Rosatom: repair optimization gave Russian NPPs 94 extra operating days in half a year

• Российские АЭС подвели итоги первого полугодия 2026 года (17 блоков, ремонты −94 дня, +7,44 млрд кВт·ч, 37 ремонтов на год) — Российская газета, 2026-07-09
• Дополнительная выработка 7,44 млрд кВт·ч сверх плана за 1П 2026 — EnergyLand, 2026-07
• Пилоты цифровых двойников Росатома: время диагностики −30% — РБК Отрасли, 2025-09
• Виртуально-цифровая АЭС (ВНИИАЭС + ИБРАЭ РАН) принята в эксплуатацию в «Росэнергоатоме» (~100 испытаний, цитата Тухветова) — Атомная энергия 2.0, 2020-02-26
• Росатом принял в эксплуатацию Виртуально-цифровую АЭС с реактором ВВЭР — Neftegaz.ru, n/a
• Атомные электростанции России (36 энергоблоков, 11 АЭС) — Росэнергоатом, n/a
• АЭС России досрочно выполнили госзадание 2025 года (215,339 млрд кВт·ч) — Атоммедиа, 2025-12
• Сокращение сроков ремонтов на Курской АЭС принесло дополнительные 180 млн кВт·ч — MashNews, n/a
• Инженеры Ростовской АЭС сократили время обмена данными при ППР с 240 до 100 минут (электронная ведомость, ПСР, 851 час/год) — Atomic-Energy.ru, 2026-01-13

Background

Rosenergoatom (Rosatom's power generation division) operates 36 power units at 11 nuclear plants across Russia. In 2025, Russian NPPs generated 215.339 billion kWh, completing the state target ahead of schedule. For scale: one large 'thousander' unit delivers around a million kilowatts of capacity, and each day of its downtime means tens of millions of kilowatt-hours that never reach the grid.

That is why scheduled preventive maintenance is the central plot of the operating fleet's economics. For an NPP, maintenance is not an option but law: refueling, mandated inspections, and replacement of major equipment such as steam generators are hard-wired into each unit's life cycle. Thirty-seven repair campaigns are planned across the fleet for 2026 (Rossiyskaya Gazeta). The question is not whether to repair — but how many days each campaign takes, and how many of those days better planning can return to generation. In this sense the nuclear industry resembles aviation: in both, the maintenance regimen is untouchable, so all competition for efficiency unfolds in the logistics around the regimen — work sequencing, parts and crew readiness, and the speed of information flow between shifts and shops.

Rosatom has been building the digital loop for this task for years. As early as February 26, 2020, the Rosenergoatom concern commissioned the Virtual Digital NPP with a VVER reactor — a software-hardware complex developed by VNIIAES and IBRAE RAN that models any unit mode on a compact Russian-made supercomputer: from normal operation to severe accidents (Atomic Energy 2.0). The state corporation's digitalization was led from 2018 to early 2025 by Ekaterina Solntseva; Kirill Komarov is Rosatom's First Deputy Director General, head of the development and international business block. In parallel, the industry runs RPS — the Rosatom Production System, an in-house lean methodology from which many local repair-optimization projects have grown.

Problem

Scheduled preventive maintenance is the largest controllable output reserve of the operating NPP fleet: new units take years and hundreds of billions to build, while cutting downtime works immediately and almost for free. The task is fundamentally constrained by the fact that safety is non-negotiable: schedules can only shrink through better planning, diagnostics, and coordination — never through the scope or quality of work. The regulator, insurers, and the industry's own culture rule out even posing the question 'skip an inspection for the schedule's sake'.

Inside a repair campaign hides an enormous volume of coordination routine. A telling example from Rostov NPP: during maintenance, specialists from the turbine, reactor, and electrical shops, the thermal automation shop, and the ventilation shop entered work-progress data into paper daily sheets, then digitized them and relayed the information by phone — the information exchange cycle took 240 minutes a day (Atomic Energy 2.0). Multiply that by dozens of departments, weeks of repair, and 36 units — and it becomes clear why 'paper' coordination eats days of the calendar schedule.

There is a third constraint: experiments on an operating nuclear unit are impossible by definition. A new operating mode, an emergency procedure, or the consequences of a modernization cannot be tried 'live' — you need an environment where a mistake costs nothing. Before virtual plant models, no such environment existed.

Solution

Rosenergoatom systematically optimizes repair campaigns across the fleet, and this is not one technology but a stack of tools at different scales.

The top layer is modeling. The Virtual Digital NPP with a VVER reactor (in industrial operation since February 2020, with about 100 standalone and integrated trials during 2019) reproduces any unit mode without consequences: it is used to build operator training simulators, run emergency drills, verify control algorithms, and check modernization projects before touching real hardware. VNIIAES director general Farit Tukhvetov called the complex 'one of the main steps on the way to full-fledged digital twins' of power units. In digital-twin pilot projects, equipment diagnostics time fell 30% (RBC Industries); twins are also being developed for small modular plants.

The middle layer is the organization of the repairs themselves. In the first half of 2026, scheduled maintenance ran on 17 power units, and the campaigns were shortened by a combined 94 days (Rossiyskaya Gazeta). Notably, the half-year tally includes not just 'light' refuelings but heavy jobs that usually stretch schedules by months: at Balakovo NPP unit 3, two of four steam generators were replaced — an operation involving cutting and welding of large primary-circuit equipment — while at Rostov NPP unit 1, life-extension work was performed. The new VVER-TOI unit 1 at Kursk NPP-2 — the first of the new generation of serial units — added further output.

The bottom layer is local RPS (lean) projects that add up to the systemic effect. The Rostov NPP example: engineers replaced paper daily maintenance sheets with an 'Electronic Daily Repair Sheet' — five shops (turbine, reactor, electrical, thermal automation, and ventilation) enter work-progress data from mobile devices, and reports assemble in real time instead of the 'paper → manual digitization → phone call' chain. The information exchange cycle fell from 240 to 100 minutes a day — a 2.4x reduction; the system was successfully tried during the unit 3 repair completed on December 21, 2025. RPS development department head Sergey Fidiy summed it up: 'The RPS project's goals were achieved at 106%. Production time was cut by 140 minutes a day — 851 hours saved per year' (Atomic Energy 2.0). Rostov NPP chief engineer Andrey Gorbunov stressed the managerial side: RPS 'stimulates employee initiative and develops their professional competencies in the search for innovative approaches' — meaning optimization ideas come bottom-up from shop engineers, not down from the corporate center.

The industry's priorities matter here: Rosenergoatom deputy director general Andrey Dementyev, commenting on the half-year results, framed them as 'the main tasks remain the same: operational stability and preventing violations.' Shorter repairs are a derivative of reliability, not its competitor; Rosatom's inspector general Sergey Adamchik separately noted that in 2024–2025 the enterprises worked more than 500 days without injuries for the first time.

Result

In the first half of 2026, repair campaigns across the Russian NPP fleet were shortened by a combined 94 days (maintenance ran on 17 power units), and output, per the June 16, 2026 forecast, exceeded plan by 7.44 billion kWh (Rossiyskaya Gazeta, EnergyLand). Station examples: Kursk NPP gained an extra 180 million kWh from shorter repairs, Balakovo about 1 billion kWh, and Rostov 40 million kWh thanks to unit 4's early grid reconnection. Digitalization's local contribution is quantified too: the electronic maintenance sheet at Rostov NPP alone saves 140 minutes a day — 851 hours a year.

Important caveats. The 94 days are the combined fleet-wide effect of repair optimization for the half-year — on average days, not weeks, per unit. The 7.44 billion kWh of above-plan output results from a combination of factors: shorter repairs, operational efficiency, and the new Kursk NPP-2 unit coming online; isolating an 'AI share' or 'digital-twin share' within it is impossible from open data, and Rosatom itself publishes no such attribution. The '−30% diagnostics time' effect belongs to digital-twin pilots, not the whole fleet.

In our view, this case is a rare example of the right sequence: first a reliability and lean culture built over decades (RPS), then a virtual environment for safe experiments (the Virtual Digital NPP since 2020), then targeted digital tools in the maintenance loop — and only atop this pyramid the aggregate result of 94 days. Tellingly, the loudest percentage in the case (the 2.4x speedup in data exchange) was achieved not by a neural network but by replacing paper and phone calls with a mobile app: in capital-intensive industries, plain process digitization still yields effects worth publishing next to the word 'AI'.

A second editorial observation: the nuclear industry publishes its effects in physical quantities — days, kilowatt-hours, minutes — which makes them verifiable. We consider such reporting a model for industrial AI cases: ruble estimates depend on prices and methodologies, while 94 days and 7.44 billion kWh can be checked against dispatcher data.

Lessons learned

1. In capital-intensive generation, the fastest lever is availability of existing assets: 94 fewer repair days yielded 7.44 billion kWh without building a single new unit.
2. Effects should honestly be counted fleet-wide, not per unit: the combined '−94 days per half-year' across 17 repaired units averages to days, not weeks, per unit — and that is how Rosatom itself publishes the numbers.
3. In nuclear, a digital twin is above all a diagnostics and simulation tool (−30% diagnostics time in pilots), not a magic repair-shortener: effect attribution must be kept separate.
4. A virtual plant model that can simulate accident modes enables training and optimization where real-world experiments are impossible by definition.
5. The fastest wins in the maintenance loop come from plain digitization of coordination: swapping paper sheets and phone calls for a mobile app sped data exchange 2.4x (851 hours a year at one plant).
6. Shorter schedules are a derivative of reliability, not its competitor: the industry logs safety records first (500+ days without injuries), schedule records second.
7. Extra output in kWh is a more honest metric than rubles: it is published officially and does not depend on price assumptions.

Frequently asked questions