Data Centres Explained: Design, Cooling, Procurement and Cost
Updated: 6 days ago
A data centre is a dedicated space, ranging from a single room to an entire building or facility, that houses IT infrastructure for running and delivering applications and services. These centres also store and manage the data associated with those applications and services.
The concept of data centres has evolved significantly over the past 75 years. As computers became increasingly compact and efficient, especially with the rise of microcomputers in the 1990s, the need for large-scale hardware diminished. This shift contributed to the widespread adoption of the term “data centre”.
The 2000s brought a major transformation with the emergence of cloud computing, which disrupted traditional data centre models. Cloud technology enabled organisations to access computing services on demand via the internet, using a flexible, pay-as-you-go model that allowed rapid scaling.
The first hyperscale data centre, a facility designed for massive scalability, was launched by Google in 2006.
Looking ahead, the data centre sector is projected to grow at a rate of 10% per annum through to 2030, driven by ongoing digital transformation and increasing demand for cloud-based services.
Are There Differing Types of Data Centres?
Yes, there are several types, each serving different organisational needs:
On-Premises Data Centres
All IT infrastructure and data are hosted on-site, giving organisations full control over operations and information security. This model is typically preferred by companies with strict compliance requirements or sensitive data needs.
Public Cloud Data Centres
These facilities host IT infrastructure that is shared across multiple customers. The largest among them, known as hyperscale data centres, are operated by major cloud service providers and offer highly scalable, on-demand resources.
Managed Data Centres and Co-Location Facilities
In a managed data centre, clients lease dedicated hardware from the provider, who also handles system administration, monitoring, and management. In contrast, a co-location facility allows clients to install and operate their own equipment within a provider’s premises. Clients retain full control and responsibility for their infrastructure. Both options are commonly used for housing backup systems or supporting hybrid IT strategies.
What Defines Modern Data Centre Design?
Modern data centres are increasingly built on cloud architecture, where physical resources such as computers, storage, and networking are virtualised. These virtualised resources can be pooled and dynamically allocated across multiple applications and workloads, often without the need for manual intervention.
This approach enables organisations to optimise infrastructure for each application without requiring physical modifications. As a result, cloud-based and virtualised designs are expected to become the prevailing standard in data centre development.
The combination of cloud architecture and infrastructure virtualisation offers several key advantages:
Optimal utilisation of computing, storage, and network resources
Rapid deployment of applications and services
Seamless scalability
Flexible solution offerings
Support for cloud-native development
What Is Data Centre Infrastructure?
Data centre infrastructure encompasses all the key components that enable the delivery of applications, services, and data. These elements work together to ensure performance, reliability, and scalability:
Servers
Servers are the core computing units that deliver applications, services, and data to end-user devices. A range of server types is available, selected based on workload demands, space constraints, power availability, and cost considerations.
Storage Systems
Modern data centres typically utilise a mix of storage configurations, including direct-attached storage (DAS), dedicated local servers, and remote network-based storage. A combination of all three is often employed to support different operational needs.
Networking
Networking refers to the physical and logical arrangement of devices within the data centre. This includes the internal connections between servers and other components, as well as the systems that manage traffic between the data centre and external clients.
Power Supply and Cable Management
Data centres must always be on. This requires dual source power supplies and battery powered uninterruptible power supplies (UPS). Proper cable management is also essential for safety, efficiency, and future scalability.
Redundancy and Disaster Recovery
High resiliency is critical in data centre environments. This is achieved through system redundancy, back-up power, and cooling solutions. The Uptime Institute classifies data centre resiliency using a four-tier system:
Tier I: Basic redundancy, including UPS and 24/7 cooling, suitable for an office setting.
Tier II: Adds extra redundant power and cooling systems for improved protection against disruptions.
Tier III: Adds further redundancy components so the facility requires no shutdown when performing maintenance and replacing equipment.
Tier IV: Offers full fault tolerance, with completely isolated and redundant systems to ensure uninterrupted operation even during equipment failure.
Environmental Controls
Environmental conditions within a data centre must be carefully managed to prevent hardware damage and downtime:
Data centres employ cooling to keep servers and other hardware within proper operating temperatures. Mostly cooling is air conditioning, computer room air conditioning (CRAC) which targets an entire data hall or specific rows or racks of servers.
Humidity control ensures levels are kept within safe ranges using CRAC systems, ventilation, and sensors, neither too high nor too low.
Static electricity is monitored and safely discharged to protect sensitive equipment.
Fire detection and prevention systems are tailored to the data centre’s unique contents, balancing safety with equipment preservation.
What About Security and Management?
Security in data centres extends far beyond traditional measures such as CCTV, access control, and on-site personnel. It also includes robust cybersecurity to protect data, applications, and processes from digital threats.
Effective data centre management involves a range of tasks and tools designed to maintain operational integrity. This includes physical upkeep, system monitoring, and the timely upgrading of both software and hardware. A variety of software platforms are available to support these functions, streamlining management and enhancing security.
Are Green Solutions Needed Today?
As data centre demand grows, particularly due to AI and other data-intensive technologies, sustainability has become a priority. Some estimates predict a 160% increase in demand, making energy efficiency more critical than ever.
Green certifications and sustainability benchmarks from recognised environmental organisations are expected to play an increasingly important role in future data centre development.
What Is Liquid Cooling?
According to a recent report, soaring artificial intelligence workloads are pushing traditional air-cooling systems to their limits. To address this challenge, data centre operators are turning to a more efficient solution: liquid cooling. Compared with conventional air cooling, liquid cooling offers benefits such as higher energy efficiency, a smaller footprint, enhanced server reliability, lower noise levels, and more.
However, there are some drawbacks, including higher initial investment and the need for new operational and maintenance skill sets. A variety of liquid cooling technologies are available on the market, each offering different levels of energy efficiency, cooling capacity, and complexity in implementation and maintenance. These include rear-door heat exchangers, direct-to-chip cooling, and immersion cooling, among others.
A recent study found that liquid cooling can reduce total data centre energy consumption by approximately 10% compared to fully air-cooled facilities.
How Are Data Centres Procured?
Data centres typically consist of two key components: the shell and core, and the technical fit-out of the data halls.
The shell and core is, in many respects, a conventional building but with enhanced features to support the redundancy and resilience required by operational data halls. These enhancements primarily relate to power infrastructure, with dual power sources being a standard requirement. This involves two separate grid connections through two substations, leading to duplication of panels, switchgear, and associated systems.
The technical fit-out of the data halls is often carried out progressively, based on user demand, either through purchase or lease agreements. Commonly, one or two halls are fitted out alongside the shell and core, with additional halls left as shell space for future fit-out.
The configuration of the data centre depends on site constraints. It may be a single-storey, warehouse-style facility, or a multi-level building. In the latter case, vertical transportation becomes a key consideration and can significantly impact project cost.
Typical packaging in data centre procurement includes:
Piling and advance works
Main contract works: Usually delivered by one main contractor responsible for the full scope. However, the use of a separate MEP contractor is also possible. Data centre experience is essential for all parties involved.
Long Lead Equipment (LLE): These are highly specialised components, typically sourced
from a limited pool of suppliers, with indicative lead times:
Power generation equipment: Approximately 6 months
MV switchgear: Approximately 3 months
Chillers: Approximately 7 months
CRAC and CRAH units: Approximately 5 months
Power skids: Approximately 5 months
Fan wall units: Approximately 6 months
Security systems: Approximately 3 months
Lifts: Approximately 6 months
How Much Do Data Centres Cost?
A number of factors influence the cost of a data centre project:
Design Considerations
In addition to the configuration of the facility, there are numerous design variations particularly in MEP systems reflecting different approaches to environmental control, redundancy, and sustainability. As a result, a wide range of pricing is expected.
Contractor Selection Choosing an international contractor typically comes at a premium compared to engaging a local contractor.
One of the key advantages of working with OSA is our broad geographic footprint. Our network of associated offices across the region enables us to provide accurate, up-to-date cost data tailored to the local context.
OSA is ready to provide Quantity Surveying support for data centre developments, regardless of project location.
OSA’s Track Record in Data Centre Delivery
Due to the confidential nature of many data centre projects, only limited details can be shared publicly. However, OSA has supported a wide range of developments, covering shell and core works, technical fit-outs, or full turnkey solutions combining both.
These projects span from in-house data centres integrated within corporate offices to stand-alone facilities designed for data hall leasing or individual ownership. Typical capacities extend up to 40MW, delivered in structures ranging from single-storey footprints to multi-level buildings of eight storeys or more.
With our proven experience, regional reach, and the shared knowledge across our network, OSA is well positioned to support your future data centre project, no matter the scale or complexity.
Whether you’re planning your next project or exploring ideas, OSA is ready to support you every step of the way.
Reach out to Hearn & Hearn Consulting for tailored support and local expertise in the Philippines.
One Solution Asia is a leading construction consultancy specialising in construction cost and project management. Operating with over 300 professionals across Southeast and East Asia, our cooperating teams are from over 10 countries around the region including: Brunei, China (Mainland), Hong Kong SAR, India, Indonesia, Malaysia, Philippines, South Korea, Thailand and Vietnam.
As former partners and directors of the distinguished Davis Langdon & Seah in Asia, we have over 80+ years of experience, earning us the excellent reputation of delivering your goals and visions effectively and efficiently.
Data Centre Projects in the Philippines
As the Philippines member of the One Solution Asia network, Hearn & Hearn Consulting provides quantity surveying, construction cost management and project management services, supported by the shared knowledge and regional expertise of OSA’s teams.
For data centre projects in the Philippines, our team can support clients with cost planning, cost estimation, procurement and commercial management throughout the project lifecycle.
To discuss a data centre project in the Philippines, contact us at theteam@hhconsulting.com.ph or visit www.hhconsulting.com.ph.
Comments