The datacentre sector is experiencing unprecedented pressure from a number of directions. Customers are demanding the highest levels of performance and sustainability, alongside financial, environmental and political pressure to improve energy efficiency.
At the same time, recent growth in the demand for datacentres is set to accelerate further with increasing investment in AI infrastructure, and growing concern about a lack of capacity in AI-ready datacentres.
Anu Kätkä an expert in environmental monitoring and control from Vaisala, explains how datacentre companies can respond to external pressures by utilizing the latest measurement technologies to optimize both energy efficiency and operational uptime.
Reducing CO2 emissions through sensor selection
Datacentres account for at least 1% of global electricity consumption, so notwithstanding escalating energy costs, and in the context of climate change, it is essential that datacentres are able to optimize their energy efficiency and reduce CO2 emissions.
Up to 40% of energy consumption in Datacentres is used for cooling and air conditioning. So, reduction in energy use for these purposes will help to improve energy efficiency.
However, the performance of computers and servers in datacentres can be adversely affected if there is insufficient cooling – overheating can cause IT equipment to fail, resulting in significant and costly downtime.
A fine balance therefore exists between minimising the energy consumption of cooling operations and maximising the uptime of IT equipment. This is why accurate, reliable sensors play such an important role.
The control of HVAC systems can only be as accurate and precise as the sensors upon which they depend. If the measurements are inaccurate, the output of the building management system (BMS) controller will also be inaccurate.
If a poor-quality sensor gives an incorrectly high temperature reading, the controller will react by over cooling and thereby increasing energy costs and emissions. If the same sensor gives an incorrectly low reading, cooling will be insufficient, which can lead to IT equipment working inefficiently or even breaking down.
Measurement stability is key to long-term performance
It is important to remember that measurement precision is not just about accuracy at the time of installation. All sensors drift over time, to a greater or lesser degree, so measurement stability is vitally important.
Some sensor suppliers claim very high accuracy, but they may not have high stability, which means that measurements from these sensors will damage energy efficiency, even after a short period.
Good-quality sensors provide accurate measurements that remain accurate in the long term. By ensuring the controller has accurate inputs it is possible to precisely control indoor conditions and optimize energy efficiency.
Highlighting the importance of sensor accuracy, Vaisala participated in a simulation based on three real datacentres in Europe, and found that a setpoint temperature difference of just 1°C (resulting from imprecise sensors) could increase annual energy usage by as much as 8,5%.
An important point to note here, is that the cost of high-quality sensors is insignificant (and almost non-existent) in comparison with both the energy efficiency that they enable and with the value of the IT assets that they help protect.
Take advantage of weather forecasts
By anticipating upcoming hot weather, datacentres can use predictive cooling to avoid early chiller startup, make optimal use of free cooling, and reduce chiller operating hours. This proactive approach translates into further energy savings and a lower carbon footprint.
Change sensors to reduce greenhouse gas emissions and cabling
Typically, the internal environments of Datacentres are monitored by hundreds or even thousands of sensors, and each instrument needs a connection with the BMS controller so that it can report the measurement values. Sensors with analog signals require a dedicated cable from each sensor to the controller.
However, if operators can switch to a fieldbus connection it becomes possible to daisy-chain the instruments so that only one cable is needed, and thereby dramatically reduce the cabling requirement. For example, a 100-meter aisle with 30 instruments, each with dedicated cables spaced three meters apart, with a three-meter cable drop, would require 1 695 meters of cable weighing 110 kg.
However, in stark contrast, a fieldbus communication protocol would allow sensors to be daisy-chained, requiring just 187 meters of cable, weighing just 12 kg. That amounts to 89% less cabling, which leads to more efficient power distribution and lower energy losses; reducing energy consumption and lowering CO2 emissions.
How to choose a sensor provider
Most reputable businesses are looking to lower their greenhouse gas emissions – in their own processes, in the use of their products, and in their supply chains. Consequently, when choosing sensors, it makes sense to look for a provider with a sustainability focus.
For example, Vaisala’s instruments are produced using renewable electricity and the company has set science-based targets for reducing emissions. In 2024 Vaisala received a gold medal from EcoVadis for its sustainability management system, and Vaisala has been ranked 38th in the 2025 edition of the World’s Best Companies – Sustainable Growth ranking by TIME Magazine and Statista.
The challenge: uptime is non-negotiable
Both temperature and humidity levels must be very carefully maintained to achieve the uptime requirements of large Datacentres. However, larger data halls can be more challenging to monitor because they have a greater potential for temperature variation – meaning it’s important that there are sufficient numbers of temperature sensors to ensure that all servers are monitored.
In addition to the temperature issues discussed above, IT equipment is also adversely affected by humidity. Low levels increase the risk of static electricity, while high levels of humidity can result in condensation, both of which can damage delicate equipment. Datacentres often have no maintenance workers onsite, so equipment reliability is crucial.
Summary
Accurate, stable sensors, weather forecast utilization, daisy chaining with fieldbus connectivity, and sustainability-focused supply chains all contribute to operational reliability and therefore uptime, whilst also minimising CO2 emissions.