Megawatts to Megabytes: Orrick’s Guide to Developing, Financing & Powering Data Centers
9 minute read | January.26.2026
How can businesses secure the right data center capacity when market dynamics favor speed, but operational requirements demand precision? This chapter examines practical strategies for evaluating and negotiating data center leases amid competing pressures of location constraints, infrastructure requirements and long-term scalability needs.
Read this excerpt from our guide Megawatts to Megabytes below or download the full PDF here
Leasing a data center, whether a fully operational site or through a build-to-suit arrangement, presents many considerations for prospective data center tenants and developers. For businesses looking to lease capacity, data centers can offer strategic advantages by providing flexibility, scalability, and cost efficiency over owning a data center outright.
Choosing the right facility will provide a foundation for future growth while minimizing risks. For developers, the leasing process requires an assessment into a number of factors such as availability of power, local regulations and political environment, environmental conditions and constraints, availability of incentives, network infrastructure, and proximity to potential users. In addition to navigating where to lease, choosing the right facility requires ensuring that the terms of the lease align with the tenant’s needs and adequately address challenges such as scalability, workforce availability, and cybersecurity risks.
Chapter 1.
In this chapter: Site selection factors including zoning regulations, water availability, regulatory compliance, tax incentives, proximity to users, disaster recovery planning, infrastructure quality and landlord capabilities.
Data centers are among the most energy-intensive infrastructure assets—and demand not just large volumes of power, but near-perfect reliability. The industry standard of “Five 9s” (99.999% availability) permits only about five minutes of downtime per year, making fully firm, uninterrupted grid supply essential. Most developers pursue grid interconnection capacity sufficient to always meet peak demand, typically through utility upgrades or direct transmission investment. Depending on region and scale, this process can take months or years for hyperscale projects. But that model is rapidly becoming unworkable.
Local zoning and land use regulations have emerged as critical determinants of data center project viability, as states and municipalities have taken differing positions on data center development. Many jurisdictions have welcomed data center development (e.g., Central Ohio) by adapting land use regulations specific to data centers. Data centers, however, typically have low employment density, high energy and water usage, and intense demand for connectivity. These characteristics often place them in tension with zoning codes that were designed for traditional commercial and industrial uses.
To accommodate data center development, jurisdictions that aim to welcome data center development may offer by-right use in utility-rich industrial zones, increase FAR and height limits in designated corridors, and remove outdated parking requirements. Jurisdictions that are restricting further development — particularly those in saturated markets — are unwilling to adapt existing land use regulations. If a party desires to acquire and then expand or redevelop an existing data center, it should ensure that such expansion or redevelopment will not conflict with existing zoning code or receive significant public opposition.
"Parcels with secure water access now command premium valuations, especially among data center users seeking to de-risk expansion pipelines."
Water availability has become an increasingly critical constraint in global data center development, rivaling power availability as a primary factor in site selection and long-term operational planning. Data centers may use a large amount of water in their cooling systems.
Water consumption in data centers has reached unprecedented levels. Virginia’s data centers alone consumed nearly two billion gallons in 2023, representing a 64% increase from 2019. Advances in technology can help manage water constraints and reduce the correlation between computing power and water use. Traditional systems — based on cooling towers and chillers — consume thousands of gallons per MW per day. In contrast, closed-loop systems target heat at the source and may require minimal water. However, advanced cooling technologies can increase capital costs, require technical expertise to manage performance risk, and sometimes involve higher power usage effectiveness.
Where data centers use large volumes of water, data center projects are under increasing scrutiny from water authorities and the public, especially in arid or semiarid regions like Arizona, Utah, and California. Conversely, there are jurisdictions, such as the Ashburn corridor in Virginia, that are investing in increased water utility capacity to accommodate further data center expansion.
Developers now face more complex environmental impact assessments and may try to acquire or lease water rights in advance to ensure adequate supply throughout a facility’s lifespan. For developers, particularly in regions with reliable and inexpensive water supplies, this represents a strategic opportunity. Parcels with secure water access now command premium valuations, especially among data center users seeking to de-risk expansion pipelines.
In addition to local zoning ordinances, many states have regulatory schemes that may impact the development or operation of data centers including consumer data privacy laws. Landlords and tenants should consider these requirements in determining the cost and/or burden of operating a data center in a particular jurisdiction.
To attract data center development, many countries, states, counties, and municipalities have offered tax or other incentives in connection with the initial construction or operation of the data center. Whether these ultimately benefit the landlord or tenant, the parties should explore all potential incentives in selecting a site for a new data center development.
Considering how quickly we receive our Google search results, it’s easy to forget that there is a physical aspect to the transfer of data. Parties should select locations for data centers that both minimize latency and optimize performance for end users.
Physical disaster risk will vary by location. The risks inherent with a site in California are different from those with a site in Iowa. Parties should confirm that the proposed data center has sufficient redundancy and failover systems in place to account for potential disasters.
Data center tenants should evaluate the quality and capacity of telecommunications infrastructure at the data center. Such infrastructure must ensure that the data center can deliver the computing power and network connectivity needed for the tenant’s business.
The landlord plays a crucial role in maintaining the data center and ensuring tenant satisfaction. Given the number of recent entrants into the data center market, tenants should adequately vet the landlord entity as well as its personnel. It’s important to confirm that the landlord has adequate experience in the operation of data centers of similar size and geography. With respect to build-to-suit projects, tenants should prioritize landlords with significant experience in managing data center construction and navigating supply chain issues.
If available, tenants should discuss other customer’s past experiences with the landlord. Unlike in a typical triple net lease, the landlord will provide significant services to the tenant throughout the term of the lease. Each tenant will have unique requirements as it relates to its intended use of a data center site, and those should be communicated at the outset of the lease discussions. Tenants should ensure that the landlord is willing to accommodate and adapt to the tenant’s specific needs and requirements.
Tenants should review the corporate structure and upper tier ownership of the landlord to assess whether any equity investors are competitors of the tenant and evaluate the landlord’s financial capacity to complete construction.
The landlord and tenant should consider the potential growth of the data center project or campus. This may be structured through multiple leases, options, or rights of first refusal/offer. Such expansion may require significant upgrades to the data center campus, including building or expanding substations to meet any increased power demands. Modular infrastructure of the initial data center development will help to accommodate future growth of the overall project.
Chapter 2.
In this chapter: Key lease provisions covering security protocols, assignment restrictions, acceptance testing, service level agreements, building management systems, planned maintenance, audit requirements and end-of-lease obligations.
Data centers are designed to support robust, scalable applications and services, often serving as the backbone for cloud providers and large enterprises. Leasing capacity in these facilities presents unique challenges and opportunities.
Security is paramount in data center operations to protect sensitive data and infrastructure.
In data center leases, assignment restrictions are typically imposed on both the landlord and tenant. The landlord should seek to prohibit direct assignments by tenants but should be prepared to permit certain indirect assignments (i.e., changes of control) of tenants. Due to the scope of services being provided by a data center landlord, data center leases often restrict the landlord’s ability to sell or otherwise transfer the data center (directly or indirectly) without the tenant’s consent. The lease should clearly outline the procedures for either party’s proposed assignment. Such procedures should include notice requirements, qualified transferees, prohibited assignees (e.g., competitors), financial requirements of a proposed assignee, etc.
Acceptance testing ensures that the leased space meets specified requirements before occupancy. The data center lease should clearly define the criteria for acceptance testing to ensure all data center systems meet specified operational standards. The lease should describe the various commissioning stages and the requirements at each stage. The lease should specify a timeline for acceptance testing and a detailed procedure for the resolution of any issues identified.
Throughout the term of the lease, the parties should maintain thorough documentation of any commissioning, acceptance testing procedures, and the results thereof.
SLAs outline the expected performance and reliability standards for data center services to be provided by (or on behalf of) the Landlord. The lease should clearly detail the various metrics for each of the data center components, including requirements for response times, temperature and cooling requirements, and acceptable data transmission delays.
The parties should establish clear, objective penalties for noncompliance with any SLA terms. These are often in the form of rent or service credits for the tenant’s benefit; however, repeated SLA violations may eventually result in an option to terminate all or a portion of the data center lease. The parties should regularly review and update SLAs to address any recurring issues or to reflect any changing requirements of the tenant.
Planned maintenance is crucial for ensuring the reliability and longevity of data center infrastructure. The lease should include a detailed regular maintenance schedule to minimize downtime or other disruptions at the data center. Whether or not provided to tenant, the landlord should maintain detailed records of maintenance activities throughout the term of the lease.
Data center audits are comprehensive evaluations of a data center’s infrastructure, operations, and security practices, designed to ensure compliance with the standards specified in the lease.