The pipeline of U.S. data center construction projects has reached more than 500 sites and counting. With that boom comes a reckoning as communities in states from Wisconsin to Virginia scrutinize — and question — the heavy electricity, water and land requirements of what NVIDIA CEO Jensen Huang has dubbed “AI factories.”

In early October, Microsoft retracted its plan to build on a site in Caledonia, Wisconsin, after local residents protested the potential conversion of agricultural land into an industrial site. The company will pursue development at another location in the state.

“One of the pieces of feedback that we heard was that the parcel was too closely situated amongst other residents,” said Kaitlin Chuzi, director of integrated technology and biomimicry at Microsoft, during a recent Trellis Impact 25 discussion about how to include nature and biodiversity in data center development decisions. 

Engaging communities early is one of the best ways to address land-use questions and co-develop strategies for habitat restoration around new data centers or at other locations, according to the panelists. It starts with landowners and local officials who can connect companies with local conservation organizations to better understand community concerns. 

“Really listening to what they’re hearing is a really helpful way for us to develop an understanding but also ferment some of the ideas about which types of work would be most impactful for the community,” said Emily Backus, sustainability director for North America at Vantage Data Centers, which operates campuses for hyperscalers, cloud service providers and other companies with large data center needs.

Another key stakeholder to include from “Day 0,” Backus said, are utilities. If a site requires new transmission lines, for example, that will dramatically increase the potential impact on nature and biodiversity. 

Part of the plan

Data center companies such as Microsoft and Vantage have integrated habitat and biodiversity impact assessments and water studies into their standard design processes from the onset, especially as communities scrutinize a flood of data center site applications. Microsoft’s Chuzi, for example, reports to the company’s land management team; her role is not part of the sustainability office. Backus, on the other hand, collaborates closely with her company’s development engineers. 

While there’s no typical land use requirement for data centers — and operators are encouraging increasingly higher server densities — hyperscale facilities can still require 200 to 500 acres.

Vantage uses unbuilt acreage to restore native habitats and encourage drought-resistant landscaping, eschewing turf or non-native plantings. By the end of 2024, it used this strategy at half of its site footprint, or about 4,000 acres.

One of Microsoft’s climate goals, set in 2020, was to protect more land than its operational footprint requires by the end of 2025. It has already met that pledge, with more than 15,849 acres permanently protected, as of its latest environmental report. Biodiversity remains a focus amid its furious AI-related expansion phase.

Biodiversity and ‘fireball squirrels’

Both Vantage and Microsoft embrace the philosophy of biodiversity net gain — the idea that development can make land more supportive for new species — to guide new projects. 

Vantage, for example, used rain gardens, vertical greenery and other nature-based features to manage stormwater at its MPX2 campus in Milan, Italy, and restore the local ecosystem. It also planted multiple species of trees and flowering meadows at the campus. Where possible, given construction codes, it adds green roofs to data center buildings. 

That could mean tradeoffs, however, from an emission standpoint, because more steel might be needed to support the weight of plants. Other tradeoffs include higher maintenance needs. 

“The big question that my team and I have been talking about recently is whether it is worth it to put a green roof data center or, for the same cost, do ecological restoration on 100 acres off site,” said Chuzi. Her preference would be to do both.

Vantage will use a metric developed by consulting firm Ramboll to measure the impact of its North American projects, starting with two data centers in the Midwest. This was inspired by projects in the U.K., where demonstrating biodiversity net gain is mandatory for developers. This requires hiring ecologists who can perform initial and ongoing site evaluations.

“You think about all those green spaces around your buildings and your parking lots,” Backus said. “All of those are opportunities to do really significant improvements that can improve the biodiversity of those areas.”

Microsoft likewise looks for ways to add biological buffers and habitats to its campuses, a practice that was initially controversial with engineers, who were worried about squirrels and other rodents chewing through electrical lines. 

“They have so many valid reasons for why you can’t bring nature closer to the technology that we’re bringing to the site,” Chuzi said. “I think my favorite one, and maybe the most memorable excuse was, ‘Kaitlin, you can’t put plants near the data center because we’re going to end up with fireball squirrels.’”   

Much of Chuzi’s role involved balancing these concerns with biodiversity goals. The rodent problem, for example, can be addressed by encouraging habitat friendly to predatory birds that can help manage the population. A buffer can be established between plantings and crucial electric and transmission systems. And electrical systems can include additional insulation.

“It’s a lot of pulling together opposing ideologies and changing how people think about technology, how they think about nature,” Chuzi said. 

Source: Microsoft

Innovation inspired by nature

Nature is also reshaping design inside data centers. 

Concerns about water, for example, inspired both Microsoft and Vantage to prioritize the installation of “waterless” cooling equipment that relies on outside air to dissipate heat or closed-loop systems that retain and recycle water. 

Both companies are also looking for ways to funnel the heat generated by servers to places where it could be beneficial. Microsoft shares excess data center heat with a municipality in Finland, for example. That takes collaboration with neighboring sites, another reason community engagement is crucial for advancing data center development.

Nature was also the inspiration for a new AI chip design at Microsoft that uses a microfluidics approach to keep cool; the company says it is three times more effective than cold plates. Tiny channels move coolant through the silicon, much the same way that the veins of a leaf deliver water and nutrients.

These ideas all have roots in biomimicry, a discipline that reflects natural worlds in new product and systems design. Two of the most famous examples are Velcro, which was inspired by burrs, and the bullet train, which takes its cues from the anatomy of owls and kingfishers. 

“If you start looking at data centers and how we design them through the lens of ‘How would nature do it in this place or how does nature solve problems like this?’ then you start to come up with really incredibly imaginative new designs that are better for the environment and better for the communities and better for the companies that are designing them,” Chuzi said.

The post No ‘fireball squirrels’: Nature and next-generation data center design appeared first on Trellis.

VF Corporation has already mapped the global supply chains of 100 products, including Dickies overalls, Smartwool base layers and The North Face hiking boots. Now, the Denver company reports significant progress toward tracing five key materials down to the commodity level by 2028.

The company disclosed in its Oct. 7 sustainability report that it has achieved 61 percent traceability by volume, the first time it has shared this metric, for certain animal, plant and fossil materials. Wool leads at 78 percent; leather is next at 67 percent. They are followed by natural rubber (56 percent), synthetics and cotton (both 51 percent).

Purchasing more “preferred and regenerative materials” and reducing production volumes helped VF cut Scope 3 emissions by 7 percent in 2024 compared with 2023, though the footprint remains 19 percent above a 2017 baseline.

“We can’t fix what we don’t know, and this work is key to compliance,” said Harsha Chenna, VF’s VP of global product stewardship. Traceability helps meet responsible sourcing policies and regulatory compliance and monitors risks including human rights, water scarcity, deforestation and pollution, he added.

Five key materials

The company said 5,003 metric tons of its five key materials were produced with regenerative practices in 2024. Among the progress it marked for each material in 2025:

• Cotton: VF traced 54 percent across all tiers, aiming to source all of the material from Australia, the U.S. or other regions with sustainable origins by 2026. Regenerative cotton represented 94 percent of the company’s 2024 supply, up from 89 percent the previous year.
• Synthetics: VF traced 56 percent, from Tier 1 to 5. Recycled material accounted for 64 percent in 2024, up from 48 percent in 2023, surpassing its 2026 goal early. Ninety-five percent of polyester and 80 percent of nylon used by The North Face came from preferable sources.
• Leather: VF traced 75 percent from Tiers 2 to 5. Timberland used 78 metric tons of regenerative leather in 2025, and 90 percent of footwear tanneries were Leather Working Group-audited.
• Rubber: VF traced 74 percent of the natural supply chain. The North Face, Timberland and Vans collectively used 319 metric tons of regenerative rubber in 2024.
• Wool: VF traced 87 percent across all tiers and counted 94 percent ZQ-certified wool in 2024.

Traceability history

VF began dedicated traceability work in 2017, piloting material mapping across 10 products. By 2021, it had expanded to 100 products across 1,200 sites. The company seeks to map materials down to the commodity level, relying on data including supplier traceability, country-level risk maps and environmental and social risk metrics.

“Traceability is still very difficult, requiring significant time, resources and analysis,” Chenna said. VF’s teams in North America and Asia-Pacific report directly to him.

VF conducted 1,070 supply chain audits in 2025. It publicly shares addresses and details for 1,214 suppliers via the Open Supply Hub, updating quarterly. Most suppliers are in the Asia-Pacific region, followed by the Americas and Europe.

VF is also advancing circular design, which represent a tiny slice of sales, as is common for apparel companies with branded resale programs. Its The North Face Renewed secondhand brand sold 96,000 items in the U.S. and U.K. last year. New The North Face products, in contrast, sell in the hundreds of millions annually.

Emissions targets, market challenges

Raw materials drive 79 percent of VF’s climate footprint: polyester makes up 25 percent, leather 19 percent and cotton 18 percent.

Validated by the Science-Based Targets initiative, VF aims for net zero by 2050. Near-term goals include cutting Scopes 1 and 2 emissions by 55 percent and Scope 3 emissions by 33 percent by 2030 from a 2017 base. The recent inclusion in the footprint of refrigerants from HVAC systems in 1,127 stores in 2017 complicates year-over-year comparisons of overall emissions.

VF has faced scrutiny from watchdog group Stand.earth for lagging climate progress. The company does not appear on track to lower supply chain emissions by 33 percent by 2030, according to Stand’s Senior Corporate Climate Campaigner Rachel Kitchin. However, she praised VF’s increased use of regenerative cotton, and tracing some materials to Tier 5.

VF’s annual revenues dropped nearly 4 percent to $9.5 billion in 2025. The company’s 2025 sustainability report followed workforce reductions a month earlier, including the departure of Senior Director of Sustainability David Quass. However, the sustainability work continues apace, according to the company.

The post VF Corporation hits 61% traceability for key materials appeared first on Trellis.

The opinions expressed here by Trellis expert contributors are their own, not those of Trellis.​

The embrace of nuclear power is often presented as a climate solution and in isolation, this characterization has merit. Nuclear energy is a clean, reliable power source that can help achieve climate goals while meeting growing energy demands. However, the current deployment of nuclear power is more accurately understood as a private solution to a public problem — one that forces us to confront questions about risk, responsibility and the kind of future we’re willing to create.

The new energy aristocracy

Let’s start with the basics: Microsoft, Amazon, Google and Meta have invested more than $20 billion in nuclear energy and its surrounding technology. In March, these four tech giants formalized their commitment by signing a pledge to triple global nuclear capacity by 2050.

This isn’t corporate virtue signaling; this is the world’s most valuable companies securing competitive advantage through energy infrastructure control. To provide a significant fraction of the terawatt-hours of electricity these tech giants consume each year, nuclear companies will likely need to build dozens of new plants, not just a couple of reactors. 

Yet the nuclear capacity being secured by tech companies represents a tiny fraction of what’s needed to decarbonize the broader economy. According to the International Energy Agency, U.S. data centers alone consumed roughly 4.4 percent of the country’s total electricity consumption. By 2030, this figure is projected to grow by 133 percent to 426 terawatt-hours. The nuclear deals announced by tech companies, while substantial, amount to perhaps six to seven gigawatts of capacity when fully operational — a meaningful addition, but less than 2 percent of projected U.S. electricity demand growth.

This creates a two-tiered energy system where a company’s ability to maintain climate commitments becomes increasingly correlated with market capitalization. When the most powerful corporations purchase their way to carbon-free electricity through long-term nuclear contracts, they simultaneously absorb significant capital and regulatory attention that might otherwise be directed toward more scalable solutions. Nuclear construction requires specialized regulatory oversight, supply chains and expertise that exist in limited supply. Every dollar and engineer devoted to private nuclear projects is, in effect, unavailable for grid-scale decarbonization.

The uncomfortable truth is that we’re witnessing the privatization of climate solutions at precisely the moment when we need coordinated, systemic action. This isn’t a moral failing on the part of tech companies — they’re rational actors responding to market incentives. But their rational behavior creates externalities that undermine broader decarbonization efforts.

The moral labyrinth

Nuclear power presents a profound moral dilemma that cannot be resolved through simple cost-benefit analysis. The arguments on both sides have genuine weight and grappling with this technology requires acknowledging the legitimacy of competing frameworks.

Advocates of nuclear power make a case grounded in consequentialist ethics: nuclear energy’s massive carbon-free output and superior safety record compared to fossil fuels create an obligation to deploy it widely. When you consider that climate change will disproportionately harm the world’s most vulnerable populations — those with the least responsibility for emissions and the fewest resources to adapt — the failure to deploy available low-carbon technologies can look like negligence. 

From this perspective, opposition to nuclear power becomes a form of moral luxury that wealthy environmentalists can afford while the world’s poor will pay the price in heat waves, crop failures and climate-driven conflict.

Data supports this position in important ways. Nuclear power has caused fewer deaths per unit of energy produced than any fossil fuel source, including natural gas. Even accounting for Chernobyl and Fukushima, nuclear energy’s safety record is exceptional when measured against the continuous toll of air pollution from coal and gas, which kills millions annually. If we’re serious about preventing catastrophic climate change while maintaining modern standards of living, can we afford to dismiss a proven technology capable of generating baseload power without carbon emissions?

Critics counter that nuclear power imposes unacceptable risks of catastrophic accidents — events whose probability may be low but whose consequences are nearly unbounded. More fundamentally, it creates radioactive waste requiring secure management for periods that exceed the entire span of recorded human history. This represents a form of intergenerational tyranny where present benefits come at the expense of future burdens we cannot fully imagine, let alone mitigate.

Something is troubling about creating materials that will remain lethally toxic for 10,000 years when we can barely predict social and political conditions a century hence. What right do we have to impose such obligations on future generations who’ll have no say in the matter and receive none of the benefits? This isn’t abstract philosophizing; it’s a fundamental question about acceptable levels of imposed risk and responsibility across time.

Concentration of risk, diffusion of benefit

Nuclear facilities typically concentrate risks in local communities while benefits flow elsewhere. The communities hosting nuclear plants — and especially those designated for waste storage — bear elevated accident risks, property value impacts and the psychological burden of proximity to hazardous facilities. Meanwhile, the electricity generated flows hundreds of miles away to power data centers and urban areas whose residents enjoy the benefits while remaining insulated from the risks.

This pattern mirrors broader environmental justice concerns, where marginalized communities disproportionately host society’s hazardous infrastructure. Nothing is necessarily malicious about this arrangement, but it does represent a form of spatial inequity that we should at minimum acknowledge.

The technology’s complexity also undermines energy democracy by requiring centralized expertise and institutions. Nuclear power cannot be deployed locally or managed by communities. It demands national regulatory frameworks, specialized engineering knowledge and institutional continuity spanning decades. This stands in stark contrast to distributed renewable generation, where homeowners and communities can directly participate in energy production. The choice between these paradigms isn’t merely technical; it’s a choice about the kind of society we want to inhabit and who holds power over critical infrastructure.

Beyond calculation

We’re confronting fundamental questions about acceptable levels of imposed risk, intergenerational responsibility and the wisdom of creating technological systems whose consequences far outlast their creators.

The current moment makes these tensions particularly acute. Tech companies deploying private nuclear capacity are, in effect, making civilizational decisions about risk and responsibility that will reverberate for millennia. These decisions are being driven by market logic and corporate procurement strategies rather than democratic deliberation about the kind of energy future we collectively choose.

This doesn’t mean nuclear power is necessarily wrong, but it does suggest that the question, “Should we deploy nuclear energy?” cannot be separated from questions about who decides, who benefits, who bears the risks and what alternatives exist. What we need is a more sophisticated framework for making collective decisions about risk, responsibility and the distribution of both benefits and burdens. This requires moving beyond individual corporate procurement strategies toward systemic solutions that ensure clean energy deployment serves broad social goals rather than narrow commercial interests.

The nuclear question, ultimately, is about what obligations we owe to those who come after us, what risks we have the right to impose on local communities, and whether our most consequential technological choices should be determined by market forces or democratic deliberation. These aren’t questions that physics or economics alone can answer. They require reasoning about the kind of world we want to create and the legacy we’re willing to leave behind.

The post The catch-22 of nuclear power appeared first on Trellis.

When the last-mile delivery team at Walmart began to craft plans to scale service, it faced problems only behemoths have to contend with: which vehicles were right for a network of 4,700 U.S. stores and 1.6 million employees, and if the company decided to go electric, could it actually set up charging infrastructure in all those locations?

“The complexity of Walmart is that we’re everywhere,” said Sai Teja Yerapothina, the company’s senior director for last-mile delivery, strategy and operations. “And we deliver everything from hot chickens to 75-inch TVs to prescriptions.”

Around four years later, Walmart is bringing groceries, electronics and other goods to customers in 2,000 electric vans that are charged at stores.

At last week’s Trellis Impact 25 event, Yerapothina shared details of how the company made the move.

Selecting the ride

Walmart’s leaders impose tough requirements on the returns of every investment, so Yerapothina’s team had to get granular in its assessment of vehicle options. To estimate the total cost of ownership, they factored in everything from vehicle price and energy used to resale value, maintenance and time spent traveling to gas stations. 

“We ran various different cost models about what this would mean for the business, for our customers, for our associates,” said Yerapothina. “EVs turned out to be our cheapest option and also the most sustainable option. So it was an easy decision for us.”

Today, Walmart contracts with Ford and GM for its fleet of last-mile delivery vehicles.

Infrastructure experiments

Level two charging, which provides speeds in the 10-kilowatt range and can recharge a vehicle overnight, was sufficient for the fleet. It is also cheaper, because the cost of electricity is often lower at night.

Beyond that, things got complicated. Permitting requirements differ across regions, as do the utilities involved. In addition, even nearby stores have slightly different designs. Some buildings are owned; others are leased. Parking lots can be for the sole use of Walmart customers or shared with neighboring businesses. “Every single location came with a different challenge,” said Yerapothina.

Rather than overanalyze the situation, his team began by deploying charging stations to a few locations then closely monitoring the results. “The best way we learn is we put something out there in the real world, give it to our associates and ask for feedback,” explained Yerapothina.

Wi-fi signals were one challenge they encountered. Sometimes the best spot for a charger lies beyond a store’s wi-fi range, and that required the team to equip some of them with cell-signal connections. The connectivity allows custom-built Walmart software to monitor the status of the charger and any vehicles connected to it. 

Building the playbook

Over time, the last-mile team worked with colleagues in energy and software to build a playbook that guided the expansion of the charging network to thousands of stores. They also established standards that hardware and software vendors need to comply with.

Though the playbook identifies preferred locations for chargers, it isn’t prescriptive. “Because the best way may not always be the right way for every location,” said Yerapothina. “The final approval of where a charger would go and the other design decisions was left to store managers.”

Guiding design principles also include simplicity and transparency. “We start with: Is this simple enough that I could stand with an early associate in a store and explain to them in 30 seconds?’ said Yerapothina. 

“Always start with the why,” he added. For example, some drivers switched off the one-pedal driving feature, which uses resistance in the motor to simultaneously brake and recharge the battery. The team responded by explaining the benefits of the approach, which include lower charging costs and reduced brake wear. “We’ve seen a huge uptick in adoption,” noted Yerapothina.

Next up: Drones

With the last-mile EV fleet in place, Yerapothina’s to-do list now turns to building out drone delivery. The company has fulfilled around 300,000 orders over the past few years using drones from stores in Dallas, Houston and Arkansas. 

For single-family homes, it’s a relatively straightforward proposition to drop deliveries in the backyard. Yerapothina said the company is still working on a solution for multi-unit buildings.

“There is a future where customers are oblivious to the mode,” he predicted. “They just say, this is the speed I want it in. And then the retailer figures out how to get it there — and drones could be one of those options.”

The post How Walmart built a playbook for deploying thousands of EVs and charging stations appeared first on Trellis.

Which feels easier: taking an empty milk jug to the recycling bin or forking over tax dollars in the name of sustainability?

If you chose the former, you’re not alone. According to new insights from Trellis data partner GlobeScan about how societies are responding to the sustainability transition, convenience is king. When asked how willing they are to make specific changes, people are most willing to make changes when they’re low-cost, familiar or relatively easy to do. More transformative changes that require financial sacrifice or lifestyle overhaul face greater resistance. People are most willing to take everyday, tangible actions that feel manageable such as:

• Recycling (89 percent)
• Avoiding plastic packaging (88 percent)
• Reducing energy use (85 percent)

Support drops significantly for more costly or disruptive actions such as minimizing living space (54 percent) or paying higher taxes (33 percent).

What this means

GlobeScan’s research shows that while people broadly support the green transition, their willingness to make hard personal sacrifices, especially financial ones, is limited. Instead, people prefer actions that feel familiar and relatively effortless. This signals a clear imperative: make sustainability frictionless. To accelerate the transition, governments and businesses must embed low-effort, high-impact actions into everyday life and design systems that reward sustainable choices without demanding sacrifice. 

Based on a survey of 31,960 people across more than 30 countries conducted July-August 2025.

The post No costly sacrifices: consumers want ‘easy’ sustainability actions appeared first on Trellis.