Highlights

• A new study shows circular server design—featuring longer life, modular repair, and easier recycling—reduces environmental impact by ~29% over 16 years and lowers demand for critical raw materials including rare earths.
• Life cycle assessment reveals electronics assemblies drive most embodied impacts, and life extension paired with design for disassembly delivers immediate gains even when recycling infrastructure lags.
• Circular design offers a faster geopolitical lever than new mines: it reduces material throughput, slows replacement cycles, and lowers vulnerability to concentrated rare earth processing—one modular upgrade at a time.

---

A new open-access paper by Deborah Andrews (opens in a new tab) and Kristina Kerwin (opens in a new tab) of London South Bank University (opens in a new tab), published in Mineral Economics, makes a practical point with geopolitical weight: if the world wants to reduce vulnerability to highly concentrated rare earth and critical mineral processing, it should not focus only on new mines.

It should also redesign the machines that consume these materials. Using a detailed life-cycle case study, the authors show that designing servers for circularity—longer life, modular repair, easier refurbishment, and better end-of-life recovery—can materially reduce environmental impacts and curb demand for virgin critical raw materials (including rare earths widely used across electronics and magnets).

Design Beats Digging: What They Actually Tested

The study compares a prototype “circular” server against a standard enterprise server. Across multiple modeled life-cycle scenarios, the circular server shows consistently lower environmental impacts, with the largest headline advantage appearing over a 16-year service period: the prototype circular server’s overall impact is ~29% lower on average than a standard approach that effectively requires two standard servers over the same 16 years (given typical refresh and replacement patterns).

This does not “solve” rare earth concentration risk by itself. But it reduces exposure by lowering total material throughput and slowing replacement cycles—two levers that matter when supply chains are brittle.

Methods: Full Life Cycle Assessment, Not Marketing Claims

The authors use a comprehensive Life Cycle Assessment (LCA) rather than corporate sustainability reporting. They built a baseline model by reverse-engineering widely used servers (including those from major OEMs) and then modeled the prototype circular server using primary data where possible, supported by established databases for secondary inputs. They ran scenarios covering product life extension (refurbishment/upgrade) and two recycling pathways: a business-as-usual (BAU) recovery case (focused on a limited set of widely recycled metals) and a more advanced multi-material recovery case.

They also emphasize that many server components contain Critical Raw Materials (CRMs); even when CRM mass fractions are small, supply risk can be high. The paper cites low recycling input rates for several CRMs—often around 0–1%—which is central to the “circularity gap” argument.

Key Findings: Why Circularity Matters for CriticalMinerals

1. Electronics dominate the embodied burden. Impacts are heavily driven by electronics assemblies (motherboards/PCBs and associated components). This is where many CRMs—including certain rare earths—hide in plain sight.
2. Life extension delivers immediate gains. Refurbishment and reuse scenarios lower impacts materially; the study reports that life-extension strategies (paired with recycling) reduce impacts versus “replace quickly” strategies.
3. Recycling helps—but design enables recycling. Even improved recovery pathways deliver limited benefit if products are hard to disassemble, components are non-modular, or materials are locked into complex assemblies. Circular design increases the chance that better recycling infrastructure can actually translate into higher recovery.
4. Circularity is a near-term lever. New mining and processing capacity can take many years. Design changes (modularity, reduced fasteners, standardized parts, reduced plastics, chassis reuse) are implementable sooner—especially if procurement and standards push in that direction.

What’s Controversial or Easily Misread

• This paper is not a China-specific processing study. It does not quantify China’s share of rare earth processing or model export controls. The geopolitics enter as an implication: when upstream processing is concentrated globally, reducing throughput and extending product life lowers downstream vulnerability.
• Circular electronics are hard. The authors are candid: many electronic components are intrinsically difficult to reclaim economically. The strongest lever today may be life extension, not “perfect recycling.”

Limitations

The prototype is a non-AI server; future AI-optimized hardware could change material intensity and refresh dynamics. Results are modeled using a European-average electricity mix and recycling assumptions, so impacts vary by region and grid. The authors also note uncertainty from reliance on some secondary datasets and recommend sensitivity analyses and further work (including circular AI servers and broader social/economic assessment).

REEx Takeaway

Rare earth security is usually framed as a mine-and-refinery contest. Andrews and Kerwin show a quieter truth: design choices that are upstream of e-waste determine downstream dependence. Circular servers won’t replace the need for diversified processing. But they can buy time, reduce demand pressure, and shrink the strategic penalty of concentrated supply chains—one chassis, one fastener, one modular upgrade at a time.

Citation: Andrews, D., & Kerwin, K. (2026). Design for circularity – a data centre equipment case study. Mineral Economics. https://doi.org/10.1007/s13563-025-00587-7 (opens in a new tab)