By Aldo Santamaria | October 3, 2025
Inside the Numbers: Project Finance for Data Centers
They power our digital world — but what does it take to model “the cloud”?
Data centers are the unseen backbone of our digital lives, supporting everything from streaming and banking apps to generative AI. Every time we mention “the cloud”, we’re really talking about large-scale facilities full of servers that store and process our online information. The global data center industry is already worth over $350 billion, with projections suggesting it could reach $1 trillion within the next decade.
But behind every glowing server hall lies a complex web of financial decisions. Many big companies lease space in these facilities — through colocation agreements — to store their IT equipment, while data center teams handle power, cooling and maintenance. Think of data centers as the physical places where the internet and digital economy live.
As a result, data centers have emerged as a unique investment class — part real estate, part infrastructure — where financial modeling plays an essential role in understanding their mix of long-term contracted revenues and high upfront costs.
How Project Finance Applies to Data Centers
Understanding project finance for data centers starts with how investors and lenders view them. At first glance, these projects may seem closer to real estate: they already have everything needed to lease space to tenants. But their structure sometimes aligns more closely with infrastructure deals.
Typical real estate loans are collateral-based. For example, if a homeowner cannot repay a mortgage, the lender can foreclose and sell the property to recover funds. By contrast, project finance is cash-flow based: lenders focus on the project’s ability to generate income rather than the value of the collateral.
Data centers lease secure, high-performance server space for companies to store and process digital information
For instance, the lender for an airport project may not be able to seize the building itself but has rights over passenger fees and landing charges. Similarly, data centers’ predictable revenues from long-term colocation agreements make them strong candidates for project finance modeling.
Key Financial Drivers in Data Center Project Finance
Data centers generate steady revenue streams by leasing space, power, cooling, connectivity and security to clients. They also earn from ancillary services, that is, the extras that keep servers running smoothly. This includes “remote hands”, “smart hands” and “cross connects”, which we will cover in our data center video series coming soon.
What About Data Center Costs?
Development of data centers requires significant capital expenditure for site selection, constructing the facility and securing power supply. Some projects even include on-site plants to meet demand. These aspects reflect infrastructure-level planning, which emphasizes how data centers straddle the line between different asset classes.
In fact, power remains the largest component of a data center’s costs, sometimes accounting for more than half of total operating costs. Facilities invest heavily in redundant sources, like battery or diesel backup systems, along with Uninterruptable Power Supplies (UPS) to ensure continuous operation in case of grid failures.
Cooling adds another major energy burden. Whether air- or liquid-based, these systems prevent overheating and ensure performance. Other costs include skilled staff, property taxes, insurance, and high-capacity fiber optic connectivity. Despite these complex requirements, the financial risk of fluctuating energy prices is largely mitigated, since costs are passed on to clients with a small markup.
Building a Data Center Financial Model
Financial modeling for data centers focuses on capacity, utilization, and power consumption. Rates are typically measured in dollars per kilowatt of IT load (kW/IT) — meaning clients are billed by power usage rather than floor space or number of servers. Naturally, more servers equal more power.
A standard data center rack might draw between 2 and 5 kW of power, roughly equivalent to running 40 to 100 light bulbs at once. But AI-ready, high-density racks can reach 20 or even 50 kW per rack. Your model should reflect how these power needs drive both capital expenditures (CapEx) and operational expenditures (OpEx).
All these variables must be considered to design a robust data center financial model — one that captures energy costs, lease terms, and service-based revenues.
Data centers earn additional revenue through ancillary services like “remote hands” and “smart hands”, where on-site technicians maintain client equipment.
Training
If you want to master project finance modeling, Pivotal180 offers hands-on courses covering infrastructure, renewable energy, and industries like battery storage and mining. Whether you’re a student building job-ready skills or a professional refining your expertise, you’ll learn to build full financial models from scratch and understand how changes in scenarios impact real-world projects.
Enroll in a Pivotal180 course! We offer a range of training programs for modelers of different backgrounds and experience levels:
Renewable Energy Project Finance Modeling
Introduction to Project Finance Modeling
Financial Modeling Fundamentals
Tax Equity & Hybrid Financial Modeling
Have questions? Reach out anytime at [email protected]
