When Virginia’s top environmental regulator received an analysis warning of data center pollution, it took him less than five minutes to forward it to seven people on his staff.
The Piedmont Environmental Council had sent the report to Michael Rolband as a courtesy two days before publicizing it in March. It ignited a scramble within the agency to poke holes in the analysis, which raised concerns about the health effects of an AI facility that stands at the vanguard of the state’s fastest-growing industry.
“Will you review and tell me what you think?” Rolband, the director of the state Department of Environmental Quality, asked members of his air permitting staff.
Internal emails obtained by POLITICO through a public records request paint a portrait of an agency that moved quickly to defend the only data center in Virginia that is powering itself — in what former officials and environmental health advocates described as an unusual effort to shape the debate around an industry whose global epicenter is in the state.
The agency’s pushback focused on the report’s findings that the facility in Loudoun County could release harmful amounts of air pollution through eight bus-sized natural gas turbines — as allowed by permits granted by DEQ itself.
Behind the scenes, Rolband’s request was met by opposition from some DEQ staffers who said that assessing the report’s assertions on health effects fell outside of their environmental training.
“We aren’t qualified to do that review,” one staffer said in an email.
The request comes as data centers are devising new ways to produce their own power as a way to temper growing public discord over rising electricity prices amid the AI construction boom. Virginia has more than 600 data centers.
The report commissioned by Piedmont raised concerns that the data center’s on-site power system, which also includes dozens of backup diesel generators, could cause tens of millions of dollars in health damages to people living around the facility, owned by Vantage Data Centers.
In the emails, Rolband told staff that the Piedmont Environmental Council “wants to criticize DEQ” and that the agency “must take the time to do a defensible critique” of the group’s analysis.
That led DEQ to issue its own report a month later that echoed arguments Vantage had made to Rolband in emails obtained by POLITICO, asserting that the data center’s emissions were below what its state permit allowed. The state’s report also made assertions that health experts said are unfounded, including one that suggested people are safe from soot if it doesn’t surpass EPA thresholds.
Robert Burnley, who led DEQ between 2002 and 2006, said the agency’s report was “unusual” and amounts to “editorializing.”
“The DEQ is not the health department, it’s not their job to determine the health of a community,” he said. “If you’re trying to defend a permit, you just say these emissions don’t contribute to a violation of the air quality standards. That’s what I would say, and that’s all I would say.”
The emails show that Rolband shared his agency’s report with officials at Vantage and the Data Center Coalition, an industry trade group. The coalition then blasted it out to reporters and lawmakers as supposed evidence that the Piedmont analysis had been wrong.
“Unfortunately, misleading claims and inaccurate information are far too prevalent in the data center conversation,” Nicole Riley, a government affairs official for the Data Center Coalition, wrote to POLITICO when sharing the state’s report in April.
The DEQ report says it was crafted “in response to inquiries” about the Piedmont analysis. But internal emails show that the agency began organizing its rebuttal before the analysis was released publicly.
Rolband did not respond to requests for comment. The state agency did not answer two dozen detailed questions sent by POLITICO this spring about why it assembled the report, who requested it and the scientific justifications behind its assertions. Weeks later, after POLITICO sent an additional request for comment about the internal emails, DEQ responded with a brief statement.
“DEQ staff did not collaborate with PEC, Vantage or any other external participants in its assessment of PEC’s report,” wrote Irina Calos, DEQ’s communications manager, referring to Piedmont.
‘A defensible critique’
Vantage’s Loudoun County facility predates what has become a nationwide push for data centers to power themselves amid public backlash over rising electricity prices from the AI boom. President Donald Trump pitched the idea of building a wave of power plants for new data centers during a White House meeting with technology companies in early March. He expanded on those efforts Thursday.
By then, more self-powered data centers were already being planned in Virginia. The region’s biggest utility, Dominion Energy, receives about 10 new grid connection requests from “large load customers” every month, it revealed in a February regulatory filing.
The result is a staggering backlog of planned data centers waiting for power. Dozens of projects that altogether need 70 gigawatts of power — equivalent to 70 nuclear plants — have asked to be connected. That amount of power is three times as large as Dominion’s all-time peak load — ever.
That could have widespread repercussions. People living near the Vantage facility in Sterling say its natural gas turbines produce constant noise and air pollution.
The permits for Vantage issued by DEQ in 2023 allow cumulative emissions of seven different pollutants, including 95 tons per year of nitrogen oxides and more than 56 tons of soot. Both contribute to asthma and heart attacks. Those figures are in line with other state permits for minor sources of pollution, but DEQ’s permits for Vantage have sparked local concerns because of the facility’s location in a residential area.
The Piedmont Environmental Council set out to quantify the potential effects. It hired Michael Cork, a research fellow in biostatistics at Harvard University’s T.H. Chan School of Public Health. Cork used publicly available information about how much pollution the facility is permitted to release to estimate its health effects within the plume’s path of travel. Cork’s analysis calculated that soot pollution from the facility could cause between $53 million and $99 million in annual health costs if it released the maximum amount of pollution allowed by its permit. The analysis also found the pollution could lead to 3.4 to 6.5 premature deaths annually.
Piedmont published the analysis online at 8 a.m. on March 4. Less than three hours later, DEQ staff had already drafted a preliminary critique, stemming from Rolband’s request two days earlier.
Internal emails show that state staff originally critiqued the computer model Cork used in his analysis to map the potential path of pollution from the facility. They also raised questions about whether the report wrongly described the facility’s potential air pollution as dangerous, when soot levels in Loudoun County are deemed acceptable by EPA.
Health experts, including former EPA air quality official Michael Koerber, told POLITICO that soot pollution can negatively affect human health even at EPA-approved levels. The World Health Organization’s standard for ambient soot pollution is nearly half of what EPA suggests is safe.
When Roland received his staff’s initial comments, on March 5, he called it “a start.” Then he pushed for more robust analyses, warning that Piedmont had been in contact with his boss, the secretary of natural and historic resources, “and wants to criticize DEQ, etc.”
“Thus, we need to make the time to do a defensible critique that will stand the test of public and critical review,” he wrote.
Piedmont’s director of land use, Julie Bolthouse, said her group had contacted the secretary’s office to schedule a meeting about data center impacts broadly, not specifically to talk about Vantage or DEQ’s permits.
‘We aren’t qualified’
Some DEQ staffers thought it was ill-advised to criticize Piedmont’s analysis. A manager in the Office of Air Quality Assessments wrote in multiple emails to other staff members that the agency was not equipped to comment on the health impacts of pollution that it regulates.
The manager, Michael Kiss, said the “magnitude of impacts isn’t necessarily surprising and are relatively low.”
Health impacts, he added, are “not something we typically comment on in modeling group.”
“I am not sure we want to get into the weeds any further than that at this point,” Kiss wrote to three colleagues. “Just my two cents.”
Later, air permit coordinator Stanley Faggert said in an email to Kiss that Rolband wanted a “detailed response to pretty much the entire document” and had told him “to use whatever resources are needed.” He asked Kiss for “up to 2-3 pages” on why a different computer model “would be the better tool for this task.”
When Kiss responded with some technical information about the computer models, he concluded by saying, “I have absolutely no comments on the health study.”
“We aren’t qualified to do that review,” he added. “I don’t need to tell you that.”
An hour later, he sent another email: “I don’t think we need to make a big deal of this.”
“The health conclusions aren’t something that I can comment on since I am not a health scientist.”
Kiss and Faggert did not respond to requests for comment. DEQ did not answer questions about Kiss’ objections. It’s unclear if his concerns ever made it to Rolband, though the final DEQ report included critiques of Piedmont’s modeling.
‘Thanks for flagging’
Rolband had been in touch with Vantage executives about the Loudoun County facility before Piedmont’s report came out.
Phillip McAuley Jr., Vantage’s senior manager for public policy, sent Rolband an email on Feb. 18 thanking him for an earlier meeting and outlining steps the company had taken to reduce noise. He also described the facility’s gas turbines as “a temporary solution given the current power constraints of the grid.”
Rolband responded by saying he would include some of the information in a briefing he was preparing for state legislators.
A couple of weeks later, Rolband sent the Piedmont report to McAuley on the same email chain.
“FYI PEC did release a report on your facility yesterday on LinkedIn (and likely other places),” he wrote. “I asked our staff to review it. If you have any comments that you want to share publicly — please do so.”
“Thanks for flagging,” replied McAuley. He sent Rolband the public statement Vantage had already provided to reporters, including to POLITICO.
In the email, Vantage highlighted how the Piedmont report was based on the maximum amount of pollution allowed by its permit — something Vantage called the “worst case scenario”— rather than the facility’s actual emissions.
Asked about the emails, Vantage’s president for global marketing, Mark Freeman, said the company “maintains regular, routine engagement” with DEQ and other local officials.
“We value our ongoing relationships with these partners to ensure we remain a responsible, long-term neighbor in the communities in which we operate,” he said.
DEQ’s final report, released in late April, included some of the same points McAuley had made to the agency. The report calculated that Vantage’s “actual emissions” of various pollutants including soot, volatile organic compounds and ozone-forming nitrogen oxides amounted to 28 percent to 69 percent of what its permit allows.
The final report also includes seemingly irrelevant information. It cites state data from 2023, before Vantage was operating, to declare Loudoun County as “one of the best (if not the best) localities in Virginia for the health of its population.”
“If I were making the statement, I wouldn’t say that,” said Burnley, the former DEQ director. “That’s not the department’s job.”
The state report also included criticism by DEQ staff about the computer model Cork used in his calculations and claimed that the facility’s pollution is “unlikely” to cause health impacts because Loudoun County ambient soot levels are lower than EPA standards.
Cork, who conducted the analysis for Piedmont, said he stands by his findings. Even if Vantage had released less pollution than is allowed, there’s nothing stopping the facility from increasing its emissions under the permit.
He did not have access to the data center’s actual pollution levels before conducting his analysis, but that changed when DEQ disclosed that information in its own report. Cork ran those numbers through his computer models and found that annual health damages could amount to between $33 million and $63 million, including two to four premature deaths.
“Even at the observed operating level, the facility is associated with a meaningful population-level health burden,” he said.
Damian Pitt, who directs Virginia Commonwealth University’s Institute for Sustainable Energy and Environment, called Vantage’s emissions “fairly significant.” He was not involved in Cork’s or DEQ’s analyses but has conducted his own data center pollution research.
“It’s comparable to the emissions you get from small gas power plants that don’t have data centers attached,” Pitt said. “The difference is those power plants aren’t in neighborhoods.”
