Low water · Chooz, Ardennes (Meuse gauge B720000001 and nuclear plant) · 12 Nov 2026
Both Chooz reactors stopped during 103 days of low Meuse flow: the limit was a Franco-Belgian agreement, not cooling
Acceptedby an independent counter-analyst, 12 Nov 2026
In 30 seconds
- What happened
- The Meuse at Chooz stayed under the 22 m3/s figure of a 1998 agreement between France and Belgium for 103 days, and both reactors of the plant were stopped, one from 11 July and the other from 1 August, until late October and early November.
- Why it matters
- The limit was not cooling but the agreement and the safety authority's rule on evaporated water, which hold as long as the plant runs. The data do not show that the stops raised the French power price, which moves with gas-fired output, nor that Belgium was first exposed on price: its premium over France narrowed.
- What to watch
- The lowest 12-day mean flow at Chooz from 12 November to 7 December and any published flow-linked measure at the plant, read on 8 December 2026; then the April 2027 mean flow, and the count of 2027 days under 22 m3/s.
Grand Est nuclear output, daily mean
RTE, via Open Data Réseaux Énergies (ODRÉ) — éCO2mix
The chain
- L1 · ObservedMeuse flow at Chooz, summer and autumn 202612-day mean under the 22 and 20 m3/s of the 1998 agreement
- L2 · Consistent12-day mean near or under the figures of the rulethe operator stops unit 2 (11 July) and unit 1 (1 August), and reconnects them on 26 October and 9 November
- L3 · Observedstops of the two Chooz unitsGrand Est nuclear output
The 12-day mean flow of the Meuse upstream of Chooz stayed under 22 m3/s for 103 days, from 14 July to 24 October 2026, the longest episode of the 2015 to 2026 record, and reached 15.558 m3/s in early October. Both units of the plant were stopped: about 7.2 TWh at rated output by the reported dates. The operator states that the river could still cool the reactors; the limit was the 1998 Franco-Belgian agreement, applied through the safety authority's rule on evaporated water.
Geography: Chooz, Ardennes (Meuse gauge B720000001 and nuclear plant) / Grand Est region (Chooz, Cattenom and Nogent nuclear plants) / Metropolitan France (RTE system) / Belgium (Meuse downstream of Chooz; Elia control area).
Premium · 49 USD a month
This page is the summary. The full analysis is in Premium.
The complete reasoning dossier: 46 constants with their sources, the evidence behind each of the 10 links, 4 conjectures with their checks, 3 dated trajectory and the independent counter-analysis, plus every organisation with its evidence and the full data series as files, for every analysis.
Unlock the full analysis · 49 USD / month → Free preview of the dossier
14-day full refund. Cancel anytime.
Behind this edition: a reasoning dossier of 46 sourced constants, 15 bounded constraints and a 10-link chain (3 untested, 2 contradicted), accepted by an independent reviewer on 2026-11-12. Read the dossier →
The three data series
Data table (the twelve points drawn for each series)
| Period | Value |
|---|---|
| 2026-10-31 | 36.10 |
| 2026-11-01 | 36.54 |
| 2026-11-02 | 36.66 |
| 2026-11-03 | 36.77 |
| 2026-11-04 | 36.15 |
| 2026-11-05 | 35.23 |
| 2026-11-06 | 30.42 |
| 2026-11-07 | 26.79 |
| 2026-11-08 | 23.77 |
| 2026-11-09 | 23.29 |
| 2026-11-10 | 23.16 |
| 2026-11-11 | 23.32 |
| Period | Value |
|---|---|
| 2026-10-31 | 5,177.70 |
| 2026-11-01 | 5,217.20 |
| 2026-11-02 | 5,224.10 |
| 2026-11-03 | 5,229.20 |
| 2026-11-04 | 5,209 |
| 2026-11-05 | 5,231.40 |
| 2026-11-06 | 4,317.30 |
| 2026-11-07 | 3,968 |
| 2026-11-08 | 3,963.50 |
| 2026-11-09 | 4,399.70 |
| 2026-11-10 | 5,433 |
| 2026-11-11 | 5,298.80 |
| Period | Value |
|---|---|
| 2026-10-31 | -1,260.40 |
| 2026-11-01 | -1,184.80 |
| 2026-11-02 | -476.60 |
| 2026-11-03 | -50 |
| 2026-11-04 | 1,407.70 |
| 2026-11-05 | 2,427.50 |
| 2026-11-06 | 1,114.10 |
| 2026-11-07 | 1,051.50 |
| 2026-11-08 | 709.40 |
| 2026-11-09 | -618.10 |
| 2026-11-10 | -1,383.40 |
| 2026-11-11 | -1,684 |
Each panel draws the last twelve points of a series its publisher lets us share, on its own scale. A dashed line marks a threshold of one of the dated hypotheses below. Restricted observations are not sent to the browser.
Why these series belong together
Three open series are read: the 12-day mean flow of the Meuse at Chooz, on which a 1998 Franco-Belgian agreement is written; nuclear output in the Grand Est region; and the physical power flow between Belgium and France. The claim under test: low water stops the reactors, and Belgium is first exposed.
What the river did
The 12-day mean went under 22 m3/s on 14 July at 21.981 [F1] and under 20 on 6 August at 19.83 [F2]. It reached 15.558 on 6 and 7 October [F3] and rose back to 24.653 on 25 October [F5]: 103 consecutive days under 22 and 80 under 20, against 46 and 35 in 2020, the previous longest since 2015. The lowest daily flow was 15.161 m3/s on 1 October [F8]. On 11 November the 12-day mean read 23.32 [F7].
The limit: an agreement and a rule on evaporated water
The agreement signed on 8 September 1998 sets two flows downstream of the plant, on a twelve-day mean: 22 m3/s, from which the first safeguard measures are taken, and 20, below which no aggravation is permitted. It names no reactor. The safety authority's decision 2009-DC-0164 limits the water the two units evaporate to 5 percent of the river flow between 20 and 22, and to nil under 20. EDF words it by unit: one stopped under 22, both under 20. Cooling was not the limit: the operator states that 6.3 m3/s suffice, and the lowest daily flow was 2.4 times that.
What the operator did, and the output withheld
Unit 2 was out of production from 11 July; EDF stopped unit 1 on 1 August to comply with the 1998 framework. On the operator's gauge the 12-day mean read 22.923 [F45] and 20.922 [F46], about 0.9 m3/s above each figure. Grand Est nuclear output fell from 7,585.8 MW on 9 July [F13] to 4,124.0 on 11 July [F14], then to 2,166.5 on 25 October [F17], its lowest day. Unit 1 is reported reconnected on 26 October and unit 2 on 9 November: 207.2 unit-days, about 7.2 TWh at the rated 1,450 MW per unit, not a measure of cost.
What the data do not show
Day-ahead prices (SMARD, restricted, read at source) rose more in France than in Belgium: the Belgian premium averaged 40.3 EUR per MWh from 1 June to 10 July and 9.2 from 2 August to 25 October, which contradicts Belgium as first exposed on price. Belgian imports from France had been falling since May. From 2 August to 11 November the French daily price correlates +0.86 with French gas-fired output and +0.40 with Grand Est nuclear output. Gas-fired output read 6,282.4 MW on 11 November [F32] against 1,223.7 a year earlier [F33].
Hypothesis · The cause of the French rise is untested, and no Chooz share can be sized.
Belgium's two exposures
Power, by contract. EDF's 2022 document lists Electrabel for 21.7 percent of Chooz B1-B2, with volumes set by a broader fleet's availability. Luminus holds 100 MW of drawing rights on the plant: 534,858 MWh in 2023 against 259 MWh in 2022, when the plant was stopped most of the year.
Hypothesis · Deduced: water, by regime. Vivaqua draws 29 percent of its output from the Meuse at Tailfer, where, as it stated in 2022, intake is capped by steps: 120,000 m3 a day, 90,000 between 19 and 16 m3/s, 60,000, then a stop. Official statements put the flow there near 19 m3/s in August and September.
Who gains if it lasts, and what locks it
Hypothesis · If gas-fired plants set the French price, an untested cause, output neither gas-fired nor sold forward earns that price. Two locks: the agreement stays in force as long as the plant operates; winter demand rises by up to 2,400 MW per degree of cold, against 2,900 MW for the two units.
Hypothesis · Deduced: TotalEnergies, GazelEnergie and RWE may gain running hours on gas-fired plants; buyers represented by UNIDEN and Febeliec carry the day-ahead price where indexed on it.
Scenario · If daily flows from 12 to 21 November average under 20.66 m3/s, the level of early November (19.475 on 7 November [F9]), the 12-day mean reads under 22 on 21 November, a first in the heating season since 2015.
What the sources cannot see
No open gauge measures the flow the rule is written on. Both Hub'Eau gauges read the river before the evaporated water is subtracted, up to 2.1 m3/s with two units running: a 12-day mean between 22 and 24.1 leaves the regulatory value undetermined. Per-unit output is absent: the regional series adds Cattenom and Nogent. Stop times rest on news-agency dispatches; reconnection times on one reading of the operator's notices, which refuse automated readers. No series covers the Tailfer flow or a daily gas price.
Kept disagreements and unconfirmed items
Hypothesis · The reviewer holds 2.1 m3/s as a maximum, so the November zone is narrower than 22 to 24.1; the analyst declines a fixed margin, the operator's gauge having read 22.898 on 10 November [F47] with no stop reported. The reviewer finds the April flow weak as a reading of 2027 exposure and attaches no price level to a Chooz date. Whether the regional shortfall is mostly Cattenom and Nogent stays open.
Hypothesis · Unconfirmed: losses reported for 2018 to 2020, from one trade-press source; output withheld in 2025; the 2026 terms of both Belgian contracts; the 2026 steps at Tailfer. This edition judges neither the agreement nor the decision; it is not investment or legal advice.
What to watch
- Meuse at Chooz, 12-day mean of the daily flows against 22 and 20 m3/s, DREAL gauge B720000001 (Hub'Eau; the query carries no end date and returns the values published to the day it is opened)
- Meuse at Chooz, same reading on the operator's gauge B720000004 (Hub'Eau; same query, no end date)
- Operator's Chooz news page: notices of a stop or power reduction naming the Meuse flow (opens in a browser)
- Grand Est nuclear output, RTE éCO2mix regional
Open hypotheses
The record across all analyses →
| Hypothesis | Confidence | Horizon | Status |
|---|---|---|---|
| The lowest 12-day mean of the Meuse at Chooz between 12 November and 7 December 2026 is below 20 m3/s. The safety authority's rule then sets the evaporated flow at nil on the thirteenth day, both units stopped in the operator's wording. Pace required, recomputed from the daily flows to 11 November: a first reading under 20 needs the days from 12 November to average under 18.26 m3/s by 21 November, under 19.22 by 22 November, or under 20 over the twelve days to 23 November; earlier dates need lower flows still (under 16.90 by 20 November). All are below the mean flow of 2 to 7 November (20.55; lowest day 19.475 on 7 November). No precedent on these dates in 2015 to 2025. Reading rule for the flow: Hub'Eau, lowest value of the 12-day mean of the daily mean flows (QmnJ) on any day from 12 November to 7 December 2026, read on 8 December 2026 with the values as then published (raw or pre-validated). The Hub'Eau query carries no end date and returns the values published to the day it is opened. If a daily value is missing on that day, the reading is made on the values published and is void if no 12-day mean of the window can be formed. The DREAL gauge B720000001 decides the branch; the operator's gauge B720000004 is read beside it and a difference of branch between the two is reported as such. Neither gauge subtracts the evaporated flow: the downstream value of the rule is lower by up to 2.1 m3/s with two units running, so a reading from 22 to 24.1 leaves it undetermined. The four branches of this reading are read in the order written (lowest mean under 20; from 20 to under 22; at or above 22 with a published flow-linked measure; at or above 22 with none) and the first whose condition holds is the branch. A 12-day mean under 22 m3/s or a measure dated 8 to 31 December 2026 belongs to no branch: those days are read in a follow-up, and the branch read on 8 December stands. Read beside the branch and not defining it: Grand Est nuclear output, mean of the daily means over the 12 days after the first reading under 20, read on 24 December 2026, against 1,098.8 to 4,998.8 MW. | LOW | 2026-12-08 | OPEN |
| The lowest 12-day mean of the Meuse at Chooz between 12 November and 7 December 2026 is below 22 m3/s and at or above 20 m3/s. The safety authority's rule then limits the evaporated flow to 5 percent of the river flow, about one unit's worth: one unit stopped in the operator's wording, or power reductions that may bear on both units. Pace required, recomputed from the daily flows to 11 November: a first reading under 22 needs the days from 12 November to average under 19.56 m3/s by 20 November or under 20.66 by 21 November; earlier dates need lower flows still (under 19.05 by 19 November). The second figure is the level of 2 to 7 November (mean 20.55); the first lies below it (one day since 31 October, 19.475 on 7 November). Local precedent for a stay between the two figures: 14 July to 5 August 2026 and the 2019 episode. Reading rule for the flow: Hub'Eau, lowest value of the 12-day mean of the daily mean flows (QmnJ) on any day from 12 November to 7 December 2026, read on 8 December 2026 with the values as then published (raw or pre-validated). The Hub'Eau query carries no end date and returns the values published to the day it is opened. If a daily value is missing on that day, the reading is made on the values published and is void if no 12-day mean of the window can be formed. The DREAL gauge B720000001 decides the branch; the operator's gauge B720000004 is read beside it and a difference of branch between the two is reported as such. Neither gauge subtracts the evaporated flow: the downstream value of the rule is lower by up to 2.1 m3/s with two units running, so a reading from 22 to 24.1 leaves it undetermined. The four branches of this reading are read in the order written (lowest mean under 20; from 20 to under 22; at or above 22 with a published flow-linked measure; at or above 22 with none) and the first whose condition holds is the branch. A 12-day mean under 22 m3/s or a measure dated 8 to 31 December 2026 belongs to no branch: those days are read in a follow-up, and the branch read on 8 December stands. Read beside the branch and not defining it: Grand Est nuclear output, mean of the daily means over the 12 days after the first reading under 22, read on 24 December 2026, against 2,548.8 to 6,448.8 MW. | LOW | 2026-12-08 | OPEN |
| The lowest 12-day mean of the Meuse at Chooz between 12 November and 7 December 2026 is at or above 22 m3/s, and at least one published document attributes a stop or a power reduction of a Chooz unit between the same dates to the Meuse flow. The published document, not the flow series, separates this branch from the next. Precedents: 11 July 2026 (22.995 on the DREAL gauge, 22.923 on the operator's) and autumn 2018 (22.161); counter-case: 10 November 2026 (22.898 on the operator's gauge, no stop reported). Reading rule for the flow: Hub'Eau, lowest value of the 12-day mean of the daily mean flows (QmnJ) on any day from 12 November to 7 December 2026, read on 8 December 2026 with the values as then published (raw or pre-validated). The Hub'Eau query carries no end date and returns the values published to the day it is opened. If a daily value is missing on that day, the reading is made on the values published and is void if no 12-day mean of the window can be formed. The DREAL gauge B720000001 decides the branch; the operator's gauge B720000004 is read beside it and a difference of branch between the two is reported as such. Neither gauge subtracts the evaporated flow: the downstream value of the rule is lower by up to 2.1 m3/s with two units running, so a reading from 22 to 24.1 leaves it undetermined. The four branches of this reading are read in the order written (lowest mean under 20; from 20 to under 22; at or above 22 with a published flow-linked measure; at or above 22 with none) and the first whose condition holds is the branch. A 12-day mean under 22 m3/s or a measure dated 8 to 31 December 2026 belongs to no branch: those days are read in a follow-up, and the branch read on 8 December stands. Reading rule for the measure, as it stands on 8 December 2026: at least one published document dated 12 November to 7 December 2026 that reports a stop or a power reduction of a Chooz unit, whatever its size or duration, and names the Meuse flow, the 1998 agreement or the limit on evaporated water as its reason. One document at any of three kinds of source counts: a notice on the operator's Chooz news page (https://www.edf.fr/la-centrale-nucleaire-de-chooz/les-actualites-de-la-centrale-nucleaire-de-chooz), opened in an ordinary browser (the page refuses automated readers), its title, date and sentence copied into the reading; a news-agency dispatch as republished by the regional and national press (no single public feed: the reading names the titles searched); or a generation unavailability message for Chooz 1 or Chooz 2 that states the cause, on the grid operator's public page of unavailability messages (https://iip.cloud-rte-france.com/production-unavailability; opened once on 12 November 2026: it answers an automated reader, but its messages load by script and are read in a browser). A trade-press article alone does not count and is reported beside the reading. A fall of output with no published reason is reported as unattributed and does not change the branch. A document published after 8 December about days of the window is reported on 24 December as a late correction. The reading names, for each source, who opened it (a person in a browser, or an agent) and on which date. If the operator's page was not opened, the fourth branch is recorded as 'no document found in dispatches and unavailability messages; operator's page not opened' and stays provisional until a person opens it. Read beside the branch and not defining it: Grand Est nuclear output, mean of the daily means of 12 November to 7 December 2026, read on 8 December 2026, against 2,548.8 to 8,574.9 MW; the regional series cannot separate this branch from the next. | LOW | 2026-12-08 | OPEN |
| The lowest 12-day mean of the Meuse at Chooz between 12 November and 7 December 2026 is at or above 22 m3/s, and no published document attributes a stop or a power reduction of a Chooz unit between the same dates to the Meuse flow. The branch states that nothing was published, not that nothing happened. Seasonal precedent: every year from 2015 to 2025, lowest value of these dates 28.8 m3/s, none of them starting from a base flow near 20 m3/s. Reading rule for the flow: Hub'Eau, lowest value of the 12-day mean of the daily mean flows (QmnJ) on any day from 12 November to 7 December 2026, read on 8 December 2026 with the values as then published (raw or pre-validated). The Hub'Eau query carries no end date and returns the values published to the day it is opened. If a daily value is missing on that day, the reading is made on the values published and is void if no 12-day mean of the window can be formed. The DREAL gauge B720000001 decides the branch; the operator's gauge B720000004 is read beside it and a difference of branch between the two is reported as such. Neither gauge subtracts the evaporated flow: the downstream value of the rule is lower by up to 2.1 m3/s with two units running, so a reading from 22 to 24.1 leaves it undetermined. The four branches of this reading are read in the order written (lowest mean under 20; from 20 to under 22; at or above 22 with a published flow-linked measure; at or above 22 with none) and the first whose condition holds is the branch. A 12-day mean under 22 m3/s or a measure dated 8 to 31 December 2026 belongs to no branch: those days are read in a follow-up, and the branch read on 8 December stands. Reading rule for the measure, as it stands on 8 December 2026: at least one published document dated 12 November to 7 December 2026 that reports a stop or a power reduction of a Chooz unit, whatever its size or duration, and names the Meuse flow, the 1998 agreement or the limit on evaporated water as its reason. One document at any of three kinds of source counts: a notice on the operator's Chooz news page (https://www.edf.fr/la-centrale-nucleaire-de-chooz/les-actualites-de-la-centrale-nucleaire-de-chooz), opened in an ordinary browser (the page refuses automated readers), its title, date and sentence copied into the reading; a news-agency dispatch as republished by the regional and national press (no single public feed: the reading names the titles searched); or a generation unavailability message for Chooz 1 or Chooz 2 that states the cause, on the grid operator's public page of unavailability messages (https://iip.cloud-rte-france.com/production-unavailability; opened once on 12 November 2026: it answers an automated reader, but its messages load by script and are read in a browser). A trade-press article alone does not count and is reported beside the reading. A fall of output with no published reason is reported as unattributed and does not change the branch. A document published after 8 December about days of the window is reported on 24 December as a late correction. The reading names, for each source, who opened it (a person in a browser, or an agent) and on which date. If the operator's page was not opened, the fourth branch is recorded as 'no document found in dispatches and unavailability messages; operator's page not opened' and stays provisional until a person opens it. If no such document exists on 8 December, this branch holds. Read beside the branch and not defining it: Grand Est nuclear output, mean of the daily means of 12 November to 7 December 2026, read on 8 December 2026, against 3,998.8 to 8,574.9 MW; that band tests the six other units of the region, not the river. | HIGH | 2026-12-08 | OPEN |
| The mean daily flow of the Meuse at Chooz in April 2027 is below 90 m3/s. Years on file under 90: 2017 (no day of the 12-day mean under 22 that summer), 2025 (3 days) and 2026 (103 days). Reading rule: Hub'Eau station B720000001, arithmetic mean of the daily mean flows (QmnJ) of 1 to 30 April 2027, read on 5 May 2027 with the values as then published. The Hub'Eau query carries no end date and returns the values published to the day it is opened. If a daily value is missing on that day, the mean is taken on the April days published and their number is stated; the reading is void if none is published. The three branches of this reading are read in the order written (below 90; from 90 to below 150; at or above 150) and the first whose condition holds is the branch. April means on file: 67.9 (2017), 70.4 (2025), 85.0 (2026), 97.6 (2021), 105.7 (2020), 116.0 (2019), 121.8 (2022), 156.9 (2015), 177.5 (2018), 186.1 (2016), 233.0 (2024), 238.7 (2023). The April mean is a starting level, not a reading of the summer: it does not separate the episode years from the others (67.9 in 2017 with no day under 22; 105.7 in 2020 with 46 days). Read beside the branch and not defining it: the May 2027 mean flow, read on 5 June 2027, against 42.7 to 60.5 m3/s, the range of the precedent years; the May ranges of the three branches overlap. | MEDIUM | 2027-05-05 | OPEN |
| The mean daily flow of the Meuse at Chooz in April 2027 is below 150 m3/s and at or above 90 m3/s. Years on file in this band: 2019 (21 days of the 12-day mean under 22), 2020 (46), 2021 (0) and 2022 (38). Reading rule: Hub'Eau station B720000001, arithmetic mean of the daily mean flows (QmnJ) of 1 to 30 April 2027, read on 5 May 2027 with the values as then published. The Hub'Eau query carries no end date and returns the values published to the day it is opened. If a daily value is missing on that day, the mean is taken on the April days published and their number is stated; the reading is void if none is published. The three branches of this reading are read in the order written (below 90; from 90 to below 150; at or above 150) and the first whose condition holds is the branch. April means on file: 67.9 (2017), 70.4 (2025), 85.0 (2026), 97.6 (2021), 105.7 (2020), 116.0 (2019), 121.8 (2022), 156.9 (2015), 177.5 (2018), 186.1 (2016), 233.0 (2024), 238.7 (2023). The April mean is a starting level, not a reading of the summer: it does not separate the episode years from the others (67.9 in 2017 with no day under 22; 105.7 in 2020 with 46 days). Read beside the branch and not defining it: the May 2027 mean flow, read on 5 June 2027, against 57.5 to 105.3 m3/s, the range of the precedent years; the May ranges of the three branches overlap. | MEDIUM | 2027-05-05 | OPEN |
| The mean daily flow of the Meuse at Chooz in April 2027 is at or above 150 m3/s. Years on file at or above 150: 2015, 2016, 2018, 2023 and 2024, none with a day of the DREAL 12-day mean under 22; in 2018 a unit was nevertheless stopped in autumn. Reading rule: Hub'Eau station B720000001, arithmetic mean of the daily mean flows (QmnJ) of 1 to 30 April 2027, read on 5 May 2027 with the values as then published. The Hub'Eau query carries no end date and returns the values published to the day it is opened. If a daily value is missing on that day, the mean is taken on the April days published and their number is stated; the reading is void if none is published. The three branches of this reading are read in the order written (below 90; from 90 to below 150; at or above 150) and the first whose condition holds is the branch. April means on file: 67.9 (2017), 70.4 (2025), 85.0 (2026), 97.6 (2021), 105.7 (2020), 116.0 (2019), 121.8 (2022), 156.9 (2015), 177.5 (2018), 186.1 (2016), 233.0 (2024), 238.7 (2023). The April mean is a starting level, not a reading of the summer: it does not separate the episode years from the others (67.9 in 2017 with no day under 22; 105.7 in 2020 with 46 days). Read beside the branch and not defining it: the May 2027 mean flow, read on 5 June 2027, against 86.9 to 174.0 m3/s, the range of the precedent years; the May ranges of the three branches overlap. | MEDIUM | 2027-05-05 | OPEN |
| In 2027 the DREAL 12-day mean of the Meuse at Chooz is under 22 m3/s on 46 days or more. Precedents: 2020 and 2026, two of twelve years. Reading rule: Hub'Eau station B720000001, number of days between 1 January and 31 December 2027 on which the 12-day mean of the daily flows is under 22 m3/s, read on 4 January 2028 with the values as then published; days after 31 October count (none on file from 2015 to 2026). The Hub'Eau query carries no end date and returns the values published to the day it is opened. If daily values are missing on that day, the count is made on the 12-day means that can be formed and the days without one are stated; the reading is void if no 12-day mean can be formed. The three branches of this reading are read in the order written (46 days or more; 1 to 45; none) and the first whose condition holds is the branch. Counts on file: 0 in 2015 to 2018, 2021, 2023 and 2024; 21 (2019), 46 (2020), 38 (2022), 3 (2025), 103 (2026). Read beside the branch and not defining it: the days under 24.1 m3/s, the edge of the regulatory zone with two units running (31 in 2018, then 38, 55, 56, 16 and 112 for the same years). The output withheld at Chooz over 2027 is an order of magnitude and is not scored. | LOW | 2028-01-04 | OPEN |
| In 2027 the DREAL 12-day mean of the Meuse at Chooz is under 22 m3/s on 1 to 45 days. Precedents: 2019, 2022 and 2025, three of twelve years. Reading rule: Hub'Eau station B720000001, number of days between 1 January and 31 December 2027 on which the 12-day mean of the daily flows is under 22 m3/s, read on 4 January 2028 with the values as then published; days after 31 October count (none on file from 2015 to 2026). The Hub'Eau query carries no end date and returns the values published to the day it is opened. If daily values are missing on that day, the count is made on the 12-day means that can be formed and the days without one are stated; the reading is void if no 12-day mean can be formed. The three branches of this reading are read in the order written (46 days or more; 1 to 45; none) and the first whose condition holds is the branch. Counts on file: 0 in 2015 to 2018, 2021, 2023 and 2024; 21 (2019), 46 (2020), 38 (2022), 3 (2025), 103 (2026). Read beside the branch and not defining it: the days under 24.1 m3/s, the edge of the regulatory zone with two units running (31 in 2018, then 38, 55, 56, 16 and 112 for the same years). The output withheld at Chooz over 2027 is an order of magnitude and is not scored. | MEDIUM | 2028-01-04 | OPEN |
| In 2027 the DREAL 12-day mean of the Meuse at Chooz is under 22 m3/s on no day. Precedents: seven of twelve years; six of them have no day under 24.1 m3/s either and no reported loss; the seventh, 2018, has 31 days under 24.1 and a reported stop. Reading rule: Hub'Eau station B720000001, number of days between 1 January and 31 December 2027 on which the 12-day mean of the daily flows is under 22 m3/s, read on 4 January 2028 with the values as then published; days after 31 October count (none on file from 2015 to 2026). The Hub'Eau query carries no end date and returns the values published to the day it is opened. If daily values are missing on that day, the count is made on the 12-day means that can be formed and the days without one are stated; the reading is void if no 12-day mean can be formed. The three branches of this reading are read in the order written (46 days or more; 1 to 45; none) and the first whose condition holds is the branch. Counts on file: 0 in 2015 to 2018, 2021, 2023 and 2024; 21 (2019), 46 (2020), 38 (2022), 3 (2025), 103 (2026). Read beside the branch and not defining it: the days under 24.1 m3/s, the edge of the regulatory zone with two units running (31 in 2018, then 38, 55, 56, 16 and 112 for the same years). The output withheld at Chooz over 2027 is an order of magnitude and is not scored. | MEDIUM | 2028-01-04 | OPEN |
Who is concerned
23 organisations, each with a verbatim public quote and a dated source; evidenced when the quote itself places the organisation on this chain, deduced when the exposure is inferred from its business. Where the universe came from →
Data you can manipulate
Free downloads carry the last twelve observations of each series its publisher lets us share; the full series, the dossier and the exports are in Premium. Series restricted by their publisher are cited with a link to read them at source.
| Series | Family | Unit | Publisher · rights | Download (sample) |
|---|---|---|---|---|
| Day-ahead wholesale electricity price, bidding zone Belgium, daily average as published by SMARD (filter 4996; SMARD takes the data from ENTSO-E); delivery day in Central European time | Economic | EUR/MWh | Bundesnetzagentur | SMARD.de · restricted | cited, read at source |
| Belgium minus France day-ahead wholesale electricity price, daily average (derived) · derived: diff of be-day-ahead-eur-mwh, fr-day-ahead-eur-mwh | Economic | EUR/MWh | Bundesnetzagentur | SMARD.de · restricted | cited, read at source |
| Physical flow on the Belgium-France interconnection (Elia ods026, control area France), daily mean of quarter-hour values per Belgian local day, null below 90 % of the day's values; positive = Belgium exports to France, negative = Belgium imports from France | Flow | MW | Elia (Belgian transmission system operator) — Open Data Portal · open attribution | JSON · CSV |
| Day-ahead wholesale electricity price, bidding zone France, daily average as published by SMARD (filter 254; SMARD takes the data from ENTSO-E); delivery day in Central European time | Economic | EUR/MWh | Bundesnetzagentur | SMARD.de · restricted | cited, read at source |
| Net commercial exchange between France and the Germany-Belgium zone (one combined column in éCO2mix; Belgium is not published separately), positive = France imports, negative = France exports; daily mean of the published values per French local day, null below 90 % of the day's values; definitive data for 2025, consolidated from 2026-01-01 to 2026-08-31, real-time telemetry (provisional, to be replaced) from 2026-09-01 | Flow | MW | RTE, via Open Data Réseaux Énergies (ODRÉ) — éCO2mix · open attribution | JSON · CSV |
| Gas-fired generation in metropolitan France, éCO2mix national; daily mean of the published values per French local day, null below 90 % of the day's values; definitive data for 2025, consolidated from 2026-01-01 to 2026-08-31, real-time telemetry (provisional, to be replaced) from 2026-09-01 | Flow | MW | RTE, via Open Data Réseaux Énergies (ODRÉ) — éCO2mix · open attribution | JSON · CSV |
| Nuclear generation in metropolitan France, éCO2mix national; daily mean of the published values per French local day, null below 90 % of the day's values; definitive data for 2025, consolidated from 2026-01-01 to 2026-08-31, real-time telemetry (provisional, to be replaced) from 2026-09-01 | Flow | MW | RTE, via Open Data Réseaux Énergies (ODRÉ) — éCO2mix · open attribution | JSON · CSV |
| Nuclear generation in the Grand Est region (Cattenom, Chooz and Nogent plants), éCO2mix regional; daily mean of the published values per French local day, null below 90 % of the day's values; definitive data for 2025, consolidated from 2026-01-01 to 2026-08-31, real-time telemetry (provisional, to be replaced) from 2026-09-01 | Flow | MW | RTE, via Open Data Réseaux Énergies (ODRÉ) — éCO2mix · open attribution | JSON · CSV |
| Meuse flow at Chooz, station B720000004 'La Meuse à Chooz - Trou du Diable "EDF"', which the publisher describes as the reference for real-time flows computed from the EDF sensor; the station record does not state a position upstream or downstream of the plant and gives the same coordinates as station B720000001: mean of the daily mean flows over the twelve consecutive days ending on the period date, computed by the collector from meuse-chooz-b720000004-flow-m3s; null for the first eleven days (first value 2026-07-03). Not the operator's regulatory computation; inputs are raw or pre-validated values, may be revised | Physical | m3/s | Hub'Eau, Eaufrance (SCHAPI, DREAL, CNR) — hydrométrie · open attribution | JSON · CSV |
| Meuse daily mean flow (QmnJ) at Chooz, station B720000004 'La Meuse à Chooz - Trou du Diable "EDF"', which the publisher describes as the reference for real-time flows computed from the EDF sensor; the station record does not state a position upstream or downstream of the plant and gives the same coordinates as station B720000001; source values in L/s divided by 1000; published from 2026-06-22 only; no validated value: 31 days pre-validated (30 good, 1 doubtful), 112 days raw (59 not qualified, 53 doubtful), may be revised | Physical | m3/s | Hub'Eau, Eaufrance (SCHAPI, DREAL, CNR) — hydrométrie · open attribution | JSON · CSV |
| Meuse flow at Chooz (station B720000001, upstream of the plant): mean of the daily mean flows over the twelve consecutive days ending on the period date, computed by the collector from meuse-chooz-flow-m3s; null for the first eleven days. Not the operator's regulatory computation, which uses the station downstream of the plant | Physical | m3/s | Hub'Eau, Eaufrance (SCHAPI, DREAL, CNR) — hydrométrie · open attribution | JSON · CSV |
| Meuse daily mean flow (QmnJ) at Chooz, station B720000001 'Trou du Diable DREAL', upstream of the power plant (does not include the flow evaporated by the cooling towers); source values in L/s divided by 1000; 2026 values pre-validated or raw, may be revised | Physical | m3/s | Hub'Eau, Eaufrance (SCHAPI, DREAL, CNR) — hydrométrie · open attribution | JSON · CSV |
Figures, recomputed at every build
| Id | Series | Statistic | Value |
|---|---|---|---|
| F1 | Meuse flow at Chooz, 12-day mean | value · 2026-07-14 | 21.98 m3/s |
| F2 | Meuse flow at Chooz, 12-day mean | value · 2026-08-06 | 19.83 m3/s |
| F3 | Meuse flow at Chooz, 12-day mean | min | 15.56 m3/s |
| F4 | Meuse flow at Chooz, 12-day mean | value · 2026-10-24 | 19.73 m3/s |
| F5 | Meuse flow at Chooz, 12-day mean | value · 2026-10-25 | 24.65 m3/s |
| F6 | Meuse flow at Chooz, 12-day mean | value · 2026-11-03 | 36.77 m3/s |
| F7 | Meuse flow at Chooz, 12-day mean | last | 23.32 m3/s |
| F8 | Meuse flow at Chooz, daily mean | min | 15.16 m3/s |
| F9 | Meuse flow at Chooz, daily mean | value · 2026-11-07 | 19.48 m3/s |
| F10 | Meuse flow at Chooz, daily mean | last | 28.59 m3/s |
| F11 | Meuse flow at Chooz, 12-day mean | value · 2026-07-11 | 23.00 m3/s |
| F12 | Meuse flow at Chooz, 12-day mean | value · 2026-08-01 | 20.70 m3/s |
| F13 | Grand Est nuclear output, daily mean | value · 2026-07-09 | 7,585.80 MW |
| F14 | Grand Est nuclear output, daily mean | value · 2026-07-11 | 4,124 MW |
| F15 | Grand Est nuclear output, daily mean | value · 2026-07-31 | 5,992.50 MW |
| F16 | Grand Est nuclear output, daily mean | value · 2026-08-02 | 3,274.20 MW |
| F17 | Grand Est nuclear output, daily mean | min | 2,166.50 MW |
| F18 | Grand Est nuclear output, daily mean | value · 2026-10-29 | 5,160.90 MW |
| F19 | Grand Est nuclear output, daily mean | value · 2026-11-05 | 5,231.40 MW |
| F20 | Grand Est nuclear output, daily mean | value · 2026-11-08 | 3,963.50 MW |
| F21 | Grand Est nuclear output, daily mean | last | 5,298.80 MW |
| F22 | Grand Est nuclear output, daily mean | value · 2025-11-11 | 7,027.30 MW |
| F23 | Grand Est nuclear output, daily mean | max | 10,826.50 MW |
| F24 | Day-ahead power price, France | max | 334.12 EUR/MWh |
| F25 | Day-ahead power price, Belgium | max | 335.27 EUR/MWh |
| F26 | Day-ahead price gap, Belgium minus France | max | 104.53 EUR/MWh |
| F27 | Day-ahead price gap, Belgium minus France | value · 2026-09-13 | 60.78 EUR/MWh |
| F28 | Day-ahead price gap, Belgium minus France | min | -11.26 EUR/MWh |
| F29 | Belgium to France physical flow, daily mean | min | -4,148.30 MW |
| F30 | Belgium to France physical flow, daily mean | value · 2026-11-05 | 2,427.50 MW |
| F31 | Belgium to France physical flow, daily mean | last | -1,684 MW |
| F32 | France gas-fired output, daily mean | last | 6,282.40 MW |
| F33 | France gas-fired output, daily mean | value · 2025-11-11 | 1,223.70 MW |
| F34 | Meuse flow at Chooz, daily mean | value · 2026-11-08 | 23.70 m3/s |
| F35 | Meuse flow at Chooz, daily mean | value · 2026-11-09 | 30.50 m3/s |
| F36 | Meuse flow at Chooz, daily mean | value · 2026-11-10 | 28.82 m3/s |
| F37 | Meuse flow at Chooz, 12-day mean | value · 2020-08-21 | 20.42 m3/s |
| F38 | Meuse flow at Chooz, 12-day mean | value · 2020-08-24 | 19.87 m3/s |
| F39 | Meuse flow at Chooz, 12-day mean | value · 2018-10-19 | 22.16 m3/s |
| F40 | Meuse flow at Chooz, 12-day mean | value · 2020-09-28 | 23.35 m3/s |
| F41 | Grand Est nuclear output, daily mean | value · 2026-10-26 | 3,992.60 MW |
| F42 | France nuclear output, daily mean | last | 41,974.40 MW |
| F43 | France nuclear output, daily mean | value · 2025-11-11 | 44,254.90 MW |
| F45 | Meuse flow at Chooz, EDF sensor station, 12-day mean | value · 2026-07-11 | 22.92 m3/s |
| F46 | Meuse flow at Chooz, EDF sensor station, 12-day mean | value · 2026-08-01 | 20.92 m3/s |
| F47 | Meuse flow at Chooz, EDF sensor station, 12-day mean | value · 2026-11-10 | 22.90 m3/s |
| F48 | Meuse flow at Chooz, EDF sensor station, 12-day mean | last | 23.06 m3/s |
| F49 | Grand Est nuclear output, daily mean | value · 2026-09-15 | 3,827.70 MW |
| F50 | Grand Est nuclear output, daily mean | value · 2026-09-14 | 4,239.60 MW |
| F51 | Grand Est nuclear output, daily mean | value · 2026-11-10 | 5,433 MW |
| F52 | Grand Est nuclear output, daily mean | value · 2025-12-02 | 9,191.50 MW |
| F53 | Grand Est nuclear output, daily mean | value · 2025-12-08 | 7,548.70 MW |
| F54 | Meuse flow at Chooz, EDF sensor station, daily mean | value · 2026-07-11 | 21.31 m3/s |
Every number above is recomputed from the published observations before this page is built; a mismatch stops the build. Sources: Bundesnetzagentur | SMARD.de, Elia (Belgian transmission system operator) — Open Data Portal, RTE, via Open Data Réseaux Énergies (ODRÉ) — éCO2mix, Hub'Eau, Eaufrance (SCHAPI, DREAL, CNR) — hydrométrie. Signal data sheet →
PREMIUM · 49 USD A MONTH
The full dossier and files for this analysis are in Premium.
Every published analysis as files you can load: the reasoning dossier with its sources, the organisations with their evidence, and the series their publishers allow us to share, plus a key for the JSON API and the MCP server.
What Premium includes →Separate from the free weekly. Summaries and samples stay free, without an account.