According to The Copernicus Climate Change Service (C3S), temperatures for Europe show long-term warming trends for both annual and seasonal averages. Although, this isn’t the only region suffering as extreme weather hit all seven continents in July 2023. The intense conditions are a problem impossible to ignore.
Recent data revealed that data centres account for around 1% of global electricity usage, according to the International Energy Agency. However, they can also act as catalysts for innovative solutions to aid their environmental footprints while functioning more efficiently.
Following climate change-related water shortages across the European continent causing governments to take action, a process industry expert is highlighting the knock-on effect on manufacturers reliant on waterways for cooling processes.
Matt Watson, Sector Lead for Manufacturing at Aggreko Europe, is emphasising the rise in extreme weather across the continent and how they may necessitate a rethink in manufacturers’ equipment procurement strategies. Specifically, with reports of heatwaves and droughts across Europe, Watson is calling for manufacturers to explore closed loop cooling solutions that do not take water from nearby communities.
“Water from riverways is vital to many manufacturers’ existing cooling processes but climate-related news across the continent is leading to a situation where this precious resource may become increasingly scarce,” said Watson. “News from Germany, Italy, Spain and France about governments enacting water preservation measures is something the continent’s manufacturers should watch with interest, especially ahead of the hot summer months.
“Consequently, manufacturing sites currently reliant on local water bodies for cooling purposes should look to adjust their processes accordingly to reduce river water use. By implementing closed loop systems that do not require as much use and subsequent cooling of water, manufacturers could mitigate costly drought-related issues that might result in expensive site disruption and shutdowns.”
With water scarcity expected to be a major concern across the continent, manufacturers are increasingly considering hire-based equipment procurement strategies to address weather-based uncertainty without limiting purchasing power. According to Watson, consulting with temporary suppliers capable of providing closed loop cooling systems promptly in response to water shortages is a vital first and incremental step to ensuring a site’s continued smooth running.
“The pressure to do more with less continues to affect all industries, including the manufacturing sector,” Watson added. “As such, the business case for purchasing expensive, permanent equipment to address something as changeable as the weather may be inadvisable with costs being so closely monitored.
“By contrast, options such as hire from Europe-wide suppliers could provide these manufacturers with the dynamism they need to maintain site productivity. I would therefore advise relevant stakeholders consult with relevant experts from organisations such as Aggreko, which have the Europe-wide expertise and reach to help mitigate these concerns.”
Stephen Donovan, Chief Development Officer, atNorth
With Europe’s climate warming twice the rate of the global average, many countries are experiencing increasingly extreme weather conditions such as droughts, storms and flooding. This coupled with energy prices reaching a record high and cooling requirements estimated to be responsible for up to 40% of total data centre energy costs, it is not surprising that the data centre industry is concerned about the impact extreme weather could potentially have on cooling operations.
Data centres already generate a considerable amount of heat, and prolonged periods of extreme hot weather mean that cooling systems will need to work harder and consume more energy (at a significant cost!) to ensure the infrastructure remains at an acceptable temperature. Should long periods of sustained heat become more prevalent, many legacy data centres may even find that the strain of optimal cooling surpasses the limits of the design of the site leading to equipment failure.
Similarly, a long-term increase in ambient temperature resulting in a huge increase in the demand for power could put a strain on the power grid, leading to potential power outages – and if these outages are prolonged, it can put a significant strain on backup systems, potentially impacting the cooling infrastructure and jeopardising the temperature regulation within the data centre.
Extreme weather events are a threat to physical infrastructure such as power stations, electricity cables and the data centre buildings themselves. Data centres heavily rely on a stable and uninterrupted power supply and use backup power systems to mitigate this risk.
Businesses that are concerned about the physical and financial effects of extreme weather on their digital infrastructure should choose the location of their data centre wisely. Newer data centres will employ state-of-the-art monitoring systems that mitigate the risks associated with extreme weather events and many employ the latest, more energy-efficient cooling technologies – for example, redundant cooling systems, smart monitoring technology and backup power generators to ensure consistency of temperature in the event of power outages. Another example is direct liquid cooling which leverages the higher thermal transfer of water to more efficiently cool equipment.
Data centres located in countries with a cooler, natural ambient temperature are able to cool the sites at a significantly lower cost. The Nordic region in particular has a stable cool climate and is powered primarily by renewable, abundant and low-cost energy sources.
Given the severity of the increase in extreme weather conditions, it is becoming apparent that even the existing design guidance that most of the industry adopts may be too historic and fails to consider the rate of acceleration of these environmental changes. As such, design considerations need to be carefully considered and adapted to meet these trends and to ensure that the data centres continue to meet the exacting internal environmental conditions demanded while still performing at optimal efficiency levels.
Gordon Johnson, Senior CFD Manager, Subzero Engineering
Data centre growth continues throughout the globe, with the IDC (International Data Corporation) predicting that the collective sum of the world’s data will grow to 175ZB by 2025. Along with this growth comes the existing challenge of energy efficiency and sustainability for data centre operators. Now there’s a new challenge that needs to be planned for sooner than later, and that’s extreme weather conditions brought on in part by climate change.
Extreme weather patterns pose threats to servers, power and especially cooling operations, with the potential threat of long downtimes and huge financial costs. For example, power grid supply failure is typically solved by backup generators when power is lost, but these can only be used for a short period of time which may not be long enough when an extreme weather event is involved.
Besides the power grid, flooding can also create potential issues with servers and cooling operations necessary to keep the servers running 24/7. It’s not just storms that could increase our outages, droughts can also complicate access to water for cooling our data centres. It’s important that data centres, no matter where they’re located, have a plan in place to deal with these unexpected events since they have the potential to cut off both power and cooling for extended periods.
One particular area of concern continues to be reliably cooling the data centre brought on by climate changes. Just recently, two large data centres in the UK experienced cooling failures amid record heat temperatures in that country. Planning for cooling failure is crucial wherever the data centre is globally located, and recently this is especially throughout Europe as increasingly hot summers have now become the new normal. If extreme temperatures remain over prolonged periods, facilities may overheat and supplemental temporary cooling may be necessary.
It’s not just large, new data centres that need to be concerned. Older, legacy data centres typically are not as equipped as newer ones to deal with cooling failure due to weather and heat-related incidents. Many of these facilities have outgrown their critical cooling and power infrastructure and are having difficulties operating 24/7 as it is. They also tend to be more reactive and thus stand to lose the most. These especially need to prepare and perform regular extreme climate and weather risk assessments and have contingency cooling plans in place if and when needed.
Climate change will continue to impact data centre operations globally, including critical cooling operations. How we prepare now to deal with this may dictate whether our data centre is operational in the near future. Whether extreme weather results in storms and flooding, extended heat waves, droughts, etc, data centre operators need to plan now to avoid costly delays and operational downtimes in the future.
Jason Matteson, Director of Product Strategy, Iceotope
Last summer saw record-breaking heat waves across Europe and this summer the US is experiencing the same. Not to mention, the global average daily temperature broke record temperatures four days in a row during the first week of July 2023. What does this mean for data centre operators? When it comes to cooling operations, it’s safe to say that previously held ideas on the viability of certain cooling solutions will not withstand the reality of today’s climate crisis.
When temperatures reach 40C, the amount of energy required to cool a data centre increases as the ambient temperature goes up. This poses a particular issue for traditional air-cooled data centres which can see their energy usage – and costs – skyrocket. Air cooling has long been the default standard within the industry, but its limits are quickly being reached. Accelerating chip thermal design power, as well as the rise of Artificial Intelligence, high-performance computing and Machine Learning, are all pushing air cooling beyond its limitations.
These systems often rely on the use of evaporative cooling towers due to their heat rejection capabilities. However, this consumes a vast amount of water which is problematic and unsustainable given the environmental pressures we face.
This is where liquid cooling technologies come in. There are a variety of technologies available on the market, each with its own set of pros and cons. Direct-to-chip offers the highest cooling performance at chip level but still requires air cooling. Tank immersion solutions require a complete rethink of data centre design. Precision liquid cooling, on the other hand, removes nearly 100% of the heat generated by the electronic components of a server, while reducing energy use by up to 40% and water consumption by 100%.
It does this through a precise delivery of dielectric fluid to the hottest components of the server. The solution uses the same rack-based architecture as air-cooled systems and fits to existing deployed infrastructure. This allows for greater flexibility in designing IT solutions as there are no hotspots to slow down performance, no wasted physical space on unnecessary cooling infrastructure and minimal need for water consumption.
Extreme weather events or not, data centre infrastructure being planned must provide more energy efficiency and sustainability benefits and contribute to the corporate bottom line. Reduction in non-IT costs can save the industry hundreds of millions per year and being more energy efficient will enable expansion to brownfield sites, enabling businesses to reuse existing real estate assets. Precision liquid cooling offers a path forward for data centre operators as they tackle the challenges of today’s climate crisis.
Steve Hayward, Vice President European Operations, CyrusOne
With Italy issuing red alerts for 16 cities and temperatures in Death Valley, USA recording 53.9C – just under the hottest temperature ever reliably recorded on earth (56.7C) – this summer serves as a stark reminder of the impact of climate change on our weather patterns. The Met Office predicts that by 2070 winters will be between 1 and 4.5C warmer and up to 30% wetter, and summers will be between 1 and 6C warmer and up to 60% drier.
Every industry has a responsibility towards sustainability and the climate, and data centres are no exception. As detailed in CyrusOne’s 2023 Sustainability Report, we have shifted our operations across Europe to 100% renewable energy and are committed to reducing our carbon emissions further.
It is important to acknowledge that increasing temperatures are already influencing cooling operations. An agile and proactive approach to mitigation against extreme weather events must be prioritised. Operators that take no action in relation to cooling operations expose themselves and their customer base to elevated prospects of server downtime and service impairment.
So, it is clear, an inability to counteract extreme high temperatures places service at risk, but it is equally important to consider the converse. Periods of extreme low temperature present the same service level risk if not proactively managed. Systems that are not correctly winterised could face component failure and pipe burst for example, so it is crucial that when considering the chance of extreme weather all seasonal risks are fully considered and mitigated.
As such, data centre operators must prioritise and maintain a proactive approach to maintaining the data hall environmental conditions whatever the weather. It is common practice to utilise ASHRAE industry standards for the heating, ventilating, air conditioning and refrigerating (HVAC&R) industry when designing and selecting cooling capacity. These standards provide expected maximum and minimum temperatures but do not account for the extremes in real time nor can they accurately predict future changes.
In acknowledging the already present influence of increasing temperatures and understanding that action is needed, what does that action look like? Examples may include increased utilisation of the building management system and onsite telemetry in real time, continuous tracking of local short- and long-range weather forecasting and ensuring well-documented trigger points exist.
These points may identify a need for greater servicing, enhanced operating procedures or headcount increases. Deployments of additional temporary cooling may be required to mitigate the high-risk summer period. Ensuring that deployments of this nature are ready for use and that the operational team is well-trained and drilled in executing its start-up is vital to maintaining uptime. The final point is to ensure that these actions and the data collated are captured and form a basis for future designs and deployments reducing future risk.
If data centre operators don’t take the time to be agile and invest in improving cooling capacity and adaptation, they are increasing risk to themselves and their customers. But with proactive planning, continuous monitoring, analysis of how extreme weather conditions might affect performance and willingness to adapt to change, they can maximise uptime and deliver on their promises, now and in the future.
Alex Brew, Regional Director, Northern Europe at Vertiv
In 2022, record-breaking heatwaves made headlines, straining economies around the globe. There aren’t any signs of temperatures cooling down in the immediate future. On June 27 2023, the Met Office reported that the UK was set to have the hottest June on record.
As well as the impact on human lives, extreme heat is taking a major toll on data centre operations and infrastructure, from outages and disruptions to reduced production at power plants, to concerns over water usage in drought-affected areas across the world. Navigating these challenges is an industry priority.
One thing is clear, data centres need to be prepared for these extreme weather conditions to be more commonplace, so adapting to this inevitability is crucial. Organisations are becoming more aware of the challenges that extreme heat brings and are rightly looking at how they can mitigate the resulting risks when it comes to their data centre operations. If operators are concerned that conditions on site may exceed the original external design criteria for their critical infrastructure, to keep data centres cool and help ensure data centre availability, here are some recommended considerations:
- For refrigerant-based cooling systems, depending on the data hall load, consider running as many units as possible to reduce the load on each cooling unit. This keeps the refrigerant pressures lower so there is no risk of trips on high pressure with the high ambient temperature.
- Schedule a regular maintenance plan. For refrigerant-based cooling units, quarterly or more frequent cleaning of the condenser coils should be carried out. The lack of proper and regular cleaning of condenser coils can cause a reduction in the overall heat capabilities of the indoor and outdoor units, which can impact capacity under extreme ambient temperatures.
As the planet continues to warm and the climate crisis intensifies, extreme heat events will become even more commonplace. Short-term actions will need to be accompanied by more long-term, systemic changes in the industry.
Other long-term considerations and best practices include:
- Consider different types of cooling systems, including immersion cooling and evaporative-free cooling systems, designed to deliver reliability and efficiency in high-temperature environments.
- If relying on solar cells as an energy source, account for the reduction of the efficacy of solar cells as ambient temperatures increase.
- Rather than solely relying on historical data to determine data centre capacity – data that fails to account for today’s extreme temperatures – use predictive modelling.
James Kirkwood, Head of Technical Sales, EkkoSense
While the Southern European Cerberus and Charon heatwaves, fortunately, missed the UK, our wet July can’t mask the fact that European and UK temperatures keep on getting hotter. Indeed, the Met Office has reported that June temperatures in the UK were the warmest on record.
Not surprisingly, higher temperatures lead to questions about how they are likely to impact data centre cooling operations. However, while the UK worries about its data centres overheating, the rest of the world seems to manage just fine at these higher temperatures. In much of the US, for example, data centres are specified to operate in ambient temperatures of up to 120F, or almost 49C, while in countries such as Saudi Arabia facilities are designed to run in a climate where temperatures can end up even higher. The right levels of cooling are applied to enable operations to run at an optimum temperature – regardless of what’s going on outside.
So, why do extreme weather conditions create such a concern for some UK data centre cooling operations? The answer is they don’t need to, but the combination of legacy design issues, the age of many data centres and the risk of further 40C plus events can certainly cause issues if left unaddressed.
The good news is that it doesn’t have to be this way. At EkkoSense, we monitor the real-time thermal status of thousands of racks across facilities of all sizes. During last year’s UK heatwave that saw record temperatures, we detected hardly any significant thermal performance anomalies in data centres that had already benefited from cooling optimisation.
But the issues that did impact performance for some operators invariably stemmed from thermal performance failures outside of the data centre compounded by lack of optimisation within the data centre. This shouldn’t cause too much surprise as much external UK cooling infrastructure is over 10-15 years old and was never designed to perform in conditions of 35C or above. Whether it’s poorly maintained heat exchangers or compressors tripping on high-pressure and causing refrigeration systems to switch off, issues with your external cooling infrastructure can quickly lead to overheating within data centres.
This can escalate quickly in extreme conditions, with multiple failures potentially impacting data centre thermal performance in under an hour. Traditional BMS systems that alert on an SLA breach, or a hard fault are often unable to pick up these problems in time. Once they have alerted you, you have often already lost resilience.
That’s why it’s essential to have a true real time, AI-powered view into exactly what’s happening across your estate from a temperature, power and space perspective. This level of insight helps you to determine exactly where you might have a problem, gives your engineers the information and time they need to resolve potential cooling issues before they become more critical and make sure they can closer observe performance, so the problem does not re-occur. And, crucially for estate managers, the insights and analytics delivered by the platform also enable them to focus on their future investments and upgrades.
