How Can Data Centres Navigate Sustainability and Energy Efficiency in the Future?

In an era where the digital landscape is expanding at an unprecedented rate, the sustainability and efficiency of data centres have become focal points of discussion within the tech community. As these facilities burgeon in number, especially in regions like Southeast Asia, the environmental implications and challenges they pose cannot be overlooked. This article delves into these challenges, alongside offering insights into innovative cooling strategies and design improvements that can pave the way for more sustainable and efficient data centre operations. The insights provided herein come from Andrew Harrop, Director of Regional Sales Enablement Center, Armstrong Fluid Technology, whose expertise sheds light on both the current state and future potential for data centres in addressing their environmental footprint while meeting the growing demand for data processing and storage.

Environmental Challenges Posed by the Rapid Growth of Data Facilities in Emerging Markets, Particularly in Southeast Asia

Data centres in Southeast Asia (SEA) face three main environmental challenges: high humidity and temperatures, risk of natural disasters, and limited use of renewable energy. 

The first challenge stems from the location of the region. Data centres run on a lot of power to process, store, and manage information. This power translates into heat generation, which leads to the increase in their operational temperature and exposing the equipment to possible damage. Traditional methods such as air cooling – which uses surrounding air to cool the data centre – is challenging to adopt or maintain due to the high humidity and temperatures of the region’s tropical climate, so other cooling methods are required to reduce operational temperature to industry-approved levels. One such method is liquid cooling, which is steadily gaining traction as a new standard. 

Separately, SEA is highly prone to natural disasters. SEA countries along the Pacific Ring of Fire, such as Indonesia and the Philippines, are in seismic zones and vulnerable to earthquakes. Major facilities such as data centres and colocations are greatly threatened by these natural disasters as the possible damage can cause long-term disruptions. Hence, data centres in this region need to develop disaster recovery plans to ensure continuous operations in the face of natural disasters.  

Finally, while many countries in the region have kick-started their energy transition, the extent of progress and opportunities vary. Take for instance, the Philippines, a country positioning itself as an emerging data centre hotspot, has a data centre market predicted to grow at a CAGR of 11.2% from 2022 to 2027 according to a report by Mordor Intelligence. However, the trajectory of the market is met by a slow growth in the use of renewables. This could be attributed to the cost of installing and maintaining renewable energy technologies such as solar and wind power, which is significantly higher than electricity generated by fossil fuels in the country. Realising the country’s ambitious goal for 35% of the nation’s energy mix to be renewable by 2030 requires the adoption of more advanced technology from industries including data centres. 

Strategies for Reducing Energy Consumption in Cooling Systems of Data Centres in the APAC Region Amidst Tropical Climate and High Humidity Challenges

As data centres in APAC are unable to rely solely on air cooling, they need to explore other more energy-efficient cooling systems. For instance, direct-to-chip cooling – where the liquid coolant is delivered directly to the hotter parts of servers to effectively extract heat from the source – can achieve up to a 45% reduction in cooling power consumption compared to air cooling. This is done through selectively cooling high-power components such as the Central Processing Unit (CPU). Direct-to-chip cooling can decrease airflow needed by up to 90% compared to air cooling. 

Separately, in countries like Thailand, cooling technologies commonly used by data centres include direct expansion, chemical-based, and water chillers. Recently, there have also been talks to use regasified liquid natural gas (LNG) to produce electricity and chilled water for data centre cooling in the country. 

Another method that we are familiar with is intelligent cooling systems, which leverage real-time data to dynamically adjust cooling equipment such as pumps based on the cooling requirements of the data centre. In a region vulnerable to changes in temperature and humidity, intelligent cooling systems are crucial to increase the reliability and resilience of data centres.

At Armstrong, our Design Envelope technology is a demand-based intelligent control solution that models equipment and system behaviour, monitors actual system conditions, and dynamically adjusts equipment operation to match system demand. The Design Envelope 4300 Pumps with Suction Guides & Flo-Trex Vales (SG/FTV) is a pipe-mounted UL 778 pumping unit with intelligent controls that directly saves energy by combining water-cooling – a method most utilised by data centres in APAC – with automation. Among our customers that have adopted this technology is Kakao, one of Korea’s largest internet portals. We installed 16 units of Design Envelope 4300 pumps with SG/FTV at Kakao’s data centre. Separately, we also provided 16 units of this solution to the data centre of a global hyperscale cloud service provider (CSP) in Indonesia. 

The Role of Data Centre Design in Achieving Energy-Efficient and Sustainable Cooling Solutions – Opportunities for Improvement

The design of data centres is important to enable operators to manage the increasing operational demands while adhering to environmental challenges and requirements. One way to ensure that the design caters to their needs is to work with technology providers to explore innovative solutions that will help develop more energy-efficient and sustainable cooling methods, such as Artificial Intelligence (AI).

AI can increase energy efficiency through predictive maintenance and automated processes. AI-powered algorithms can analyse past power consumption data and predict cooling demands, while AI-enabled tools can provide continuous analysis of sensor data, identifying optimal operating conditions to maintain, and thus allow operators to increase their power usage effectiveness (PUE). 

We are also seeing the emergent of more green data centres. Also known as sustainable data centres, they are facilities specially designed to minimise environmental impact while continuing to manage data efficiently and reliably. In terms of cooling efficiency, green data centres employ a series of methods that may include water-based cooling systems, optimised airflow management, liquid immersion cooling, and more. 

In Singapore, the government has been encouraging the development of green data centres through the country’s Green Mark scheme for data centres. The Building and Construction Authority (BCA) and the Infocomm Media Development Authority (IMDA) rolled out this scheme to encourage the adoption of more energy-efficient operation, management, and design of data centres. 

Expectations and Demands for Data Centre Sustainability in the Future: Insights from a Technology-Savvy Generation

With the rise of a technology-savvy generation, we will see greater volumes of data that would need to be processed. And due to the large amount that will be collected and used, the industry will expect higher standards of cybersecurity. At the same time, as the number of servers and data centres continue to grow, more cooling solutions will be expected, especially in countries in the tropics. All these, added with the increased demand for sustainability, put a lot of pressure on operators. 

In fact, according to a 2022 JLL study, sustainability and social responsibility will be the top priority for data centre operators in APAC over the next two years, preceding traditional productivity and efficiency metrics. It also identified leading innovations for more sustainable data centre operations, including the use of airflow management solutions such as hot-cold aisles to improve cooling efficiency, as well as the use of Artificial intelligence (AI) or machine learning (ML) for optimised temperature control.

A technology-first world will also require continuous investment and time into product innovation, for the improvement of servers and equipment that will minimise downtimes. We can expect more Tier 4 data centres, which are the most complex and feature concurrent cooling, electrical, and control circuits that allow for redundant components. While Tier 4 data centres offer better reliability, they require more equipment and infrastructure than the lower-tier data centres, leading to higher energy consumption and creating a need for more efficient cooling systems.  

Balancing the Increasing Demand for Data Processing and Storage with the Reduction of Environmental Footprint: Strategies for Data Centre Operators

The use of offsite-built modular systems can help data centre operators to meet the ever-increasing demand for data processing and storage while making strides in reducing their environmental footprint.

Deploying critical data computes in phases over time allows for the scaling up of modular systems when processing higher loads of data. They can also be designed and constructed to occupy a smaller footprint compared to traditional methods. In the process, the scalability also eliminates the need for overprovisioning capacity, minimising any unnecessary energy consumption. 

Additionally, these modular systems can help to ensure data centre redundancy in managing and processing data. Implementing a well-planned redundancy design is crucial, where components in the data centre are duplicated so that in the event that one fails, the other equipment within is not affected. This allows the data centre to maintain uptime and recover more quickly from disruptions. With the increasing demand for data at all hours, companies need to have reliable systems that ensure redundancy to stay competitive. 

Finally, the standardisation in offsite-built modular systems allows for optimal energy performance. When coupled with standardised control solutions for in-process cooling and water flow, they can reduce power consumption significantly.