Why Lufeng Nuclear Plant Builds Its Units Out of Order

A nuclear plant where Unit 5 is almost ready to generate power while Unit 3 has not yet received government approval to break ground is not a construction project running behind schedule. It is a deliberate sequencing choice, and understanding why it was made tells you something important about how China builds nuclear capacity at scale.

Lufeng nuclear power plant in eastern Guangdong Province is simultaneously one of China’s most advanced nuclear construction sites and one of its most architecturally unusual. Six units are planned, but they are not being built in order, they do not share a single reactor design, and they were not approved in a single regulatory decision. Each of those facts is intentional.

The construction sequence at Lufeng maps a fuller picture than the unit numbering suggests, running from the units nearest to generating electricity down to those still awaiting final approval. What you will finish with is a clear read on where each unit stands, what the recent cold functional test milestone actually means in technical terms, and why running two reactor designs on one site is a feature of China’s nuclear strategy rather than a complication.

Six units, two reactor designs, and a construction sequence that runs backwards

The first thing that confuses most people looking at Lufeng is the numbering. It is the first nuclear energy project in eastern Guangdong, and it hosts six planned units drawn from two separate Generation III reactor families. Generation III refers to the current class of pressurised water reactors built to enhanced post-Fukushima safety standards.

Here is the part that looks like chaos and is not: the plant is not being built in numerical order.

Units 5 and 6 use the Hualong One design (also called the HPR1000), China’s home-grown reactor developed by China General Nuclear (CGN). Those two units broke ground first. The lower-numbered Units 1 to 4 use the CAP1000, China’s localised version of Westinghouse’s AP1000, developed with China National Nuclear Corporation (CNNC) involvement, and they entered construction years later.

The reason is regulatory. Each design family followed its own approval pathway through separate State Council decisions, and the Hualong One units simply cleared that gate first.

The phased approval logic at Lufeng is not unique to this site; it is a feature of fleet-based nuclear construction, the system by which China coordinates dozens of concurrent reactor projects through standardised procurement, shared component pipelines, and centralised regulatory sequencing.

CAP1000 Units 1-4: later start, phased approvals

The National Development and Reform Commission (NDRC) approved the four-reactor CAP1000 project back in 2014. That is worth pausing on, because it is where the common misreading of Chinese nuclear announcements begins.

NDRC approval established the project concept, not permission to build. Each pair of units still needed its own State Council sign-off before a single foundation could be poured. That decision for Units 1 and 2 did not arrive until August 2024, a full decade after the project was first approved in principle.

Once ground was broken, progress moved quickly. Unit 1 received its CA01 supermodule (a 950-tonne prefabricated section of the containment structure) in September 2025, and its reactor pressure vessel was installed on 14 July 2026. Unit 2 installed its own 950-tonne CA01 supermodule on 21 July 2026.

Units 3 and 4 remain a different story entirely. As of September 2026, they have no State Council approval and no published construction schedule.

For anyone tracking Lufeng as a demand signal for reactor components or uranium supply, that decade-long gap between policy intent and physical construction is the lesson. “Approved” means very different things at different regulatory stages, and the actual construction sequencing matters far more than the unit numbering ever will.

What the Unit 5 cold functional test actually proved, and what it did not

From the outside, the milestone looked like a single line in a press release. CGN announced on 29 April 2026 that Lufeng Unit 5 had completed its cold functional test, describing the result as laying a “solid foundation” for what comes next.

To understand why that matters, you need to know what the test actually does.

A cold functional test (CFT) subjects the reactor’s primary circuit to high pressure at low, non-operational temperature. It is the first time the reactor systems run together with their auxiliary systems, and it cleans the main circulation pipes in the process. The core purpose is to verify the integrity and leak-tightness of the primary-circuit welds, equipment, and high-pressure sections before any nuclear fuel enters the building.

At Unit 5, the test began on 17 April 2026 and completed at 19:58 on 27 April 2026, when the primary circuit pressure was brought back down to atmospheric pressure to confirm success. During the structural integrity phase, the system was pressurised to 110% of design pressure, with an integrated leak rate test conducted at roughly 0.39-0.42 MPa, the standard band for Hualong One designs.

Those numbers had a construction history behind them. The inner containment dome was installed in April 2024 and the outer containment dome in October 2025, both structural prerequisites for reaching CFT readiness.

“Solid foundation for subsequent hot testing, fuel loading, and commercial operation.” CGN press release, 29 April 2026.

Passing CFT confirms something concrete: the reactor coolant system can hold design-basis pressures without leaking, which clears what CGN and CNNC describe as the structural gate of commissioning. The unit has formally entered the commissioning phase.

The same structural gate cleared at Lufeng was reached just weeks earlier at another Chinese site: Changjiang Unit 4 passed its own pressure verification and entered the commissioning phase in September 2026, offering a near-simultaneous benchmark for how the milestone plays out across different reactor programmes.

What CFT does not do is equally important. It runs without fuel and at low temperature, so it tests none of the thermal stresses, high-temperature material behaviour, or full-power dynamic performance that a working reactor experiences. Several high-stakes gates remain ahead:

1. Cold functional testing verifies leak-tightness of the primary circuit at pressure. (Complete for Unit 5.)
2. Hot functional testing raises temperature and pressure toward operating conditions, exercising pumps, valves, and control systems under thermal load.
3. Fuel loading introduces nuclear fuel into the reactor for the first time.
4. First criticality achieves a self-sustaining nuclear reaction.
5. Grid connection synchronises the unit to the electricity network.
6. Commercial operation begins sustained power generation.

For an analyst watching the timeline, CFT completion is a genuine de-risking event. It eliminates one class of construction failure. But the projected 2027-2028 grid connection for Unit 5 sits several gates away, and you should read that date as a working projection rather than a firm commitment.

Why Lufeng runs two reactor designs at the same site

Running two reactor families on one site looks inefficient. Two supply chains, two training programmes, two sets of spare parts. The logic resolves once you see it as industrial policy rather than a planning accident.

The two designs serve different purposes. Hualong One is China’s domestically developed Generation III reactor, promoted jointly by CGN and CNNC as an export-capable product. The CAP1000 is China’s localised adaptation of Westinghouse’s AP1000, a design chosen to absorb foreign technology while building domestic manufacturing competence around it.

Academic analyses of Chinese nuclear governance argue that policymakers deliberately maintain more than one Generation III line. Running parallel designs reduces technological and supply-chain concentration risk, preserves bargaining power with foreign vendors, and lets competing technologies be benchmarked under identical grid and regulatory conditions.

Co-siting is where that benchmarking becomes real. By building both designs at Lufeng, Chinese industry can generate genuine performance and cost data for Hualong One alongside a well-established international benchmark. That side-by-side record strengthens the commercial case when Hualong One competes for export contracts.

Hualong One export positioning has been reinforced by the diplomatic agreements China has signed with multiple partner countries, with state-level energy MoUs functioning as the precursor step that opens national regulatory processes to Chinese reactor technology.

Both designs are Generation III pressurised water reactors built to post-Fukushima passive safety standards, so the choice is not about safety tiers. It is about which supply chain each reactor feeds.

Supply-chain and training trade-offs of a parallel-design site

The costs of a two-design site are real. Operating two reactor families increases complexity in operator training, spare-parts inventory, and maintenance scheduling. Each design carries its own procedures and its own qualified personnel.

The compensating advantage is resilience. A broader supplier base reduces dependence on any single reactor vendor or component source, and it means a problem in one supply chain does not stall the entire site.

There is a regulatory cost worth flagging. Maintaining quality assurance across multiple reactor designs built simultaneously is one of the capacity pressures the National Nuclear Safety Administration (NNSA) has identified as China’s build programme accelerates. For you, the practical takeaway is that milestones on Units 5 and 6 signal a predominantly domestic supply story, while progress on the CAP1000 units signals demand across a partially internationalised component base.

Lufeng’s clean energy case: what the numbers mean at full capacity

Start with what Lufeng is projected to deliver at full strength, then work outward to where it sits in China’s national plan.

CGN’s projections for the completed six-unit plant are substantial:

• Approximately 52 TWh of electricity generated annually.
• Nearly 16 million tonnes of standard coal displaced per year.
• Exceeding 42 million tonnes of CO2 emissions avoided annually.

“Exceeding 42 million tonnes of CO2 reductions per year at full capacity.” Projection attributed to CGN.

Here is the timing dependency that reframes those figures. Every number above assumes all six units are operating. Units 1 and 2 are scheduled for commercial operation in 2030, and Units 3 and 4 have no construction approval at all. The earliest realistic scenario for the full plant running is 2030 at the very earliest, and full six-unit output is genuinely a multi-decade prospect.

You should treat the 42-million-tonne figure as an upper-bound scenario, conditional on regulatory, construction, and commissioning outcomes that stretch at least a decade ahead. It is real in its potential, not in its near-term delivery.

Where Lufeng fits in China’s 2030 and 2035 nuclear build

China aims for roughly 110 GWe of nuclear capacity by 2030 and around 150 GWe by 2035. Against a mid-2026 operating base of roughly 60-62 GWe, hitting the 2030 target requires nearly doubling the operating fleet in under four years.

The World Nuclear Association’s China profile, updated through September 2026, confirms the 110 GWe target as a 15th Five-Year Plan commitment and documents the State Council, NDRC, and NNSA roles that govern the layered approval process each Lufeng unit must clear.

That maths puts Lufeng squarely in the critical delivery window. Around 38 units are under construction nationwide, adding some 44 GWe, and the projected grid connections for Lufeng Units 5 and 6 in 2027-2029 fall exactly when China needs new capacity switching on.

The pace raises a regulatory question China itself has acknowledged. The volume of concurrent projects means NNSA review and inspection resources are under strain, a pressure flagged in China’s own national reporting under the Convention on Nuclear Safety. Financing operates at a similar scale, with Asia, primarily China, expected to account for hundreds of billions of dollars in nuclear investment to 2030, and Lufeng is one of several large multi-unit sites drawing on that pool.

Lufeng’s financing draws from the same capital pool shaping global nuclear capacity expansion, with Asia expected to account for the largest share of the hundreds of billions of dollars in nuclear investment projected through 2030, making site-level execution directly relevant to international capital allocation decisions.

What to watch as Lufeng moves from construction to power generation

The overall project is less useful to monitor than the specific gating events inside it. Four events will tell you whether Lufeng’s timeline is tracking as projected:

1. Unit 5 hot functional test completion and outcome. This is the next major commissioning gate, exercising pumps, valves, and control systems under thermal load. It reveals vibration and thermal expansion behaviour that cold testing cannot, ahead of the projected 2027-2028 grid connection.
2. Unit 6 containment and commissioning milestones. First safety-related concrete was poured on 26 August 2023, and Unit 6’s commissioning sequence should trail Unit 5’s by roughly 12-18 months toward a projected 2028-2029 grid connection.
3. State Council approval for Units 3 and 4. No approval and no construction schedule exist as of September 2026. This is the single most consequential unknown for the full-capacity scenario.
4. Units 1 and 2 construction progress. Unit 1’s reactor pressure vessel was installed on 14 July 2026 and Unit 2’s CA01 supermodule on 21 July 2026, both targeting commercial operation in 2030.

One quieter indicator is worth adding to that list.

China plans to add 102 personnel to the NNSA to manage the expanded regulatory workload, per its national report under the Convention on Nuclear Safety.

That staffing increase is a concrete signal that China’s own regulators are acknowledging the capacity pressure of this build rate. For anyone using Lufeng as a proxy for China’s broader nuclear execution capability, the two events that will most sharply test the optimistic projections are the hot functional test outcome at Unit 5 and the State Council decision on Units 3 and 4. Each is independently trackable through CGN releases, World Nuclear News, and NucNet, and each will either confirm or revise the timeline.

Lufeng’s place in the nuclear transition: conditional on execution, not just ambition

Lufeng’s promise runs along three axes of conditionality, and it is worth stating them plainly:

• Commissioning execution for Units 5 and 6, the units nearest to generating power.
• Construction and regulatory progress for Units 1 and 2, on track but still early stage.
• Regulatory approval for Units 3 and 4, entirely unresolved.

The mixed-reactor fleet and phased approvals are deliberate risk management, and the logic holds up. But deliberate complexity cuts both ways. Every additional design line, approval stage, and commissioning gate is another point where a delay can start, and delays across a six-unit programme can compound.

Unit 5 is the leading indicator to watch. Its hot functional test, fuel loading, and grid connection will set the performance benchmark for Hualong One at this site and signal execution capability to the market. The national 110 GWe target for 2030 leans directly on units like Lufeng 5 and 6 connecting on their 2027-2029 schedule.

The full-plant figures, roughly 52 TWh annually and the CO2 reductions that come with it, are a sequence of dependent events rather than a guaranteed outcome. The pace of that sequence is what separates an ambitious pipeline from delivered decarbonisation, and Lufeng is a compressed version of the same tensions running through China’s entire nuclear expansion: impressive scale, genuine technical progress, and a set of regulatory, workforce, and financing pressures that will decide whether ambition converts into operational megawatts.

This article is for informational purposes only and should not be considered financial advice. Investors should conduct their own research and consult with financial professionals before making investment decisions. Past performance does not guarantee future results, and forward-looking projections are subject to regulatory, construction, and market conditions.