Phase II Environmental Site Assessments for Data Center Land: When You Need One and What It Costs

A Phase I ESA is a records review, site visit and interviews. It identifies where contamination may be but does not test for it. A Phase II ESA is the follow-up investigation: it collects and analyzes samples to verify the presence or absence of suspected contamination in the potential source areas the Phase I identified.¹

The governing practice is ASTM E1903-19, issued in November 2019.² It covers hazardous substances, pollutants, petroleum products and certain other substances, and it describes a process based on the scientific method intended to characterize conditions in an objective, representative and reproducible way.² Unlike the Phase I standard, E1903 does not prescribe a fixed checklist. The scope is negotiated between the user and the environmental professional and documented in a statement of objectives.

That flexibility cuts both ways. A Phase II can be a few soil borings around one farm fuel tank or a multi-month program across a former industrial plant. The questions a buyer brings to it, and the decisions it must support, determine its size.

The usual trigger is a Phase I finding of a recognized environmental condition that matters to the deal. On data center land, common triggers include:

• Underground or aboveground fuel tanks at farmsteads, equipment yards or former fuel stations.
• Former industrial, rail, landfill or power generation use, including ash and fill areas on retired plant sites.
• Old electrical equipment, chemical storage or burn and dump areas.
• Upgradient sources such as dry cleaners or industrial plants where vapor or groundwater could migrate onto the site.
• Historical use of firefighting foam, plating, paper or textile processes, or landfills, which raise PFAS questions.⁵

PFAS deserve a specific mention. EPA’s rule designating PFOA and PFOS as CERCLA hazardous substances took effect on July 8, 2024, and practitioners note that the designated substances should now be evaluated in Phase I reports to preserve CERCLA defenses, which can lead to Phase II soil and groundwater testing.⁵ EPA announced on September 17, 2025 that it would retain the designation, and on August 18, 2026 the D.C. Circuit denied the petitions challenging it.⁶⁷

E1903 organizes a Phase II as a sequence of questions and tests. The conceptual site model, a working picture of where contaminants came from, how they move and who or what could be exposed, is built before sampling and revised afterward.² The sampling plan is tailored to the concerns from the Phase I and to the property’s characteristics, and it includes protocols for field handling, sample preservation and transport so the data hold up.¹

Regulatory context shapes the scope. Site-specific state and federal requirements often influence the extent of the investigation, so a Phase II intended to support entry into a state cleanup program may need more samples, specific methods or a state-approved laboratory.¹ If the property may end up in a voluntary cleanup program, tell the consultant at the start; see our guide to brownfield cleanup programs and liability protections.

Sampling design matters as much as the number of samples. EPA’s quality guidance describes judgmental sampling, where locations are chosen by professional knowledge, and systematic or grid sampling, where points follow a set pattern from a random start and can be used to delineate boundaries and search for hot spots.¹² Most data center Phase IIs use judgmental locations around known features, and add a grid only where the source is diffuse, such as a fill area.

Vapor needs its own logic. EPA’s 2015 vapor intrusion guide recommends developing a conceptual site model when vapor-forming chemicals lie beneath or near buildings and weighing multiple lines of evidence, including groundwater and soil gas data. It also notes that petroleum hydrocarbons degrade rapidly under aerobic conditions, which limits their potential for vapor intrusion.¹¹

Laboratory results are then screened. EPA’s Regional Screening Levels are generic, risk-based values for residential and industrial soil and tap water, calculated at a target cancer risk of 1E-06 and target hazard quotients of 1.0 and 0.1.⁸ Practitioners describe them as values EPA relies on to help determine whether remediation is necessary.⁹ Many states publish their own screening levels. A data center is an industrial or commercial use, so industrial values often apply, but using them may require a recorded land use restriction.

Costs scale with the number of areas of concern, the media sampled, the analyte list and whether wells are needed. Oklahoma’s brownfields program puts Phase II costs at $10,000 to more than $100,000 for sites not enrolled in its program.⁴ An environmental consultancy describes $10,000 to $30,000 as the usual range, stretching to $100,000 or more on complex sites.¹³ PFAS analysis, multiple groundwater rounds and large assemblages push toward the top.

On timing, the City of Longmont, Colorado estimates 6–12 weeks to complete a Phase II.³ Fieldwork itself is often a few days. One public contract in South Bend, Indiana, for example, called for drilling and sampling within two calendar days, laboratory analysis within 14 days of sampling, and a draft report within 14 days after the lab results.¹⁴ Drilling permits, where required, add time and cost and should be built into the schedule.¹⁰

A Phase II usually ends in one of four places: no significant impact, so the condition is closed in the record; limited impact that can be managed during construction; impact that needs more delineation; or contamination that calls for a cleanup program, a price change, an indemnity or walking away.

Finding contamination does not by itself end federal protection. EPA’s bona fide prospective purchaser provision allows a party to acquire property with knowledge of contamination, provided it performs All Appropriate Inquiries and meets continuing obligations such as taking reasonable steps to stop continuing releases and prevent threatened ones.¹⁵ Knowledge does change what those reasonable steps are, so the construction plan should reflect the Phase II data.

For data center sites, the Phase II often shapes layout as much as price. Large building pads, deep utility trenches, stormwater ponds and substation foundations all disturb soil. Mapping impacted areas lets the design team keep data halls and ponds off them, plan soil handling and dewatering, and decide whether vapor mitigation is needed in occupied spaces. Coordinate the Phase II with the geotechnical investigation, since both involve borings and can sometimes share mobilization.

1. 01Secure access rights for drilling and well installation in the option or purchase agreement, including restoration duties.
2. 02Size the diligence period for a 6–12 week Phase II plus one possible follow-up round.³
3. 03Ask the consultant for a written statement of objectives, the conceptual site model and the sampling rationale for each area.²
4. 04Decide early whether PFAS analysis is in scope, given the CERCLA designation.⁵
5. 05Agree on which screening levels apply and whether the results may be submitted to a state program.
6. 06Ask for a schedule that separates permits, fieldwork, lab time and reporting.¹⁴

Our guide to due diligence periods and purchase contingencies covers how to build environmental time into a contract, and our methodology explains where environmental screening fits in a site review. If a parcel’s Phase I raised questions, you can get a site reviewed to see whether power, water and layout justify the cost of sampling. Confirm scope and conclusions with a qualified environmental professional and counsel.