Data Center Land Due Diligence Checklist: Power, Fiber, Water and Entitlements

Data center land diligence starts with power, but the deal only works when utilities, title, water, fiber and permits line up.

Data center land due diligence proves that a site can support the power, cooling, connectivity, permitting and construction requirements of a data center before a developer commits capital. It is not industrial land diligence. A site can look clean on zoning, acreage and access, then fail because the utility cannot deliver capacity for 48 months.

That is the market now. JLL's 2026 Global Data Center Outlook projects 97 GW of new data center capacity between 2025 and 2030, roughly doubling global capacity. The same report says average grid connection waits in primary data center markets now exceed 4 years. That turns diligence into a race against power, not a real estate exercise with power added later.

A good diligence process kills weak sites quickly and preserves optionality on strong ones.

1. Confirm the power path before anything else

Power is the first diligence gate because every other assumption depends on it.

The team needs to verify current capacity, future capacity, delivery schedule, required upgrades, substation proximity, transmission constraints and the legal status of any utility commitment. A casual 'capacity available' comment is not enough. Developers need will-serve letters, load study status, queue position, upgrade responsibility and a clear energization timeline.

Morgan Lewis noted in its March 2025 guidance on undeveloped data center land that developers must assess average and peak electrical demand and negotiate with utility providers to confirm adequate supply and infrastructure capacity. If new capacity is needed or power lines must be rerouted, delay risk can change the whole project.

AI can help here by tracking utility correspondence, extracting commitments from emails and letters, comparing queue milestones and building a timeline risk register. Human judgment still owns the utility negotiation, power engineering and decision on whether a schedule is bankable.

2. Check title, survey and easements like the site already has a building on it

Title and survey review is more complicated for data centers because the project may require new easements for power, fiber, water, access, drainage and security.

The diligence checklist should include:

1. Deed restrictions and use limitations

2. Existing access easements

3. Utility easements that conflict with planned buildings

4. Rights of way for power and fiber extensions

5. Mineral rights, rail rights or legacy industrial encumbrances

6. Boundary issues and parcel aggregation risk

7. Offsite easements needed for infrastructure delivery

Morgan Lewis specifically warns that title and survey diligence periods are often longer for data center projects because easements may need to be relocated, terminated, granted or negotiated before the site can be developed.

AI is useful for first-pass extraction from title commitments, surveys and recorded documents. It can build an exception table and compare legal descriptions. It cannot decide whether an easement relocation is commercially acceptable. That needs counsel and the development team.

3. Prove fiber, not just proximity

Fiber diligence should answer 4 questions: who is nearby, what route reaches the site, what rights are needed and whether diverse paths are possible.

A site next to a highway or rail corridor may still need complex rights of way to reach usable fiber. For hyperscale, AI and high-performance compute workloads, diverse paths matter because the risk is resilience, not only speed.

The checklist should include carrier availability, dark fiber options, route diversity, meet-me points, latency requirements, easements, make-ready work and the time required to bring service to the building.

AI can map potential routes, extract provider commitments and compare fiber risk across shortlisted sites. Humans still need to validate commercial terms, route feasibility and operational resilience.

4. Treat water and cooling as entitlement risks

Water is no longer a secondary technical item. It is a permitting, community and operating risk.

BDO's 2026 data center due diligence checklist says investors should evaluate cooling method, system efficiency, water use and noise pollution as AI chips create higher thermal loads. Morgan Lewis also says developers must assess average and peak water demand, inbound and outbound capacity and treatment needs.

For land diligence, that means checking:

• Potable and non-potable water capacity

• Wastewater discharge capacity

• Cooling technology assumptions

• Water rights and restrictions

• Drought exposure

• Community sensitivity to water use

• Backup or onsite storage requirements

• Noise from cooling equipment and generators

AI can compare parcels against water-stress data, cooling assumptions and municipal capacity documents. It should not replace engineering judgment on cooling design or stakeholder judgment on community risk.

5. Validate zoning, entitlements and political path

A site can be technically viable and still fail because the entitlement path is too slow or too exposed.

The diligence team should confirm permitted use, conditional-use requirements, height limits, setbacks, lot coverage, noise limits, emergency generator rules, stormwater requirements, wetlands, floodplain, air permitting, environmental review and local public-hearing risk.

Data centers often trigger local concern around power, water, noise, tax incentives and jobs. That does not mean the site is bad. It means the entitlement strategy needs to be explicit before closing.

AI can scan zoning codes, meeting minutes, planning agendas and comparable approvals. Humans decide whether the political path is acceptable and who should carry the message locally.

6. Build a decision-ready issue log

The output should not be a folder of PDFs. It should be a ranked issue log.

Each issue should include the source, risk level, responsible owner, decision needed, deadline and effect on cost or schedule. The best logs separate 3 buckets:

1. Fatal issues. Power cannot arrive, title blocks access or zoning prohibits the use.

2. Commercial issues. The site works, but only with cost, delay or concessions.

3. Open diligence. More evidence is needed before a final decision.

This is where AI creates leverage. It can read the materials, maintain the log, flag contradictions and keep the diligence clock visible. The development team still owns the call.