Part I — Situation overview

By 1 August 2026 the shutdown of the Paks nuclear power plant moved from the phase of announcements to the phase of implementation. On 29 July, at a Danube water level of −118 centimetres, the shutdown of unit 3 began; on 30 July from 7 a.m., at −121 centimetres, the output of unit 2 was halved; and on 1 August at 05.45 a.m. the still-operating turbine generator set of unit 1 was also shut down. At his press conference in Paks on 31 July, Prime Minister Péter Magyar stated that according to the forecasts the Danube will reach on Tuesday or Wednesday the level — the threshold indicated is −133 centimetres — at which the last unit must also be shut down. In the plant’s 44 years of operation there has been no instance of a complete shutdown. The technical cause is not the quantity of water: cooling requires about 100 cubic metres per second, and the Danube’s flow has fallen from 950 cubic metres in June to 650 cubic metres. The constraint is the water level: at this level the water intake pumps can no longer be operated safely.

On the same day two further steps became public which are the subject of today’s analysis. One is the publication of the content of the rotational load-shedding order (RKR): under this, in a situation of last resort — if the power plant reserve, imports and industrial consumption reduction together are not sufficient — curtailment may be ordered by MAVIR, that is the domestic electricity transmission system operator, and is implemented by the territorially competent distribution companies. Apart from the essential users, the order divides the national load into 18 groups of roughly equal size; the groups are aligned not to municipal boundaries but to the technical structure of the distribution grid, and any one switch-off period may last at most three hours. The other step is the launch of the voluntary phase: the Ministry of Economy and Energy sent a letter to large consumers using more than 50 gigawatt hours (GWh) a year, and according to the prime minister 31 large consumers have so far announced voluntary power curtailment; the government hopes for 400–600 megawatts (MW) of saving from this. The prime minister put the cost of the shutdown, “as a conservative estimate”, at at least 50 billion forints, adding that no increase in household electricity prices is envisaged.

MIAK’s reading is the following: the government treats every single day of the crisis as an individual event — writing a letter, publishing an appeal, convening a council — while in its essence the situation is a recurring, modellable problem in advance, which calls not for an ad hoc reaction but for a rule fixed in advance. The rotational order now announced itself proves this: there is a technical logic for it, but alongside it there is neither an exemption list fixed at the level of a legal source nor a contractual framework for the companies undertaking voluntary curtailment. The problem is therefore not that the government does not communicate, but that the rule and the framework are asymmetrical: the technique of switching off has been worked out, whereas responsibility and compensation have not.

Part II — Literature foundation

Before turning to MIAK’s proposals it is worth fixing the conceptual frame in which today’s situation can be interpreted. Peter F. Drucker, the Austrian-born American author on management theory, identifies in his work The Effective Executive the first step of an effective decision as the question whether we are facing a generic (that is, recurring, traceable to an underlying regularity) or a genuinely exceptional event: for the generic a rule must be made, the exceptional may be handled case by case — and the most frequent managerial error is precisely that generic situations are treated as a series of individual events. Ray Dalio’s volume Principles turns the same thought into practice: the principle must be recorded in advance, in writing, because a rule formulated after the fact is no longer a rule but the justification of the decision. And the OECD’s March 2026 report Testing Resilience adds the macroeconomic side: the effect of the 2026 energy price shock and supply disruption is determined not by the size of the loss suffered but by the economy’s capacity to adapt. The detailed treatment of the literature — author by author, with quotations — can be found in the 6.4 Literature in detail section.

Part III — MIAK’s concrete proposal

MIAK proposes three measurable measures, each of which can be implemented even during the crisis now under way.

The technical content of the rotational load-shedding order is by now known, but its legal status is not. The promise has been made that hospitals, social institutions and other essential public services cannot figure among the consumers subject to curtailment — but this promise was made at a press conference, it did not appear in legislation. According to MIAK this must be promulgated at the level of a decree within 14 days, in such a way that the exempted circle is named by type of institution (inpatient care institution, residential social institution, water utility pumping station, electronic communications backbone node, pharmaceutical and food cold chain). The reason for this is not formal: the 18 groups of the order now published follow the technical structure of the distribution grid, not municipal boundaries — that is, the group as a unit cannot distinguish between a hospital and a residential property connected to the same line section. The exemption is therefore achievable only if the grid separability of the exempted consumers is surveyed in advance, and where this cannot be solved technically, the gap is recorded publicly. In Drucker’s distinction (see 6.4.1) this is the rule given for the generic problem — as against the practice that leaves every curtailment decision to individual consideration. The proposal is linked to the K7 energy market shock resilience programme point.

3.2 A contractual framework and a compensation rule for voluntary consumption reduction (within 30 days)

Today 31 large consumers have undertaken voluntary power curtailment, and the government hopes for 400–600 megawatts of saving from them. This phase currently operates entirely outside a legal framework: there is no published set of conditions on the extent and duration of the undertaking, on when and with what notice period it may be called on, and on whether the undertaking company receives any compensation — a grid use fee discount, a capacity payment or an offset on the tax side. MIAK proposes that within 30 days a public demand-side response programme based on a model contract should be prepared, in which the extent of the undertaking, the threshold for calling it on and the amount of compensation are fixed in advance and are the same for every applicant. This has two benefits. One is economic: a factory that knows at what threshold and with how much notice it has to shut down can reorganise its shift pattern; one that does not know accumulates reserves or postpones production. The other concerns transparency: if the voluntary undertaking remains informal, it cannot be verified after the fact whether the same yardstick was applied to every large consumer. The proposal builds on the G9 strategic industrial policy and the K2 energy transition planning programme points.

3.3 A public methodology for the fiscal cost of the shutdown and its monthly updating (first publication within 30 days)

The estimate of 50 billion forints is today a figure without a methodology. MIAK asks the Ministry of Economy and Energy to publish which items make up this sum — the additional cost of imported electricity, lost producer revenue, grid balancing cost, compensation payments — with what electricity price and duration assumptions it is calculated, and what is not included in it (the prime minister himself indicated that the estimate contains only the primary effects). Thereafter the actual data should be updated monthly, and at the end of the shutdown an after-the-fact accounting should be prepared comparing the estimate with the actual cost. This is the direct application of the after-the-fact impact assessment under programme point G20 — the Drucker audit: an estimate is worth something if we later look at how accurate it was. The machine-readable publication of the data fits the logic of the D9 decision-support system and the A1 public money dashboard.

The three proposals are bound together by a single principle: the quality of crisis management is measured not by how quickly the government reacts on a given day but by whether its decisions are guided by rules knowable in advance. This is the practical translation of Drucker’s proposition, and at the same time the yardstick MIAK applies to every government.

Part IV — Expected effects and risks

Dimension Expected effect Risk
Economy The predictable call-on threshold and compensation reduce corporate stockpiling and production postponement; the public cost methodology improves fiscal plannability Compensation is a budgetary burden; if the thresholds are set too low, the programme can be called on more often than necessary
Society The named exemption list makes it knowable for the population what to expect; the exemption turns from a promise into an entitlement The absence of technical separability may come to light, and this may cause a loss of trust in the short term — even if uncovering the gap is the right step in the long term
Public administration Fixing the procedure reduces the scope for discretionary decision and makes the curtailments verifiable after the fact The 14-day deadline for making the decree is tight; a rushed text may result in imprecise exemption definitions

The main trade-off is tense between speed and precision. At the moment of curtailment the system operator has to decide within seconds, and every exemption fixed in advance narrows its room for manoeuvre: the more consumers receive an exception, the fewer groups remain from which the shortfall can be covered, and the more often the same districts have to be switched off. This is a real technical constraint, not an excuse — which is precisely why the exempted circle must be defined narrowly, by type of institution and not in sectoral generalities. The proposal works if the size of the exempted circle remains one the grid can still bear; if it expands under political pressure, the order loses its workability.

The second trade-off stretches between the voluntary and the mandatory phase. A compensated voluntary programme is more expensive than a cost-free appeal — in return it is predictable and legally unassailable. The current solution, without compensation, looks cheap, but its cost is merely shifted: onto the companies, which price it in as planning uncertainty. The proposal tips towards the risk side if the amount of compensation is set in such a way that undertaking curtailment becomes more attractive than avoiding it — this is excluded by the objective, technical definition of the call-on threshold.

Part V — Measurability and summary

5.1 What is worth tracking? (proposed performance indicators)

According to MIAK the following performance indicators (KPIs, in English Key Performance Indicator) are worth tracking:

  • The appearance of the curtailment order in a legal source: whether the consumer order and the exemption list appear at the level of a decree — proposed deadline 14 days, the yardstick being the presence of named institution types, not a general reference to “essential public services”.
  • The technical separability ratio of the exempted consumers: what percentage of the named exempted circle can actually be separated from the rotational groups — proposed first survey within 60 days, with a public summary.
  • The contractual coverage of the voluntary programme: how many of the announced voluntary undertakings rest on a model contract, and how many megawatts of capacity this covers — proposed target that after 30 days the overwhelming part of the undertaken capacity should rest on a contractual basis.
  • The estimate–actual divergence of the shutdown cost: the difference between the published estimate of 50 billion forints and the actual cost, broken down quarterly — this is the single most important figure of the proposed after-the-fact impact assessment.

5.2 Summary

The key message: MIAK asks the government to promulgate the rotational load-shedding order now published at the level of a decree within 14 days, with a named exemption list; to give a contractual framework and a compensation rule to voluntary consumption reduction within 30 days; and to publish the methodology of the 50 billion forint cost estimate, then to account for it after the shutdown. These three steps do not cost any appreciable money, and do not require a political decision about whether nuclear energy is a good or a bad direction — all three are procedural questions, and all three can be carried out now, even during the crisis.

Two MIAK foundational values are in play in this matter. Accountability, because curtailment always affects someone and not someone else: if the order and the exemption are not written down in advance, then it cannot be established after the fact whether the decision was guided by a rule or by discretion — and where this cannot be established, responsibility cannot be located either. And transparency, because the figure of 50 billion forints is today a claim about public money without evidence: without the methodology of the estimate and its after-the-fact accounting the reader has no instrument for deciding whether the crisis management was expensive or cheap. MIAK asks the same yardstick of every government: a crisis situation is not a dispensation from rule-making but precisely the moment when the absence of a rule is most expensive.

Part VI — Justifications and further sources

6.1 The press framing by spectrum

The liberal-left band raised uncertainty and long-term risk to the centre of its framing. 444.hu’s piece entitled The catastrophe is not the shutdown of Paks but that we do not know when it can be switched back on made the unknown time horizon of the restart the main stake of the news, and with an expert statement (Attila Aszódi, former government commissioner for Paks) emphasised that not a single model had forecast the present flow rate. Telex, in its G7 section, under the title Paks’s shutdown will be a very tough stress test for Hungarian electricity supply, put the load-bearing capacity of the grid in the foreground, while as an economic item it reported the further output reduction in the early hours of Saturday. This band therefore framed the modelling failure and the systemic vulnerability.

The economic and public-affairs band focused on the operational measures and the figures. Portfolio followed the Paks press conference minute by minute, and carried in a separate item the ministerial letter sent to the large consumers — that is, the technical detail of the demand-side intervention, which in the other outlets appeared only at the level of a mention. 24.hu was the only one to set out the operation of the rotational load-shedding order in a stand-alone explanatory article (18 groups, cycles of at most three hours, MAVIR competence) — that is, this band gave the reader the most procedural information. ATV highlighted the unprecedented nature of the situation (“next week something will happen for which there has been no precedent”).

The conservative band emphasised household exposure and the limits of imports. Magyar Nemzet, under the title Rolling blackout: household curtailment may come, put in its headline the consequence affecting households — that is, the layer which government communication consistently steered towards the large consumers. Mandiner put the emphasis on the risk of import dependence (“it is not certain that the Paks shortfall could be made up from imports”), thereby questioning the government’s argument about 3,600–3,800 megawatts of import capacity. It is noteworthy that neither end of the spectrum placed the legal status of the order now published at the centre of its framing — the fact that the exemption promise does not appear in legislation. The press described the technique and the consequence; the question of the legal source was left out.

6.2 Facts and data

Data Value Source
Danube water level at Paks (29 July 2026) −118 cm statement of MVM Paks Nuclear Power Plant
Danube water level at Paks (30 July 2026) −121 cm statement of MVM
The indicated threshold for complete shutdown −133 cm Telex, 1 August 2026
Danube flow rate (June 2026 → end of July) 950 → 650 m³/s statement of Attila Aszódi, 444.hu, 1 August 2026
The water intake required for cooling approx. 100 m³/s statement of Attila Aszódi, 444.hu
The plant’s nominal and current output (31 July 2026) 2000 MW → approx. 960 MW 444.hu, 1 August 2026
The number of rotational groups 18 24.hu, 31 July 2026
The maximum of one switch-off period 3 hours 24.hu, 31 July 2026
Large consumers announcing voluntary curtailment 31 press conference of Péter Magyar, HVG, 31 July 2026
Saving hoped for from the large consumers 400–600 MW Portfolio, 31 July 2026
The circle of addressees of the ministerial letter consumption above 50 GWh/year Portfolio, 31 July 2026
Available import capacity 3600–3800 MW statement of Péter Magyar, 29 July 2026
The estimated budgetary effect of the shutdown at least HUF 50 bn (conservative, primary effects only) press conference of Péter Magyar, Telex/Portfolio, 31 July 2026
The statutory limit on cooling water temperature 30 °C, in exceptional cases 31.5 °C with ministerial consent HVG, 30 July 2026
Romania: the loss of Cernavodă both units shut down, approx. 1400 MW, approx. 20% of the country’s generation Portfolio, 29 July 2026

The cooling water temperature limit deserves separate attention because it shows that a threshold fixed in advance in legislation and a named decision-maker already exist in the system today (30 °C measured 500 metres from the discharge point, which may be exceeded up to 31.5 °C on security-of-supply grounds, with the consent of the minister responsible for energy policy). MIAK’s proposal does not ask for a new type of regulation, but that consumer curtailment should receive a rule of the same level as the one cooling water discharge already enjoys today.

6.3 Policy dimensions

  • Environment and climate (programme points) — energy market shock resilience and the planning of the energy transition; cooling water exposure as a recurring rather than extraordinary planning condition (programme point ID: K7, K2);
  • Economy (programme points) — the industrial policy frame of the demand-side response programme and the after-the-fact impact assessment of estimates (programme point ID: G9, G20);
  • Public administration and e-government (programme points) — the audit of the organisational routines of crisis decision-making: who orders, who implements, who is answerable afterwards (programme point ID: KI11);
  • Digitalisation and AI regulation (programme points) — the machine-readable publication of load and cost data (programme point ID: D9);
  • Transparency and anti-corruption policy (programme points) — the public traceability of public spending in the case of compensation payments (programme point ID: A1).

6.4 Literature in detail

6.4.1 Peter F. Drucker: The Effective Executive

Drucker identifies as the first element of the effective decision process the question whether the event that has arisen is the manifestation of a recurring regularity or a genuinely one-off exception. In his view the two call for different treatment, and the most frequent managerial error is that a generic situation is treated as a series of individual events — that is, that people react pragmatically where the underlying principle is missing.

“All events but the truly unique require a generic solution. They require a rule, a policy, a principle.”

Drucker gives as an example precisely a power outage: he calls the great north-eastern blackout a case that was not an exception but the first appearance of a new type of failure that will henceforth be frequent — and for which a generic solution must therefore be found. In the case of the Paks cooling water constraint this means that the 2026 drought is not an extraordinary event but the first sharp manifestation of a recurring exposure: the response to it must therefore be a rule, not daily discretion. Drucker also warns separately that “emergency temporariness” tends to become lasting — this is the principal risk of today’s voluntary, contract-free curtailment practice.

📖 Source: Peter F. Drucker: The Effective Executive

6.4.2 Ray Dalio: Principles

The central claim of Dalio’s volume is that a principle is a concept that can be applied again and again in similar situations — as opposed to a narrow answer given to an individual question. From this follows the practical requirement: the principle must be written down, because only a principle set out in writing can be tested, disputed and enforced.

“Principles are concepts that can be applied over and over again in similar circumstances as distinct from narrow answers to specific questions.”

Dalio is also emphatic about the output of diagnosis: it must take a form recorded in writing so that patterns can later be identified. In the case of the Hungarian energy crisis this is the argument for the public dashboard and the contractual framework: a voluntary undertaking that does not live in a contract is not a principle but an occasional agreement — it cannot be invoked at the next drought, and it cannot be verified after the fact whether it was concluded on equal terms.

📖 Source: Ray Dalio: Principles

6.4.3 OECD: Testing Resilience

The OECD’s March 2026 interim report examines the resilience of the global economy after the Middle East conflict that also affected energy infrastructure. According to the report’s summary, the disruption of energy transport and the damage to energy infrastructure together caused energy price increases and supply disruptions, and persistently higher energy prices appreciably increase corporate costs and consumer prices, with adverse growth consequences.

“A prolonged period of higher energy prices will add markedly to business costs and raise consumer price inflation, with adverse consequences for growth.”

The report makes the testing of resilience its central concept: the question is not whether a shock hits the economy but how quickly and at what cost it adapts to it. In the present Hungarian situation this gives the yardstick for the estimate of 50 billion forints: the loss in itself does not measure the quality of the crisis management — what measures it is how far the corporate sector was able to adapt in a planned way, and this depends on plannability, that is precisely on rules promulgated in advance.

📖 Source: OECD: Economic Outlook, Interim Report — Testing Resilience (March 2026)

6.5 International comparison

Rotational curtailment is not a Hungarian peculiarity, and international practice differs from the present domestic situation precisely in what MIAK is asking for. The European Union’s electricity market regulation (2019/943) prescribes the preparation of risk-preparedness plans in which the member state must fix in advance the crisis scenarios and the conditions for non-market-based measures — including consumer curtailment. The aim is precisely that curtailment should not be a discretionary act but a procedure published in advance.

In Japan, during the capacity shortage after the 2011 Fukushima accident, the system of keikaku teiden, that is planned power cuts, was operated with a schedule published in advance by district: those affected knew days beforehand which group they were in and when their turn would come. The experience is twofold: the publicity of the schedule improved cooperation, but it also emerged that separating hospitals and facilities operating cold chains is technically considerably more complicated than the planning had assumed — precisely the problem for which proposal 3.1 asks for a separability survey.

The direct regional parallel is Romania: because of the Danube’s low water level both units at Cernavodă have shut down, which means about 1,400 megawatts, some 20 per cent of the country’s electricity generation. This is significant from the point of view of Hungarian import cover: the figure of 3,600–3,800 megawatts of import capacity relates to a normal situation, while the cooling water constraint hits the countries along the Danube at the same time. In MIAK’s reading this is a further argument for putting the demand-side programme on a contractual footing: imports are not a substitutable guarantee in a shock that is regional.

Environment and climate

  • K7 — Energy market shock resilience
  • K2 — Energy transition plan

Economy

  • G9 — Strategic industrial policy
  • G20 — Economic policy impact assessment system (Drucker audit)
  • G15 — Counter-cyclical fiscal stabiliser

Public administration and e-government

  • KI11 — Organisational behaviour audit

Digitalisation and AI regulation

  • D9 — Data-driven public policy decision-support system

Transparency and anti-corruption policy

  • A1 — Public money dashboard

Proposed new programme point: A demand-side response programme with a model contract and a public compensation rule — for the Environment and climate area.

6.7 List of sources

Press sources (MIAK press monitor, 1 August 2026 — topic 1):

Knowledge-base references (books):

  • 📖 Peter F. Drucker: The Effective Executive
  • 📖 Ray Dalio: Principles
  • 📖 OECD: Economic Outlook, Interim Report — Testing Resilience (March 2026)

Note: the local file path of the book does NOT appear in the visible text of the blog — only the author and the title. The file path is an internal matter of the generation process, not the reader’s.

MIAK internal materials:

  • MIAK policy area: Environment and climate (programme points; programme point ID: K7, K2)
  • MIAK policy area: Economy (programme points; programme point ID: G9, G20, G15)
  • MIAK policy area: Public administration and e-government (programme points; programme point ID: KI11)
  • MIAK policy area: Digitalisation and AI regulation (programme points; programme point ID: D9)
  • MIAK policy area: Transparency and anti-corruption policy (programme points; programme point ID: A1)
  • MIAK press monitor, 1 August 2026 — topic 1, score: 93/100

Additional public data sources:

  • MAVIR system operation data; ENTSO-E Transparency Platform
  • Water level time series of the National Directorate for Water Management
  • Regulation (EU) 2019/943 on the internal market for electricity — risk-preparedness plans

Generation metadata