When I wrote recently about Project Spero here in Spartanburg and the unfolding “resource question,” the story still felt open, and we didn’t have many details beyond platitudes, so my thoughts were suspended between promise and caution.
This week, it moved. Spartanburg County Council approved the next step for the proposed artificial-intelligence data center after a packed, tense public meeting, advancing the roughly $3 billion project despite vocal opposition from residents concerned about its environmental and infrastructural impacts. The meeting stretched for hours, with hundreds of people filling the chamber and hallway to voice concerns about the scale of the facility planned for the Tyger River Industrial Park. In other words, the decision process is no longer theoretical. It is unfolding in real time (and hopefully with more transparency), and that matters for the path ahead.
Large data center announcements are consistently appearing in public discourse (at least here in the Carolinas), wrapped in abstraction and NDAs, surrounded by investment totals, job counts, and innovation narratives that feel distant from everyday life. But once approvals begin, the conversation shifts from what might happen to what must now be managed. Water withdrawals stop being projections, and power demand stops being modeled. Land use stops being conceptual while all of this becomes material. The movement of Project Spero into the next phase signals that Spartanburg is entering precisely that transition, moving from imagining a future to negotiating its physical cost.
One of the most striking claims emerging from the latest reporting is the developer’s insistence that the proposed AI data center will be “self-sufficient,” operating without straining local infrastructure or putting upward pressure on energy bills. On the surface, that language sounds reassuring, suggesting a facility that exists almost in isolation, drawing only on its own internal systems while leaving the surrounding community untouched.
However, this is precisely where the deeper resource questions I raised earlier become more important, not less. Infrastructure rarely, if ever, functions as an island. Power generation, transmission agreements, water sourcing, fuel supply, and long-term maintenance all unfold within shared regional systems, even when parts of the process occur on-site.
The broader context makes that reassurance harder to take at face value. Large data centers elsewhere have been documented consuming millions of gallons of water per day, and electricity costs have risen sharply in regions where such facilities cluster, with those increases often eventually distributed across customers rather than absorbed privately. That does not mean Spartanburg will necessarily follow the same pattern, but it does mean the conversation cannot end with a press release promise. If anything, the national trajectory suggests the need for clearer disclosure, not simpler assurances.
Local concerns voiced at the council meeting point to exactly this tension. Questions about transmission agreements, cost structures, and regulatory oversight are not abstract procedural details. They are the mechanisms through which “self-sufficiency” is tested in practice. The reported rejection of a large transmission proposal by federal regulators because of potential cost-shifting onto ratepayers highlights how easily infrastructure investments intended for a single industrial project can ripple outward into the broader grid. What appears contained at the planning stage can become shared responsibility over time, particularly when long-term demand growth, maintenance needs, or energy market shifts enter the picture.
The developer’s plan to generate some power on-site using natural gas, along with a closed-loop cooling system designed to limit water use, is significant and worth taking seriously. Those design choices suggest an awareness of public concern and an attempt to mitigate resource draw. But even here, the key question is not simply how much water or power is used inside the facility’s literal boundary fence. The real issue is how those systems connect to fuel supply chains, regional water tables, transmission reliability, and emergency contingencies. A closed loop still depends on an initial fill and ongoing operational stability. On-site generation still relies on pipelines, markets, and regulatory frameworks beyond the site itself. “Self-sufficient” in engineering terms doesn’t mean independent in ecological or civic terms.
This is exactly why the earlier framing of Project Spero as a resource question still holds. The challenge is not whether the developer intends to minimize impact. Most large projects today do for a variety of reasons, from economics to public goodwill to tax incentives. The challenge is that digital infrastructure, such as data centers, operates at scales where even minimized impacts can be structurally significant for smaller regions. Spartanburg is not just deciding whether to host a facility, but is deciding how much of its long-term water, energy capacity, and landscape stability should be oriented toward supporting global computational systems whose primary benefits may be distributed far beyond the county line.
The Council meeting itself was contentious, emotional, and at times interrupted by public reaction. It would be easy to read that as dysfunction, but I read it differently. That level of turnout suggests something deeper than simple opposition or support. Instead, local turnout for this sort of decision signals that residents recognize it touches fundamental questions about the region’s future and what counts as development in a place defined as much by rivers, forests, and communities as by industrial parks. Public tension often marks the moment when a community realizes that a project is not just economic but ecological and cultural.
Data centers, in this sense, are simply the visible tip of a broader shift. Across the Southeast (and especially here in South Carolina), AI-scale computing is accelerating demand for electricity, land, and cooling water at unprecedented levels, asking local governments to balance economic incentives against long-term utility strain, short-term construction jobs against enduring resource commitments, and technological prestige against environmental resilience. Project Spero brings that global tension directly into Spartanburg County. The deeper question is not whether this one facility should exist, but whether communities like ours have the ecological, civic, and ethical frameworks needed to evaluate infrastructure built primarily for planetary digital systems rather than local human (and more-than-human) needs.
Approval of another procedural step does not mean the story is finished. It means the story has entered its consequential phase. This is where transparency, ecological assessment, and long-range planning matter most, not least. Decisions made quietly at this stage often shape regional water use, grid load, and land development patterns for decades. If the earlier phase asked whether we should consider this, now the question is more likely to be how we will live with what we choose (or our elected officials “choose” for us).
What encourages me most is not the vote itself but the turnout. Packed rooms mean people care about the future of this place. They care about rivers, roads, power lines, neighborhoods, taxes, and the invisible infrastructures that shape daily life. That is not obstruction, but is civic life functioning. Project Spero may ultimately prove beneficial, burdensome, or something in between, but the real measure of success will be whether Spartanburg approaches it with clear eyes about both its opportunities and its ecological realities.
The true cost of a data center is never only measured in dollars. It is measured in attention, in energy, and in the long memory of the land that hosts it.