Data Centre and computer room environment management

1. Why does environmental monitoring matter?

Computer room hardware (servers, switches, storage arrays, UPS systems, and cooling units) all produce significant heat and often operate to tightly controlled conditions, set by manufacturers and others.

Factors affecting performance can range from issues with room layout and operation to internally and externally generated contaminants and poor management. These can include:

• Thermal shutdowns and hardware damage caused by overheating equipment;
  Electrical short circuits and corrosion caused by condensation, high humidity, zinc whiskers (and other metallic contaminants);
• Electrostatic discharges due to low humidity;
• Long term temperature issues, power fluctuations and contamination leading to data loss, reduced hardware performance/life span; and
• Complete service outages.

For most businesses uptime is very important, for others it is critical (telecoms, healthcare, finance, etc.). Minimising potential downtime both proactively and reactively will always be an investment, bearing in mind costs can very quickly escalate should capability be lost.

2. What are the key environmental factors to monitor?

The tolerance of IT hardware (thermal and environmental) has changed over the years, with most equipment being more robust than its predecessors, although no less fickle. Rooms themselves have also changed with equipment size and levels (remember the old Storagtec units?!) having reduced then increased again – volume of kit in particular has really increased.

Almost all computer rooms and data centres come with some form of built in capability, all though sadly the old wall mounted analogue thermo-hygrographs (usually with a pronounced hunting pattern), are a thing of the past. CRAC units, cabinets and BMS sensors to name but three all have monitoring capacity and provide ongoing information on certain, but not all parameters.

Despite these changes, the parameters you need to be on top of, when looking at your rooms and datacentres, have not changed significantly.

• Temperature

Whether from the servers themselves or other room dynamics, temperature is a key component for uptime. How the room is set up (hot aisles and cold aisles), whether the room users have understood the room when stacking, equipment re-arrangements, cooling failure or airflow blockages, as examples can all cause thermal variations, which in turn can lead to issues.

Measuring ambient temperatures at multiple heights (top, middle, bottom of racks), monitoring server inlet and outlet temperatures and general room temperatures (rationalised against layout), all provide real time and targeted information.

• Airflow

As important as temperature, but often forgotten, regardless of your temperature performance, if your racks are not getting enough air, problems can occur.

How the air is distributed, the condition of plenums and the stacking of racks all can play a part in how well the air is flowing through your room. Considerations should include monitoring, airflow rates into and out of hardware and racks as well as from supply grilles throughout the room. Rack and IT equipment position, in relation to up/down, front/back and side to side airflow is also important for localised hot spotting.

• Pressure

Your data centre/room needs to be active in discouraging external contaminants. Your primary air supply, entrance doors and other possible ‘holes’, come into play here. Differential air pressure measurements in and across these areas, as well as between, for example, hot and cold aisles all contribute to keeping the room balanced and dust and debris out.

• Room condition/cleanliness

Not just macroscopically, with the larger items of cardboard, cables, food/drink (you still see them), and manuals, etc., but microscopically, the collective dust and debris you often cannot see.

As well as carbon and organic matter entrained from the outside, contaminants can include plastic, metal fragments/ferromagnetic particles, crystalline material, textile fibres, paper dust and possibly zinc whiskers (see white paper 5th September 2025 - Assurity Consulting Whitepaper - Zinc whiskers 2.0). Their prevalence and composition can provide lots of information on the factors influencing the room, from poor quality filtration to wall and floor degradation.

• Relative humidity

Humidity levels in a computer room, much like temperature can have a direct effect on hardware/hardware reliability and other room, particularly metal surfaces, and this parameter is another with a range of manufacturers and others guidance.

It is also a parameter that needs to be considered in a range, as too high relative humidity can increase the risk of condensation occurring, which can then lead to short circuiting of hardware or corrosion on relevant surfaces across the room. Conversely too low relative humidities sees the risks of electrostatic discharges occurring, which is also very undesirable in this environment.

• Air quality

There are numbers of standards/best practice/manufacturers guidelines that can be used to assess the quality of air in a controlled environment (e.g. ISO 14644 and EN 50660-2-3). These provide levels for the room cleanliness, usually through the maximum permissible amount of particles/particle concentrations in a volume of air.

Whereas, most monitoring of rooms will use these guidelines, not everyone sense-checks the room to look at where the readings should be taken for best effect. Often factors such as pressure readings, rack configuration and CRAC unit condition are not considered, but do play a big part in what air quality actually looks like. Room cleanliness (see above) is another big contributory factor.

• Water

Unwanted water in a computer room is never good and can emanate from several sources, e.g. condensation (see above), pipework bursts or leaks, and external leaks.

Most rooms will today have some form of leak detection system (sensing cables/rope and/or spot/probe detectors), usually with single or multiple zones. Making sure it is all still in the right or relevant place is of course fundamental to getting a suitably early warning, but you would be amazed as to how much this equipment can move about (or have been badly positioned in the first place).

• Fire

With electrical fires being a significant risk in commercial buildings, good quality detection and suppression is essential.

It is true to say that most of the room we assess have an active approach to fire and smoke with aspirating air and other detectors employed and suppression systems designed specifically for IT environments, be they gaseous, foam or water based. However, making sure control panels are in good working order and detectors heads free to do their job should be part of routine checks. Significant hot spotting could also mean a modification/addition to the system is needed to reduce the risk of overheating and fire.

It would also be wise to acknowledge that other factors such as power/electrical supplies and physical security in computer rooms and data centres are vital components in room management, if not room environmental management (these are not discussed as part of this white paper therefore).

Similarly, there are a number of aspects of wider computer-room and data centre management, outside the hardware-controlled environment, that are equally important in producing a quality and safe facility. These should cover aspects such as, fire, Legionella, noise and general health and safety, and again are not discussed as part of this white paper).

3. In summary

Almost all computer rooms and data centres come with some form of built in capability, although, sadly, the old wall mounted analogue thermo-hygrographs (usually with a pronounced hunting pattern), are a thing of the past.

Having the ability to be able move round a room, and pass an educated eye over set-ups and layouts, adds a further dimension to the data, as it allows for more local variations and their influence to be directly checked/challenged.

A strategy for monitoring environmental performance should not only cover parameters such as, thermal conditions, airflow, pressure, cleanliness, humidity, air quality, water/leaks and fire, but the interaction of each element and the part they are playing in maintaining a stable environment and so uptime.

With the right combination of sensors, analytics, and operational processes, coupled with an independent, specialist overview of performance, more data centre and computer room owners and operators should be able to minimise problems, prevent outages, and optimise their computing environments for both performance and cost efficiency.

If you want to benefit from over 40 years of experience in independently monitoring computer-rooms and datacentres, please get in touch.

Assurity Consulting is the UK’s leading independent consultancy specialising in workplace health, safety and environmental solutions. As your partner in compliance management you will reap the benefit of our more than 40 years’ experience of helping customers across a range of different sectors – manage their compliance responsibilities as effectively as possible. If you need any help with your health, safety or environmental compliance, or if you would like more information on the services Assurity Consulting offer, please get in touch.

DOWNLOAD PDF