As data centres pursue ambitious efficiency and ESG targets, Shawn Gilstorf, Vice President Products at Rensa Filtration, explains why reducing operational waste could offer an overlooked opportunity.
Ask a data centre operator about sustainability and they’ll tell you about power. PUE, renewable procurement and cooling efficiency have defined much of the industry’s environmental focus for the better part of a decade, and for good reason.
Power remains a defining operational and environmental consideration, while the rapid expansion of AI infrastructure is increasing the pressure to improve efficiency. But there is another sustainability story playing out beyond the energy dashboard: the waste generated by the routine materials and consumables required to keep a data centre operating.
As AI workloads drive up compute density, cooling requirements and the amount of equipment operating within a facility, the volume of materials required to maintain that infrastructure also deserves greater attention. Material circularity – keeping resources in use and out of landfill – is becoming an increasingly important consideration for operators looking to reduce the environmental impact of their sites.
More equipment and more intensive operations can mean more maintenance and more of the everyday consumables that keep a site running. Yet these recurring waste streams can be difficult to see in the same way that energy and water use are measured. Air filters offer a useful example of how a routine maintenance item can become a significant waste stream when multiplied across a large facility.
The tonnage hiding in plain sight
The first problem is simply seeing what is there. Maintenance waste can be easy to overlook because individual items rarely seem significant. This picture changes when operators track what is purchased, replaced and discarded over the course of a year. A useful step is to identify recurring consumables and group them by volume, frequency of replacement, material composition and current end-of-life route. Depending on the facility and its operations, these materials can include:
- Air filters
- HVAC maintenance materials
- Packaging and shipping materials
- Batteries and electrical components
- Cooling system consumables
- PPE and general maintenance supplies
Air filters are a useful example. A large facility may use hundreds or thousands of filters over time as part of HVAC and cooling maintenance. Because many filters combine different materials, including frames, media and other components, their end-of-life management can be challenging, particularly where local recycling infrastructure is not equipped to process them.
Individually, each filter may seem insignificant. Collectively, however, recurring replacements can represent a substantial material stream that may receive little attention in broader sustainability reporting.
Why ‘recyclable’ isn’t always straightforward
Here’s the catch that trips up a lot of good intentions: something is not necessarily recyclable just because it’s made of recyclable materials. A typical filter combines a frame, synthetic media, adhesives, gaskets and sometimes wire backing into a single unit. Separating these components economically can be challenging, and the contaminants collected during use can make it harder still.
For years, much industrial equipment has been designed primarily around how it performs in use. What happens afterwards was often not part of the specification, and procurement teams accordingly optimised for performance and unit price.
Make end of life part of procurement
Reducing operational waste often starts with procurement. When evaluating recurring consumables, operators can ask suppliers not only how a product performs and what it costs, but also what happens to it at the end of its useful life. What materials does it contain? Can those materials be separated or recovered? Is there a practical recycling route available in the locations where the product will be used? And can the supplier provide evidence of that pathway?
Filters that combine multiple materials can present greater challenges for recovery than products designed with fewer material types. That does not mean a particular material or product is automatically the better choice. Performance, availability, total cost and local recycling infrastructure all need to be considered together. The important shift is to make end-of-life management part of the purchasing conversation rather than treating disposal as an issue to address after a product has been selected.
Scale can create an advantage
The volume of waste generated by large data centres can be a challenge, but it can also create an opportunity. Recurring maintenance materials generated at a single site may provide enough volume and consistency to make collection and recovery more practical than they would be for smaller facilities. Predictable replacement cycles, established supplier relationships and existing site logistics can also help operators build repeatable processes for collecting and handling materials.
This doesn’t make circularity automatic, however. Operators still need to establish whether there is a viable recovery route, how materials will be collected and stored, who will transport them and what the process will cost. But where sufficient volume exists, the scale of a data centre can make it easier to move from an occasional recycling effort to a measurable waste-diversion programme.
Creating a practical way in
Addressing hidden waste doesn’t require an entirely new sustainability programme. It starts with a repeatable process that can be incorporated into existing maintenance and procurement practices.
- Audit the waste: Identify recurring maintenance consumables, track how many you replace in a year and estimate how much of each stream is sent to landfill.
- Prioritise the opportunity: Not every waste stream will justify the same effort. Consider volume, replacement frequency, disposal costs, material composition and the availability of practical recovery options when deciding where to start.
- Put the question to suppliers: Ask how their products are built and what recovery or recycling options exist – and do it before you buy, not after.
- Measure the result: Track a few clear KPIs: tonnes diverted from landfill, recycling participation, the share of procurement that supports circularity and year-on-year waste reduction.
These metrics can also support ESG disclosures and zero-waste commitments, helping operators understand and report on the impact of their waste-reduction efforts.
Account for the operational trade-offs
Circularity also has to work within the realities of operating a data centre. Recycling infrastructure varies by region, so a recovery route that works at one site may not be available at another. Collection and storage require space and processes, while transportation can add cost and environmental impact.
Procurement teams also need to balance sustainability objectives with performance, availability and total cost of ownership. Maintenance teams may need new procedures for handling and separating materials, while collecting reliable data can require additional effort at the outset.
These practical considerations do not eliminate the opportunity, but they do mean that waste-reduction programmes need to be designed around how a facility actually operates.
Forging ahead to the next frontier
The next chapter of data centre sustainability won’t be written on energy efficiency alone. It will be about how well operators manage materials across the whole life cycle. As the industry pursues ambitious ESG and zero-waste targets, there may be opportunities to divert waste that sit outside the metrics operators already track.
Air filters are only one example of how a routine consumable can be rethought through circular design, recovery that scales and more complete accounting. Handled well, the very growth that makes the AI era so resource-hungry could push the industry somewhere genuinely more circular – one unglamorous waste stream at a time.
