Why some data centres deteriorate faster than others

A data centre can appear fully operational right up until the moment it is not. Cooling systems may be running, power supplies stable, and the monitoring dashboards normal. Yet, across the data centre, degradation may already be underway. The cause might not be mechanical failure or electrical overload, but in fact, corrosion. 

Data centres are designed around tightly regulated environments. Factors like temperature, humidity, airflow, and power demand are carefully managed. However, the quality of the air itself is not always examined with the same scrutiny. Airborne pollutants such as sulphur dioxide, nitrogen dioxide, ozone, acidic aerosols, and hydrogen sulphide, combined with humidity, salinity, and dust particles, can accelerate the corrosion of sensitive components. 

The effects of corrosion are gradual. They affect circuitry, connectors, and, over time, essentially any metallic components. Over months and years, it can greatly shorten equipment lifespan, leading to costly maintenance and increasing the likelihood of unanticipated failures.

The variable changes by location

One of the biggest challenges with atmospheric corrosivity is that it changes by location. This can be significant throughout regions and even within the same facility. For example, coastal locations will likely see increased airborne salinity.  For data centres in urban or industrial areas, there may be higher concentrations of sulphur or nitrogen-based pollutants. A more obvious factor is local meteorology, which may affect how pollutants are dispersed or change moisture levels in the air. 

The variations are not limited to different regions and countries. Within a single data centre, conditions can vary greatly among technical spaces and plant rooms. Data centre designs must be done holistically, as ventilation, filtration choices and other design choices that affect the air will influence exposure levels. 

When operators lack site-specific data, the decision is usually to manage risk conservatively. This means changing filters, seals, and components typically at fixed schedules, regardless of the environmental conditions. This can reduce the risk, but it might also lead to unnecessary costs and disruption. 

Measurement in practice

Atmospheric corrosivity testing assesses environmental factors that affect corrosion and converts them into data. This process commonly involves a mix of on-site monitoring and laboratory analysis to determine current conditions and predict long-term trends. 

A central function of testing is the use of corrosion test coupons. These are typically copper and silver strips. These materials respond to how corrosive the environment is and give an indication of how sensitive components could degrade over time. After exposure, the coupons are analysed to determine the corrosivity rate. 

This data will then be combined with background pollutant measurements. The particular targets are sulphur compounds, nitrogen oxides, and hydrogen sulphide. Temperature and humidity will also be recorded, as moisture can accelerate corrosion. 

Sometimes, particulate matter mass and particle size distribution are assessed. This determines whether airborne particles might pose additional risks, either by causing corrosion or by causing dust to collect more quickly within equipment. The results from these studies will inform decisions about ventilation strategies and filter performance. 

Finally, the gathered data will be classified against recognised regional standards. These include ANSI/ISA-71.04 and ISO 9223/9224. These frameworks categorise environments from low to extreme corrosivity, and by aligning findings with established standards, operators and designers can correlate findings more consistently and compare sites objectively. 

Supporting resilience

The value of corrosivity testing is in its application. By identifying high corrosivity early, intervention can take place before degradation leads to failure. It could also influence maintenance schedules to focus on higher-risk zones rather than applying a blanket across the data centre, which increases costs. 

Whilst the focus is on the risk of corrosion and failure, there can be benefits to identifying low-corrosive environments. In areas where this is true, operators could extend service intervals with improved confidence and reduce the unnecessary replacement of functional components. 

Over a long period of time, which data centres are designed for, consistent monitoring builds a record of the environmental conditions. From there, trends can be analysed to identify emerging risks or changes in pollutants. This can be seen with visual tools like time-based corrosion rate charts and can help correlate environmental conditions and equipment performance. The results will also benefit future site and component selection decisions. 

An evidence-led approach

Power availability and cooling resilience will always be central concerns in data centre design. However, air should not be ignored. Air is everywhere in the facility and is an equally important factor in long-term reliability. Whilst corrosion does not attract the same attention as electrical or mechanical faults, its cumulative effect is significant.

By means of atmospheric corrosion testing, the industry can move from precautionary maintenance to confident decision-making enabled by data. This approach will benefit operators looking for resilience and cost efficiencies in their valuable infrastructure portfolio. 

With data centres expanding through different locations, and with the atmosphere itself changing to become more extreme, knowing local conditions will be vital. The industry needs to understand that if a risk is unseen, then it likely deserves closer attention.