Comment on Z2026-05-01 Data Center Ordinance and Industrial Performance Standards

July 7, 2026 6:36 PM to planning

Dear Planning Commission Members,

I am writing regarding the July 8 public hearing for Z2026-05-01, the proposed zoning text amendment to add data centers as a conditional use in the Industrial zoning district.

I am unable to attend in person due to a prior commitment, but I ask that this email be entered into the public record.

My main position is simple: I do not want Warren County to open the door to data centers. I do not believe this use fits the long-term character, infrastructure, or public interest of Warren County.

That said, even if the County continues to review data center language, I strongly urge the Planning Commission to step back and begin a broader review of industrial compatibility standards. Many of the concerns residents have raised about data centers extend beyond data centers. Noise, vibration, generators, rooftop equipment, substations, lighting, construction impacts, water use, traffic, utility strain, and long-term compliance problems can come from many industrial uses.

One of my biggest concerns is noise. Warren County should not rely on promises that a use “will not be a problem.” If the County Code does not clearly regulate industrial noise, low-frequency hum, generator testing, rooftop mechanical equipment, substations, and continuous 24-hour operations, then residents may have little practical protection after a project is built. Any ordinance should include objective standards, independent testing, citizen complaint procedures, meaningful enforcement, and penalties that actually apply to industrial uses.

The draft data center ordinance includes a number of noise standards, but that only proves the larger point: if these concerns matter for data centers, they also matter for other industrial uses. Warren County should develop an Industrial Compatibility Performance Standards Ordinance that applies broadly, not just one industry at a time.

I am also concerned about any Payment-in-Lieu-of-Taxes arrangement. The applicant’s materials reference an estimated annual gross tax revenue of about $2.458 million and propose a PILOT to protect county revenue. If any PILOT is indefinite, extends too far into the future, or does not increase as property values, reassessments, public costs, inflation, infrastructure demands, and emergency service burdens increase, then residents could end up subsidizing the difference. Everyone else’s property gets reassessed. Any special tax structure for a data center should not leave ordinary homeowners and businesses footing the bill for what the data center does not cover.

The County should ask hard questions before moving forward:

• Will the PILOT increase over time?
• Will it adjust with reassessments, inflation, and actual public costs?
• Will it cover emergency services, road impacts, utility impacts, monitoring, inspections, and enforcement?
• Will it be limited in duration?
• Will the public be able to review the full terms before approval?
• Will the County have enforcement tools if the project creates impacts after construction?

My biggest fear with any SUP or CUP is the wording of the text, how a data center can still find a way through, and more especially how it can create the same problems seen elsewhere in the country.

1. The “small data center” definition is ambiguous.
The Industrial district already allows several by-right technology-, computer-, and data-related uses, including data processing activities, hardware and software, internet service providers, content developers, and telecommunications-related uses. It leaves a loophole for smaller or cleverly described facilities unless the ordinance clearly states whether small data centers under 10,000 sq. ft. are prohibited, accessory, by-right, or subject to the same review. A few data centers have started out with small footprints and then tried to expand, arguing they were already allowed by right when they were smaller.

2. Some standards use soft language.
The draft says data centers are subject to standards that address impacts, but also says that “whenever possible,” resource protections are encouraged. Later, the water-cooling provisions state that resource-efficient cooling is encouraged and that the systems should reflect a “good-faith effort.” Those phrases are not as enforceable as “shall,” “must,” “prohibited,” “limited to,” or “requires approval before.”3. Setbacks are too small for the impacts.
The draft’s setbacks include 125 feet from residential zoning, 150 feet from PND, 100 feet from A-1, 200 feet from the Town corporate limits, 25 feet from commercial/industrial, and only 25 feet between buildings in a common development. It also says buffers are inclusive of required setbacks, meaning the buffer may not add meaningful extra distance. For a use involving 24-hour cooling equipment, generators, substations, lighting, and large buildings, those distances are inadequate. 

4. Noise language is detailed, but weak enforcement.
The draft recognizes that data centers create continuous mechanical noise and includes dBA/dBC measurements, tonal analysis, generator testing windows, and post-development evaluation. That is good. But post-development noise evaluations are conducted by the campus operator. That raises obvious questions: Who independently verifies the testing? Are reports public? What happens after ten years? What happens after equipment ages, expands, or changes?

5. Noise and vibration limits can be relaxed.
The draft allows noise limits to be relaxed when no residential or noise-sensitive receptors are reasonably foreseeable within 500 feet, and vibration limits have similar relaxation language. That may become a loophole, especially if impacts extend beyond 500 feet, if future residential/recreational uses emerge, or if “reasonably foreseeable” is interpreted narrowly. 

6. Emergency operations are exempted from sound limits.
The draft exempts emergency operations, including a power outage. That may be understandable for public safety, but a data center outage could mean many generators running for extended periods. The ordinance should probably require logging, reporting, fuel/emission planning, and limits or mitigation after extended generator operation.

7. Generator testing is too thin.
The draft states that generator testing must occur between 7 a.m. and 7 p.m., unless DEQ requires a different time. But it does not appear to set a maximum number of testing days, cumulative testing hours, number of generators tested at once, public notice, fuel storage standards, emissions reporting, or complaint process. If a site has enough generators, you’ll hear testing non-stop 12 hours per day.

8. Water protections are incomplete.
The draft bars the permanent use of potable water for industrial cooling, but it allows temporary use of potable water through a Water Service Agreement and uses soft language encouraging efficient technologies. It does not clearly establish a total water-use cap, drought response requirements, public monitoring, groundwater protections, or a new public hearing if cooling technology changes later.

9. It does not appear to lock down fiscal protections.
The applicant materials discuss a PILOT and an estimated $2.458 million in annual revenue, but the zoning text amendment itself does not appear to require a PILOT, escalation, reassessment adjustment, duration limit, cost recovery, infrastructure reimbursement, emergency services funding, or public disclosure of long-term fiscal terms.

10. The “impact analysis requirements“ may not be strong enough.
The ordinance title says it establishes data center impact analysis requirements, but the language I found mostly says additional testing or documentation “may be required” if the application is ambiguous or lacks detail. That is discretionary. A stronger ordinance would require specific studies up front: noise, water, power, fiscal impact, traffic, emergency services, emissions, lighting, viewshed, decommissioning, and cumulative impacts.

11. There is no clear citizen complaint and enforcement process.
The draft includes technical standards, but I do not see a robust complaint procedure, an investigation deadline, graduated civil penalties, a corrective action process, permit modification, revocation, stop-work authority, or recovery of county enforcement costs. My ICPS recommendation directly fills that gap by calling for citizen complaint procedures, monitoring, independent verification, graduated civil penalties, corrective actions, repeated-violation procedures, and stop-work/permit remedies, where authorized.

Any industry can claim it will not cause problems. But if the County has no clear rules, no objective standards, no monitoring, and no meaningful penalties, then the County and its citizens may be left with the impacts while the applicant moves on. 

For these reasons, I ask the Planning Commission to recommend denial of Z2026-05-01. At minimum, I ask the Commission to recommend that staff begin work on a broader Industrial Compatibility Performance Standards Ordinance,before Warren County approves any new utility-intensive industrial use.

I’ve enclosed a request that I hope you will consider adding to the agenda in an upcoming meeting. If you are interested, I can develop it further. However, I believe I am at a point where I would need the county’s interest to protect itself from the issues citizens have been raising before continuing to add to it. My recommendation for ICPS is not intended to support a SUP, CUP, ban, or support for data centers. Please do not base your decision to recommend or deny the zoning solely on the hopes of having ICPS. It is a supplement to guard and protect the citizens of our county and to improve the county’s ability to hold industry accountable.

Thank you for your time and service.

Respectfully,
Lewis Moten
North River District, Front Royal, Virginia