ARPA-E's UNLOCK Wants to Double Shale Recovery. The Gate Is a Field Site You Need Within 30 Days.

The shale revolution made American oil and gas production the largest in the world, but it left most of the resource underground. ARPA-E's newest program puts a number on that: in shale and tight oil, recovery efficiency is "often below 20%," with primary recovery typically in the 5% to 10% range. The other 80% or more stays in the rock.

On October 21, 2026, ARPA-E released UNLOCK, short for Unlocking New Levels of Output and Capacity for Key Subsurface Resources, under two parallel notices:

• DE-FOA-0003630, the open NOFO for universities, national labs, companies and individuals
• DE-FOA-0003631, an SBIR/STTR version for small business concerns

The goal is stated plainly: double the per-well recovery efficiency of energy and critical resources from low-permeability formations while holding cost and wastewater at or below today's levels. The two notices share about $45 million and an expected 9 to 12 awards.

Concept papers are due at 9:30 a.m. ET on November 20, 2026, 30 days after release. That timeline would be tight for any ARPA-E program. It is especially tight for this one, because UNLOCK will not consider a concept paper that lacks a signed letter of commitment from a field test site.

The numbers

A quiet period is in effect. Until selection letters go out, the only channel for questions is ARPA-E-CO@hq.doe.gov. Emails to program staff are disregarded.

What ARPA-E is actually buying

UNLOCK is not a stimulation program or an enhanced recovery program. It is a program for integrated stimulation and recovery, and the NOFO makes that the first requirement. Each submission must combine a stimulation approach and a recovery approach, with "at least one of these" being "novel and transformative, not incremental." Each must also tie sensing and site-calibrated numerical modeling into that workflow.

The list of technologies of interest is broad:

• Waterless or low-water stimulation: energized gas injection with supercritical CO2 or nitrogen, chemical stimulation, electric discharge and pulsed plasma fracturing, high-energy laser stimulation, thermal and acoustic stimulation, self-propagating fracture networks
• Advanced recovery fluids: supercritical, nanoparticle-enhanced or microbe-assisted fluids; adaptive or phase-changing fluids such as temperature- or pressure-activated gels
• Engineered proppants: shape-adaptive, reactive, catalytic, in-situ generated, multifunctional (sensing or scale-inhibiting), high-temperature stable
• Diagnostics and modeling: distributed chemical sensing, quantum sensors, swarm-based and micro-sensors, quantum computing, and what the NOFO calls "super intelligence"

The last item reflects the administration's current preferred term for AI, which now appears across DOE solicitations. The NOFO also requires applicants to disclose in the project summary whether and how generative "SI" tools were used in writing the application.

Two categories, unequal odds

Applicants choose one category.

Category A: Shale and Tight Oil & Gas. ARPA-E "anticipates that Category A will comprise the majority of projects selected." The targets are specific:

Category B: Other Low-Permeability Subsurface Resources. This covers geothermal, in-situ mineral extraction, coalbed methane and "emerging fields such as geologic hydrogen." The NOFO gives two worked examples:

• Enhanced geothermal: thermal recovery of at least 70% (versus SoA below 35%), water leak-off below 1% (versus 1% to 20%), and production cost at or below the $59 to $101 per MWh SoA range
• In-situ leaching of uranium: recovery of at least 80% (versus SoA below 40% in low-permeability formations), wastewater at or below roughly 50 to 250 gallons per pound of U3O8, and cost at or below $13 to $60 per pound

Category B applicants must build their own metrics table modeled on these examples. That is extra work, and a lower expected selection share. But Category B also likely draws fewer applicants, since geothermal developers and ISR uranium operators with university partners are a smaller pool than shale operators. If your technology really applies to both, consider whether a strong Category B paper beats a middling Category A one.

The field site requirement is the real filter

Read the "Submissions Specifically Not of Interest" list and the program's shape becomes clear. ARPA-E will not merit-review concept papers that:

• Lack a committed field trial site, shown by the required letter of commitment, or do not integrate the site operator into Phase 1
• Rely on cores from sites other than the proposed field test location
• Propose technologies that cannot be field-tested downhole at relevant depths within the period of performance
• Propose to fund drilling technologies or activities
• Offer standalone sensing, monitoring or modeling not coupled to stimulation and recovery
• Focus on above-ground recovery or surface facility optimization
• Lack a demonstrated lab or bench-scale proof of concept
• Omit any of the recovery, water or cost metrics

Each project runs in two phases of about 18 months. Phase 1 is laboratory validation using test-site cores, fluids and data. Phase 2 is a downhole field test at well scale. Funding for both phases is obligated at the start, but the move to Phase 2 is "a portfolio-wide down-select, so advancement is not guaranteed even if all milestones from Phase 1 are met."

Put together, an eligible UNLOCK team needs, within 30 days:

1. A technology already proven at bench scale
2. An operator willing to sign a commitment letter for a downhole trial. ARPA-E will not fund drilling, so the well has to exist already or be drilled at someone else's expense
3. Access to that site's core and fluid samples for Phase 1
4. A permitting plan and the current permit status
5. Site-specific baseline numbers for recovery, water and cost
6. A techno-economic analysis in ARPA-E's category-specific Excel template

For most university labs, item 2 is the bottleneck. If you do not already have a relationship with an operator, your best route is through someone who does: a petroleum engineering department's industrial affiliates program, a state geological survey, a DOE field laboratory, or an existing NETL or Utah FORGE-style test site partner. Note that cores must come from the same site you propose to test, so a team cannot use convenient archived cores for Phase 1 and a different well for Phase 2.

The NOFO also requires the operator to be integrated into the project in both phases. A letter that only grants access to a well, without any ongoing role, risks being judged nonresponsive. Describe what the operator will actually do in Phase 1, such as supplying samples, sharing production history, or reviewing the field test design.

Which notice to use

Universities, national labs, large companies and nonprofits apply under DE-FOA-0003630. FFRDCs and DOE labs may lead a team but cannot apply alone.

Small businesses face a choice. A small business concern may lead an application under only one of the two notices, and ARPA-E "strongly encourages" small businesses to use the SBIR/STTR version. That version uses combined Phase I/II awards (up to $2,477,017) or Phase I/II/IIS awards (up to $4,630,944), with SBIR/STTR work-share rules: under SBIR, the small business performs at least 66.7% of Phase I and 50% of Phase II; under STTR, at least 40%, with a research institution partner.

Under the open NOFO, cost share matters:

• Base: 20% of total project cost
• IHE or nonprofit standalone, or teams made up only of IHEs, nonprofits and labs: 5%
• Small business lead performing at least 80% of the work, or small business consortia: 0% for the first 12 months, then 10% overall
• Teams where universities, nonprofits, small businesses and FFRDCs do at least 80% of the work: 10%, though a large business on such a team holding a patent waiver still owes 20% on its share
• Other transaction agreements: normally 50%

A large operator that leads its own application should expect to bring more than 20%. The NOFO "strongly encourages" it. A university-led team with an operator as a subrecipient may fall into the 10% tier, depending on how the work is split. Model your cost share before deciding who leads.

There is no cap on submissions per organization, as long as each one is "scientifically distinct."

How concept papers are scored

The concept paper is four pages for Sections 1 to 4, plus a one-page table of underlying variables, a one-slide summary, the TEA workbook and the commitment letter. Two criteria, weighted equally:

• Impact relative to NOFO targets (50%): whether the concept is transformational, whether it reaches the category targets, and whether it shows awareness of competing approaches
• Scientific and technical merit (50%): feasibility backed by data, an approach that explains why it beats alternatives, clear deliverables, and team capability

ARPA-E also requires that the underlying variable data be "verifiable with clear references, statistically representative," and "not selectively chosen," and it reserves the right to request the full dataset. Applicants who set their state-of-the-art baseline low to make a doubling look easy should expect that to be checked. Use the operator's actual production history as the baseline.

The policy context

UNLOCK fits the administration's energy agenda: ARPA-E describes success as "reducing foreign energy imports and enhancing the nation's energy infrastructure resilience," and Category A is fossil-fuel production work. But the program is also aimed at water, one of the oil field's biggest cost and environmental problems, and Category B gives geothermal and in-situ mineral recovery a real place alongside oil and gas. Researchers in enhanced geothermal and critical minerals should not assume the program isn't for them. The example targets were written with their resources in mind.

ARPA-E previously floated this concept through a teaming partner list for a potential NOFO on advanced stimulation and enhanced recovery in low-permeability formations, using the same Category A and B split. If you joined that list, check who else did, since those contacts are now your most direct route to an operator or a technology partner.

A 30-day plan

• This week: Pick a category. Contact operators and test-site managers. Register in DOE eXCHANGE and download the concept paper, TEA and slide templates. Confirm SAM registration.
• By November 3: Have a draft commitment letter in the operator's hands that names their Phase 1 role. Get core and fluid sample access in writing.
• By November 10: Send any remaining questions to ARPA-E-CO@hq.doe.gov before the question cutoff.
• By November 17: Finish the TEA and the underlying variable table using site data. Have someone outside the team check the baseline sources.
• November 19: Submit. The deadline is 9:30 a.m. Eastern, not 5 p.m., so a deadline-day upload leaves almost no room for portal problems.

Related reading

For other ARPA-E and DOE opportunities, see our analysis of ARPA-E's fusion commercialization programs and the DOE Office of Science FY2027 open call. Small businesses weighing the SBIR/STTR track should read our overview of the 2026 SBIR/STTR reauthorization.

Granted can help you turn a bench-scale result and an operator relationship into a four-page concept paper mapped to UNLOCK's metrics before November 20. Start with Granted.

Sources: ARPA-E, Notice of Funding Opportunity DE-FOA-0003630, "Unlocking New Levels of Output and Capacity for Key Subsurface Resources (UNLOCK)," and DE-FOA-0003631, UNLOCK SBIR/STTR, both issued October 21, 2026 (Grants.gov opportunities 364136 and 364135).