” model, and has been able to enter costs into the rate base when the microgrid serves the distribution grid, rather than splitting costs with a host client.

Now, this Duke project is not a microgrid, yet. That means the only benefit the school gets that the grid doesn't is the heat, which the school will pay for. Critics could still charge that if the utility wants to spend time on projects with a more localized benefit, it should extract more in return for the ratepayers at large.

A plant that only served one customer wouldn't be a good candidate for the rate base, Wheeless said, but this is different.

"In this case, it's connected to a substation that serves that customer but also serves other people in the community too," he said. Down the road, if the Dukes pursue the microgrid expansion, they could work out an appropriate cost-sharing agreement on that service.

The way the NC Utilities Commission rules on CHP cost sharing will shape the utility's options in paying for projects like this, which will affect its willingness to pursue them. One lesson from this experience is already clear: Rolling out new CHP infrastructure isn't a technical challenge so much as a social one.