Data center boom

A digital economy requires a robust digital infrastructure, which includes data centers. Data centers, like any other type of property, pose both opportunities and challenges for the professional property manager. As businesses increasingly rely on robust data solutions, the intersection of real estate and technology allows property managers to leverage their expertise to capitalize on this growing sector. However, these properties also have challenges that property managers must be aware of. 

A data center, generally, is a facility used to house computer systems and associated components, such as telecommunications, storage systems, and network infrastructure. It provides critical infrastructure to process, store, and disseminate data. Data centers date back to the 1940s. An early example would be the U.S. military’s Electrical Numerical Integrator and Computer (ENIAC), completed in 1945 at the University of Pennsylvania, which required dedicated space to house its massive machines. 

U.S. versus Europe

There are slight variations in the definition of a data center between the U.S. and Europe. Chris Petosa, senior vice president, U.S. Data Center Management at Colliers International, AMO^®, notes that in the U.S., a data center is typically defined by its size (often measured in square footage or power consumption), redundancy capabilities, and tier classification (tier 1 to tier 4 based on uptime and fault tolerance). 

Robert Fuller, CPM^®

Robert Fuller, CPM^®, senior facility manager at CBRE, Inc., AMO^®, said that in the U.S., data centers are defined by their megawatts (MW) of power available or that can be consumed. Typically, small-size data centers have a power consumption of 20MW, mid-size ones are 50MW to 70MW, and larger ones are 100MW to 250MW.

In Europe, data centers are often defined with a greater focus on energy efficiency and sustainability standards, driven by regulations like the EU Data Center Code of Conduct. European data centers tend to focus more on carbon neutrality and compliance with green energy policies.

Data center stats

The U.S. is home to about 2,600 data centers, generating between $10 and $15 billion in revenue annually. Worldwide, there are more than 8,000 data centers; the global data center market is valued at about $200 billion. According to Data Center Frontier, in 2024, the four largest “hyperscalers”—companies that deploy massive networks of infrastructure to accommodate many thousands of cloud service users— include Amazon, Microsoft, Alphabet, and Meta. These companies spend a combined $50 billion per quarter on digital infrastructure, including data centers.

Fuller pointed to a 2017 report by Forbes highlighting that U.S. data centers alone use more than 90 billion kilowatt-hours of electricity a year, requiring roughly 34 giant (500-megawatt) coal-powered plants. This consumption is anticipated to double every four years.
A 2024 report from the Electric Power Research Institute (EPRI) estimates that data centers could grow to consume up to 9% of U.S. electricity generation annually by 2030, up from
4% in 2023.

According to Fuller, the most commonly used energy efficiency metric of data center energy efficiency is Power Usage Effectiveness (PUE), calculated as the ratio of total power entering the data center divided by the power used by IT equipment.

Data center categories

Chris Petosa

Petosa said that the type of data center will affect how the property is managed.

Cloud data centers are operated by cloud service providers like AWS, Google Cloud, or Microsoft Azure. Management focuses on large-scale automation, high redundancy, and scalability, often dealing with multi-tenancy and rapid resource allocation.

Managed data centers are operated by third parties for clients who lease the infrastructure. Property management here involves providing high availability and regular performance updates, focusing on security and uptime.

Enterprise data centers are built and operated by a single organization (such as a bank or hospital) for internal use. Property management in these locations is often tied closely to the organization’s specific security, compliance, and performance needs.

Critical data center requirements

Fuller says that the typical data center has an occupied administrative area of the facility, although there are “dark data centers,” which are unoccupied and maintained remotely. In tandem with the engineering staff, a data center facility manager is responsible for providing continuous power and cooling to the data center and maintaining 100% uptime. Losing power could be catastrophic to an organization, industry, or government, says Fuller, which is why data centers, hospitals, and power plants are dubbed “mission-critical facilities.” 

Fuller added that their mission-critical facility status is also based on their impact on the economy if any facilities were to lose power and, in effect, go down. For example, when an international airport is shut down because of extreme weather or a computer glitch, flights are canceled, and millions of dollars are lost.

Substantial penalties or forfeiture of revenue can be imposed on a data center operator who fails to meet the KPIs (key performance indices) agreed upon with the client of the managed data center.

Data center PM duties and skill sets

Fuller says that the data center is a uniquely complex and dynamic facility requiring the facility manager to possess more than a cursory knowledge of IT infrastructure and the mechanical, electrical, and plumbing (MEP) equipment in the facility to be successful and grow. An effective property manager of data centers must carry out these duties:

• Operations oversight: Ensuring all systems (cooling, power, network) are running optimally.
• Maintenance management: Overseeing preventive and corrective maintenance of equipment like generators, HVAC, and backup systems.
• Vendor management: Coordinating with third-party service providers (security, janitorial, maintenance).
• Compliance and reporting: Ensuring the facility meets regulatory requirements, like Service Organization Control (SOC) 2, a cybersecurity compliance and reporting framework, ISO certifications, and other industry standards.
• Energy efficiency: Managing energy consumption, including ongoing analysis of utility energy usage and measurement of electrical and water usage, as well as managing sustainability initiatives, especially in Europe.
• Risk management: Preparing for emergencies and ensuring high availability and disaster recovery processes.
• Financial budgeting and asset management: Managing both planned and unplanned work efforts, cost reporting, and measuring.
• Environmental reporting: Both Environmental Protection Agency (EPA) and state regulations can be highly restrictive. Data centers face financial penalties and reputational risk if the operator adversely affects the environment in which they operate.

Data center property managers should understand electrical, mechanical, and IT systems (e.g., UPS, cooling systems, fire suppression, network architecture). In addition, they should have working knowledge of physical and cybersecurity best practices; disaster recovery planning and business continuity practices; and energy efficiency strategies, like PUE optimization.

Specific data center certifications like Certified Data Center Professional (CDCP), Data Center Operations Manager (DCOM), and BICSI certifications are valuable, as well as the Data Center Technician (DCT) designation issued by the Data Center Dynamics Academy based in the UK.

Adaptive reuse

Many communities are looking into the possibilities of adaptive reuse, taking abandoned or underused properties and converting them to another more productive use. Petosa highlights the advantages and disadvantages of converting older properties into data centers. Advantages include lower upfront costs, strategic locations in urban areas, and access to existing infrastructure such as power and fiber. Disadvantages include the cost of retrofitting for power density, cooling, and security. Zoning issues or limitations of the original design can also pose challenges.

Fuller said that urban areas with local technical schools, universities, and colleges have the advantage of a supportive workforce to draw from with knowledge and skills to work in a data center environment.

In an unpublished white paper, Petosa notes that converting a commercial real estate building into a data center involves considering various factors to ensure that the facility meets the technical, operational, and environmental requirements of modern data center infrastructure. The considerations include:

• Power infrastructure (electrical capacity, redundancy, and power distribution)
• Cooling and HVAC systems (cooling capacity and
  redundancy in cooling)
• Space and layout (floor load capacity, ceiling height, and column spacing)
• Security (physical security and fire suppression)
• Connectivity (network infrastructure and carrier neutrality)
• Compliance (zoning and building codes and
  environmental compliance)
• Scalability (room for expansion and modularity)
• Accessibility (proximity to network hubs)
• Environmental considerations (including green
  building features)
• Stranded power opportunity

Petosa also notes that when there’s stranded power in an office building (power capacity that is allocated but not used), it represents an opportunity for the building owner or operator to explore alternative uses or strategies to leverage that excess capacity. The additional electrical capacity can be utilized for a data center.

Building owners and operators must work closely with utility providers, energy consultants, and relevant stakeholders to assess the feasibility of utilizing stranded power effectively and explore potential opportunities for enhancing energy efficiency and sustainability.

Evolution of the data center

Petosa notes that as data centers move closer to end users for faster processing, property managers may need to handle smaller, more distributed data centers. There will be increasing pressure on managers to meet sustainability goals through green energy, efficient cooling, and low-carbon operations. Also, property managers should expect more AI-driven automation in data centers for predictive maintenance and monitoring, which will reduce downtime and increase efficiency.