The Baltic Microreactor Initiative
Lithuania does not need another technology assessment. It needs to own the one thing no one in the region is building: a clear, risk-proportionate framework that lets an advanced reactor be licensed once and deployed many times.
"The reactors are the easy part. The Initiative builds the path they travel on — and makes sure that path only has to be built once."
The Shift: from "can it be licensed?" to "can it be licensed once?"
The region is moving. Estonia has passed its nuclear law and stands up a regulator in 2027.[a] Lithuania's strategy sets a decision point in 2028, and a ministerial working group is already live.[b] Memoranda are stacking up — the BWRX-300 with GE Vernova Hitachi and SGE, Newcleo's lead-cooled design, and more.[c]
What is missing is not interest, capital, or technology options. It is the cross-cutting regulatory architecture that turns any of these into something deployable at speed and at scale. That gap is the opportunity — and it is bounded by the 2028 decision.
The question of whether an advanced microreactor can clear a regulator is closing. Last Energy has completed a Preliminary Design Review with the UK regulator and entered site licensing, is in pre-application with the US NRC, and holds a place in the US Department of Energy's reactor pilot programme.[d] Basic licensability is being demonstrated — by others, now.
What remains wide open is repeatability. Today every unit is licensed project-by-project, site-by-site, country-by-country. The same reactor faces fresh design scrutiny in Vilnius, then again in Tallinn, then again in Warsaw. That repetition — not the physics, not basic permission — is the cost and time sink. Eliminating it is the unclaimed lane, and it is the one worth occupying.
"Eliminate regulatory repetition, and the economics of small reactors change entirely."
The Levers: A Clear, Risk-proportionate Framework
The goal is not lighter-touch nuclear. It is proportionate nuclear — high, predefined safety standards scaled to actual hazard, so that a reactor holding a fraction of the radioactive inventory is not regulated as though it were a gigawatt plant. This is the established international direction, not a fringe position: the IAEA's graded approach is long-standing, the US has reduced emergency-planning requirements for small reactors and built a risk-informed licensing track, and the UK and Canada are developing parallel graded frameworks.[e] The opportunity is to implement that direction faster and more cleanly than anyone else — to lead, not to copy-paste.
Within that framework, three levers do the work:
- Type-certification— the flagship
Certify a standardized design once, exhaustively, and license the design rather than each unit, with only site-specific review thereafter. Aviation does this with airframes; France did it with standardized fleets. This is what turns 'licensable' into 'repeatably deployable' — the unclaimed lane described above.
- Right-sized emergency planning zones— the highest single cost unlock
Hazard scales with the core's radioactive inventory. Below a certain size, passive heat removal can exceed decay heat: the reactor cools itself with no operator action and no power, and cannot melt down. Where the credible worst-case offsite dose is small enough, the emergency planning zone can shrink toward the site boundary. Because so much of nuclear's regulatory and siting weight exists to protect a large surrounding population, this one change removes a disproportionate share of the cost and constraint.
- Standardization on proven designs and fuel— the enabler
Novelty is what slows licensing. The genuinely fast and light path is small and conventional — a proven reactor platform on standard low-enriched fuel — not small and exotic. Designs reaching for unconventional coolants or highly-enriched fuel reintroduce exactly the review burden and proliferation scrutiny that proportionality is meant to relieve.
The smaller the reactor, the stronger this case becomes — which is why microreactors are not just the proving ground for type-certification, but the place where a risk-proportionate framework is most defensible.
Why This Is an Economic Argument, Not a Regulatory Nicety
Take the most credible low-cost design on the table. Last Energy claims roughly $5,000–6,500/kW for a 20 MWe unit — an order of magnitude below generic microreactor estimates — achieved by using a conventional pressurized-water reactor on standard low-enriched fuel rather than exotic coolants and fuels.[f] Strip away the headline number and the structure is what matters: the hardware cost is broadly fixed, while the licensing layer splits into a one-time design cost and a recurring per-deployment cost that repeats in every country.
Type-certification barely touches the hardware. What it moves is decisive:
- Time-to-revenue and returns
Compressing per-deployment approval from years to months pulls revenue forward and collapses carrying cost — worth far more to project economics than any single line-item saving.
- The cost curve itself
You cannot order twenty identical units, and trigger the factory learning curve that takes a design from first-of-a-kind to nth-of-a-kind, if each one faces a fresh design review. Type-certification is the precondition for the volume that makes the hardware cheap.
- Bankability
A repeatably-deployable design carries a lower risk premium — and a lower cost of capital — than one licensed bespoke per site.
And the value compounds regionally. A design certified once and recognized across, say, five follower states eliminates on the order of hundreds of millions in avoided per-country licensing and delay — value that accrues to every developer, not a single vendor. That is precisely why it must be built as public-interest infrastructure rather than a company's private filing: a vendor cannot credibly ask a regulator to "certify my product once for everyone," but an independent, technically credible body can propose the pathway in the public interest.
The Prize: A Private-led Center for Nuclear Innovation
The deeper point is not that Lithuania licenses a handful of reactors. It is that a clear, first-mover regulatory framework is itself an economic asset — one that attracts private capital, developers, supply chain, and talent, and concentrates them in one place. Regulatory certainty is the scarce input in advanced nuclear; the jurisdiction that supplies it first becomes the centre of gravity for the industry.
This is a well-worn path in other sectors:
Transposed the EU's UCITS fund passport first, in 1988, and became Europe's largest fund domicile — an industry that now underpins a substantial share of its economy.
Built the global aircraft-leasing industry on a clear legal and tax framework; most of the world's leased aircraft are domiciled there today. Aircraft, like reactors, are physical assets — proof the model works for hardware, not just paper.
Became the legal home of most large US companies not by being the loosest regime but the most predictable one, with specialised courts and deep case law.
Turned digital-government leadership into a global brand that compounds into talent, software exports, and reputation.
In each case the winner sold certainty and recognition, moved before the larger players, and captured value far out of proportion to its size. That is precisely the opening for the Baltics in advanced nuclear: the standard is not yet set, and the first credible framework will anchor the ecosystem around it.
Critically, this is private-led. The state's contribution is the framework; private players bring the capital, the technology, and the execution risk — the model Last Energy already runs, taking no government financing. For a government, that is industrial development without a state-balance-sheet bet: clarity offered in exchange for investment, jobs, and supply chain.
What Lithuania would actually capture is worth stating plainly, because a reactor is not mobile the way a fund or an aircraft is. The country does not earn paper-domicile rents; it earns the position of certification origin, deployment hub, and standards author — and the GDP accrues industrially: a clustered supply chain, high-skilled jobs, exportable expertise, and, not least, the cheap and reliable baseload that pulls energy-intensive industry such as data centres into the country. The closest growth model is Ireland's aircraft leasing, not pure fund domiciliation.
The Vehicle: An Independent Coordination Body
The IEA has already recommended that Lithuania establish an independent nuclear coordination body — mandated to oversee stakeholder engagement and, together with the safety authority, identify the regulatory and licensing adaptations needed to efficiently license advanced reactors.[j] The Baltic Microreactor Initiative is a concrete proposal to seed exactly that body. Its work:
- Define the pathway
A risk-proportionate, type-certification licensing pathway for advanced reactors, with microreactors as the first application.
- Convene
Bring together the actors who must align, and keep them aligned.
- Anchor in rigor
Technical-support-organization-grade work, so its output is a document a regulator can act on — not another position paper.
The roles are distinct and complementary:
- Ministry of Energy
Provides the mandate and the fit with the 2028 decision.
- VATESI
The ultimate owner of any licensing framework — and Lithuania's decades of regulating Ignalina make it uniquely equipped to lead one.
- Altra
Brings operational competence and the institutional will to make new nuclear real.
- LEI
Provides the technical spine — safety-case and severe-accident rigor, and access to EURATOM and Horizon funding.
The Initiative coordinates and authors. It does not duplicate any of them.
What We Will Not Pretend
An independent body earns its standing by being honest about what it does not yet know.
Four truths
- 01.
The low-cost claims are first-of-a-kind and unproven.
Today's headline numbers are credible targets, not demonstrated outcomes, and nuclear first-of-a-kind projects have a long history of overrunning. The first real proof points are imminent — not banked.
- 02.
The cost curve is a hypothesis.
It depends on order volume that does not yet exist. Which is exactly why removing the regulatory barrier to that volume is the highest-leverage move available.
- 03.
Type-certification is necessary, not sufficient.
It removes regulatory repetition. It does not, by itself, create demand or make the hardware cheap.
- 04.
Some burdens do not scale down.
Physical security against sabotage and material theft, safeguards and non-proliferation obligations, third-party liability, and waste and decommissioning responsibilities persist largely regardless of reactor size. As an EU member, Lithuania also cannot legislate beneath the Euratom and IAEA safeguards floor; the pathway must be built inside those obligations, not around them. 'Proportionate and clear' is the realistic ceiling — not 'minimal.'
Naming these openly is the point. An honest, independent body is more useful — and more bankable — than another booster.
Why Now, and Why the Baltics
The 2028 decision is the clock. Estonia's regulator stands up in 2027; desynchronization from the Russian grid has made sovereign baseload a live political priority; EURATOM funding is open.
The region is small enough that one regulator's precedent is respected, and serious enough that the precedent travels. The window to build the licensing architecture is the next 18–24 months — before the technology choices harden around imported, per-site licensing by default.
What Success Looks Like
Not a press release. A regulator adopting a type-certification pathway that did not exist before. A credible, independent body recognized as the place where repeatable advanced-reactor licensing gets defined. And the Baltics established as the first region in Europe to make advanced reactors deployable at scale — setting the standard, rather than importing it.
"The reactors are the easy part. The Initiative builds the path they travel on — and makes sure that path only has to be built once."
- [a]Estonia's nuclear law and 2027 regulator — World Nuclear News, "Estonia's Parliament passes new nuclear energy law"; NucNet, "Estonia's Parliament Passes Legislation That Paves Way For Nuclear Energy" (June 2026). world-nuclear-news.org; nucnet.org
- [b]Lithuania's 2028 SMR decision point, ~1.5 GWe target, and ministerial working group — World Nuclear Association, "Nuclear Power in Lithuania"; World Nuclear News, "Lithuania to look into nuclear energy options" (2025). world-nuclear.org; world-nuclear-news.org
- [c]BWRX-300 assessment MoU (GE Vernova Hitachi / SGE / Altra) and Newcleo lead-cooled MoU — Nuclear Engineering International, "Lithuania considers SMR technology" (March 2026); NucNet, "Lithuania To Study Deployment Of Newcleo Small Modular Reactors" (July 2025). neimagazine.com; nucnet.org
- [d]Last Energy: UK Preliminary Design Review and site licensing, US NRC pre-application, DOE Reactor Pilot Program — Neutron Bytes, "Last Energy Earns PDR Milestone at UK ONR" (Aug 2025); US NRC pre-application activities page (Last Energy); Nuclear Engineering International, "Last Energy plans Texas microreactor" (Oct 2025). nrc.gov; neimagazine.com
- [e]Risk-proportionate / graded regulation precedent (IAEA graded approach; US NRC alternative emergency-preparedness rule for SMRs, 2023; Part 53 risk-informed framework, 2026) — US Federal Register, "Emergency Preparedness for Small Modular Reactors and Other New Technologies" (2023); US Federal Register, "Risk-Informed, Technology-Inclusive Regulatory Framework for Advanced Reactors" / 10 CFR Part 53 (2026). federalregister.gov
- [f]Last Energy PWR-20 cost (~$5,000–6,500/kW; ~$130m/unit; EXIM $103.7m letter of interest) and design (single-loop PWR, 4.95% LEU) — Neutron Bytes (Aug 2025); American Nuclear Society, "Last Energy seeks $103.7M for Welsh microreactor project"; Last Energy company technology page. ans.org; lastenergy.com
- [g]Luxembourg first to transpose the UCITS Directive (1988); largest EU fund domicile — Luxembourg for Finance, "Asset Management"; Delano / Funds Europe, "1988: transposition of the UCITS directive." luxembourgforfinance.com; delano.lu
- [h]Ireland's ~60–65% share of the global leased-aircraft fleet, GPA origins, Cape Town Convention — Irish Times, "Ireland can't afford to throttle back on aircraft leasing" (Jan 2025); Aircraft Leasing Ireland / Ibec data release (Nov 2025). irishtimes.com; ibec.ie
- [i]Delaware as legal home of ~66–68% of the Fortune 500; incorporation revenue ~29–33% of the state budget; Court of Chancery — Delaware Division of Corporations statistics; WHYY, "Delaware moves to protect lucrative incorporation 'franchise'" (Mar 2025). corp.delaware.gov/stats; whyy.org
- [j]IEA recommendation that Lithuania establish an independent nuclear coordination body — International Energy Agency, Lithuania country profile. iea.org/countries/lithuania