Tempo: how a red day is made

In March 2026, thirteen red Tempo days followed one another, with weekends excluded, even though residual electricity demand was relatively low. France’s energy regulator, the Commission de régulation de l’énergie (CRE), explains why: the season’s allocation of twenty-two red days had to be used up. For subscribers, each colour still brought its price and the usual demands on their daily routines. A calendar rule had become a decisive part of the signal sent to consumers. G01

On 21 October 2026, the CRE adopted a new selection method. Published in the Official Journal on 24 October, it will apply to red days from 1 November 2026, alongside the exclusion of 24 and 31 December. The official notice arrangements will change a year later, on 1 November 2027. The algorithm and the information available to help customers plan their day will therefore change on different schedules. G01

The investigation leads to a fundamental economic choice: how can the most expensive days be reserved for times when reducing consumption is useful, while giving customers a calendar they can live with? RTE’s published rules provide a much more precise answer than an app’s colour display. They combine electricity forecasts, the remaining allocation of days, monthly limits and the value of waiting. The tariff then converts that selection into euros. G04 R02 T01

The reform changes how expensive days are chosen. Its effect on a bill will still depend on how much electricity the customer uses on those days. This dossier follows that journey from the electricity system to a customer’s budget, distinguishing observed outcomes, RTE’s simulations and our illustrative examples.

Research closed on 29 October 2026. The selection reform has been adopted and has yet to take effect. Monetary examples use EDF’s Tarif Bleu schedule effective from 1 August 2026 in mainland France. Fictional scenarios are identified in the text, figures and data.

A public colour, individual supply contracts

Tempo spreads a commitment across the year. Some periods offer relatively low unit prices; others make consumption much more expensive and encourage customers to shift or reduce their use. This demand reduction, known in French as “effacement”, means consumption avoided during a given period compared with what would otherwise have occurred. When an activity is simply rescheduled, the energy may be consumed later. When it is forgone, the total volume falls. The distinction matters both for the electricity system and for calculating a saving. T09 T10

A Tempo year runs from 1 September to 31 August. Its standard calendar contains 22 red days, 43 white days and the remaining days in blue: 300 in a 365-day year. Red days are selected between November and March, on weekdays excluding national public holidays. White days can occur over a broader period, while Sundays remain blue. The number of red days is therefore a rule of the product, known before the winter’s weather is known. G04 G01 T07

RTE, France’s electricity transmission system operator, has selected the colours since 1 November 2014. The transfer from EDF was intended to let several suppliers offer contracts based on a shared signal. The CRE sets the framework; RTE applies the method and publishes the colour; suppliers sell the offers that attach a price to it. For regulated tariffs, setting and adopting those prices follows a separate public procedure. These responsibilities meet on the customer’s bill, but operate at different stages. G05 T03 T05

This arrangement addresses competition as well as electricity-system needs. An entrant can use the common calendar without relying on a competitor’s daily commercial decision. Customers on different offers receive the same colour signal, with prices and conditions that may vary according to their contract. Every monetary calculation in this investigation therefore refers explicitly to EDF Tarif Bleu Tempo. G01 T09

The scale of the scheme makes its rules worth examining closely. The CRE reports about 900,000 customers in 2025 across the regulated residential and non-residential Tempo options. That figure has a specific scope: it covers neither households alone as of October 2026 nor all market offers using the signal. It nevertheless shows that a change in selection affects an economic relationship already in widespread use. G01

Twenty-two days to allocate through an unknown winter

Choosing the relevant days once winter was over would be relatively straightforward: rank them and retain those with the highest values under the chosen criterion. RTE must decide before each day, with imperfect knowledge of what lies ahead. Selecting a red day now reduces the stock available for a potentially more demanding period. Waiting preserves that option, while increasing the risk of having to allocate the remaining days at the end of the season.

The Tempo method documented in January 2025 makes this tension explicit. It starts with forecast national consumption and subtracts forecast wind and photovoltaic generation. This residual demand, also called net demand, is the electricity requirement left to cover after those two sources of production. The algorithm uses an average over the Tempo day, from 06:00 to 06:00 the following morning, then normalises the data to make them comparable with the historical observations used for calibration. G04

The decision then passes through thresholds that depend on progress through the year and the remaining stock of colour allocations. A threshold is the level at which a day becomes valuable enough to select. In the documented model, having more red days left or moving closer to the end of the season changes that level. The colour therefore reflects both the forecast for the next day and decisions already made during the winter. G04

The calendar adds another constraint. A Saturday, Sunday or national public holiday removes an opportunity to allocate a red day. The stock must also be exhausted. RTE’s note states that, at the end of the period, the rules can lead to a day being selected even when the required demand threshold has not been reached. This is clearance of the remaining quota, or “apurement”: days are selected to meet a numerical obligation, potentially with less value for the electricity system. G04 G01

The March 2026 episode now has a concrete explanation. According to the CRE, 13 of the 22 red days were selected in succession, excluding weekends, to exhaust the remaining stock despite relatively low residual demand. That represents 59.09% of the annual quota. The limit of five consecutive red days is compatible with this run of thirteen weekdays: weekends break the sequence of calendar days. The documents examined establish the aggregate finding; this investigation did not obtain the primary daily series for the episode. G01 G04

There is independent confirmation of the winter’s mildness. Météo-France ranks December 2025 to February 2026 as the fourth-warmest winter since 1900. That meteorological period differs from the November–March electricity window. It provides context. Explaining each individual selection would still require the demand, wind and solar forecasts available at the time, together with the remaining stock of days. W01

That distinction matters when judging the algorithm. A comparison with demand as ultimately measured combines forecast quality and selection quality. A decision may have been consistent with the information available and then prove less useful because the weather forecast was wrong. Conversely, a day may be selected to satisfy a known constraint despite a low demand forecast. The CRE’s assessment identifies precisely this second situation for the thirteen March days. G01

The hourly peak becomes the selection criterion

The reform aligns red Tempo days with peak-period days, or PP days, used in the capacity mechanism. This mechanism organises the availability of generation and demand response during periods that matter to the system. PP days are used, among other things, to assess participants’ contributions and available capacity. The new arrangements also select twenty-two days between November and March, making it easier to align the signals. G01 R02

The alignment comes with a change in measurement. RTE’s PP note of 30 September 2026 uses the forecast hourly peak of residual demand. It starts with forecast total French demand before demand response, then subtracts forecast wind and photovoltaic generation. The former daily Tempo average and this hourly maximum capture different features of a day. Measuring demand before demand response preserves the underlying requirement that reductions in consumption are intended to address. R02 G04

Two entirely fictional profiles make the difference clear. The first day has twelve hours at 70 gigawatts of residual demand and twelve at 50: its average is 60 GW, its peak 70 GW. The second has one hour at 85 GW and twenty-three at 45: its average is about 46.67 GW, its peak 85 GW. Ranking the days by their averages puts the first ahead. Ranking them by their maxima reverses the order.

These profiles illustrate the statistics involved. RTE’s full selection also depends on thresholds, remaining stock, date restrictions and the forecasts available. The examples nevertheless show why the phrase “high-demand days” leaves an essential decision unexplained: which feature of demand is the method trying to capture? Sustained high demand and a brief peak place different patterns of demand on the system. G04 R02

Residual demand itself has a specific definition. With total demand unchanged, lower wind or photovoltaic output increases the residual. Outdoor temperature, considered on its own, therefore explains only part of the calculation. Assessing the risk of power cuts also requires examining available resources, potential imports and system margins. Ecowatt follows that security-of-supply logic; Tempo provides a tariff incentive. R02 R06 R07

The hours also need to be distinguished. The documented PP windows for the capacity mechanism are 07:00–10:00 and 17:00–20:00. Tempo peak-tariff hours run from 06:00 to 22:00. Aligning the selected days preserves these different uses of time: the capacity windows leave the hours billed at the red peak rate unchanged. R02 T07

Setting bounds before spring arrives

The new system adds a more explicit seasonal structure. RTE’s note publishes monthly cumulative bounds: a minimum and maximum number of PP days that must already have been selected by the end of each month. Within those bounds, a guide curve indicates the intended path. The bounds limit how heavily selections can be concentrated; the guide influences daily decisions. R02

At the end of November, the cumulative total can range from zero to three, with a guide value of one. At the end of December, all three values meet at six. At the end of January, the range is thirteen to nineteen, around a guide value of sixteen. By the end of February, twenty to twenty-two days must have been selected, with a guide value of twenty-one. The total reaches twenty-two at the end of March. Every figure is a cumulative month-end value. R02

Data and calculations

Bounds published by RTE. The guide is a distribution reference; lines between month ends aid reading.

PP days accumulated since the start of the season

Month end	Minimum	Guide	Maximum
November	0	1	3
December	6	6	6
January	13	16	19
February	20	21	22
March	22	22	22

The clearest consequence concerns spring. Under these monthly constraints, a minimum of twenty PP days selected before the end of February leaves at most two days for March. That prevents a reserve of thirteen PP days from waiting to be used in the final month. It also moves part of the constraint to earlier deadlines, especially the requirement to reach six by the end of December. R02

The guide corresponds to one, five, ten, five and one days per month, from November to March. These are distribution targets, which RTE distinguishes from the selections that will actually occur. Reading them as five guaranteed monthly quotas would erase the flexibility allowed by the bounds. The requirement for six cumulative days by the end of December has a different status: the minimum and maximum coincide on that date. R02

Daily thresholds are calibrated in advance using prospective scenarios. RTE describes comparing the expected gain from selecting a day now with the gain from keeping it available for later. The dynamic method, based on the Bellman equation, formalises this value of waiting. For calibration of the day-ahead thresholds, the note refers to 200 climate scenarios. The procedure is repeated annually. R02

A restoring mechanism then adjusts the thresholds around the guide. If too few days have been selected, it makes further selections easier. If the calendar is running ahead, it raises the hurdle. The aim is to act progressively, before a large gap forces the remaining stock to be cleared. RTE publishes a formula with proportional and quadratic components; our collection did not retrieve their numerical coefficients or the complete operational threshold tables. R02

This structure offers greater predictability, with an explicit economic trade-off. The CRE reports that some suppliers fear the monthly bounds may reduce the value of demand response: a constraint can force the selection of a moderately useful day, or limit how much flexibility can be retained for a later period. The best calendar identified with hindsight and a calendar that can be operated before winter unfolds face different constraints. The reform strikes a balance between those objectives. G01

RTE’s evaluations show fewer clearing days

The CRE’s decision draws on RTE analyses: a retrospective comparison across four winters and a prospective assessment using 1,400 scenarios. The exercises address complementary questions. The first tests the methods against known winters; the second explores possible conditions. The results remain RTE’s, as reported by the regulator. G01

Across the four winters studied, average effectiveness is reported at 95% for the former Tempo method and 94% for PP. The CRE defines effectiveness as the ratio between residual demand on the twenty-two selected days and residual demand on the twenty-two highest-demand reference days. A measure close to 100% means the chosen days capture a large share of the quantity being targeted. It can remain high even when several selected dates differ from the ideal dates. G01

The number of clearance days provides another perspective. In the published comparison, the maximum is thirteen for Tempo, reached in winter 2025–2026, against three for PP across the sample studied. The improvement concerns days used to satisfy stock constraints. Three is an evaluation result within a particular sample; it does not establish a guaranteed ceiling for every future winter under the reform. G01

In the prospective simulations, 90% of scenarios have an effectiveness ratio above 93%; 90% have fewer than three clearance days. The two proportions are published separately. The text does not provide their joint distribution, so it cannot establish that the same 90% of scenarios meet both criteria simultaneously. The CRE itself stresses the statistical and imperfect nature of the exercise. G01

The indicator’s definition deserves particular attention because the May consultation also included a chart showing the proportion of red days selected from among the twenty-two highest-demand days. That measure of overlap between dates and the final decision’s residual-demand ratio answer different questions. Combining them into one series would create a misleading comparison. G03 G01

The public documents therefore support scrutiny of the method, its constraints and its reported results. Independently reproducing the evaluation would require the input series, forecast versions, parameters and detailed outputs. Our work reproduces the calculations in its own examples and treats RTE’s simulations as attributed results. That leaves a useful task for the next assessment: compare the announced method with the calendar actually produced and the information available at each decision.

The signal becomes six prices per kilowatt-hour

For a customer, a colour takes effect through the tariff schedule. Under the EDF offer used here, a Tempo day starts at 06:00 and ends at 06:00 the following morning. Peak-tariff hours continue until 22:00; the following eight hours are off-peak. The period from midnight to 06:00 therefore still belongs to the tariff day that began the previous morning. That detail can change the cost of an activity scheduled overnight. T07

The schedule effective from 1 August 2026, including all taxes, charges €0.1356/kWh for blue off-peak hours and €0.1654 for blue peak hours. White-day prices are €0.1536 and €0.1921, respectively. On red days, the rate is €0.1615 off-peak and €0.7295 during peak hours. These six rates apply to the energy consumed; a standing charge is added to the bill. T01

The most visible opportunity appears within a red day. Shifting 10 kWh from peak to off-peak hours on a red day, while keeping the volume unchanged, reduces the energy charge by €5.68 under this schedule. The calculation is ten times the difference between 0.7295 and 0.1615. It assumes the activity can actually be rescheduled, its energy requirement remains identical and the standing charge is unchanged. Vehicle charging or programmable equipment can be assessed with this formula; heating and occupancy constraints require analysis of the individual situation. T01

The price’s date matters as much as the colour’s date. A CRE explanatory page retains a table for the February 2026 tariff change, when the red peak rate was €0.7060/kWh. EDF’s August schedule shows €0.7295. This dossier uses the latter for current calculations. Comparing historical bills would require applying each successive schedule over the period when it was actually in force, instead of using a single price for the whole year. T04 T01

Calendar reform and price changes have separate causes and decisions. In its January 2026 proposal, the CRE explained that Tempo had moved away from the tariff level implied by the sum of its cost components. It proposed a 6.2% increase in the overall level, including taxes, to halve that gap, with further gradual adjustment to follow. This percentage described the option’s average change under the proposal; it did not mean an identical increase in every unit rate. T05

The price paid contains several components. EDF’s schedule states that it covers network access and use, with the applicable levies added. The CRE’s August annex provides the pre-tax rates; adding excise duty and VAT reconciles them with the six tax-inclusive prices. The ratio of the red peak rate to the blue off-peak rate is about 7.01 before taxes, then 5.38 including taxes: the same excise duty per kilowatt-hour narrows the visible ratio between the high and low prices. T01 T02

This composition explains why comparing a household’s red-day price directly with the wholesale price of a particular hour cannot establish a supplier’s margin. Procurement, network charges, commercial costs, taxes and remuneration for supply each have their own rules and timetables. The selection signal and the economic construction of the tariff need to be read together, with each function kept clear. T05 T01

The timing of consumption determines the bill

The bill is calculated by multiplying the kilowatt-hours in each of the six tariff categories by the relevant price, adding those amounts and then the standing charge. With rates held constant, the order of the colours would make no difference to a customer whose consumption was identical every day, with the same peak and off-peak split. Timing becomes decisive when volumes vary with weather, occupancy or activities.

A fictional example isolates the effect. Take two eligible days, with 8 kWh during peak hours on the first and 20 kWh on the second. Under one calendar, the first is red and the second blue: the energy considered costs €9.1440. Reversing the colours raises that cost to €15.9132. Total consumption remains 28 kWh, and exactly one of the two days remains red. The difference is €6.7692, or €6.77 after rounding. T01

Data and calculations

Volumes are fixed before colours are assigned. Other days and consumption remain identical.

Fictional example · euros including tax · before cent rounding

Colour assignment	Cost of the 28 peak kWh
A red / B blue	9.1440
A blue / B red	15.9132
Difference	6.7692

To put that pair back into a standard year, keep the other twenty-one red days and all other consumption unchanged. The comparison then isolates the effect of placing this one red day. It shows why a selection that tracks national peaks more closely can increase the exposure of a customer who consumes more at those times and has little scope to shift their use. It provides no estimate of the reform’s average effect on real customers.

The reasoning also works in the other direction. A customer who can readily adapt their activities may reduce consumption at high prices. Their bill will depend on that response as well as on the sequence of colours. The burden of organising a response also matters: rescheduling an appliance, charging a vehicle earlier and adjusting heating involve different efforts. The published electricity-system simulations assess day selection; they do not measure this full range of domestic constraints. G01

The economic value therefore has to be assessed against a consumption profile. To illustrate an annual calculation, we use a wholly artificial case: 8 peak kWh and 4 off-peak kWh every day, with 300 blue, 43 white and 22 red days. Holding August prices constant for twelve months and using a contracted capacity of 6 kVA gives a calculated total of €984.39 with Tempo, against €1,066.76 with Base at 6 kVA for the same volume. Standing charges are calculated as twelve times the published monthly amounts. This fixed-price example checks the calculation method; it describes neither a historical bill nor a representative household. T01

The accompanying calculator lets readers replace those assumptions with six consumption volumes and examine the effect of shifting red peak consumption into red off-peak hours. It keeps the prices, their date and the limits of the comparison visible. Its purpose is to show how a bill responds to changes in consumption volumes, without attempting to predict forthcoming colours. The data remain in the browser used to read the dossier.

Estimate a bill from your consumption

The six annual volumes and the selected shift remain in your browser. The starting profile is fictional; calculations hold VAT-inclusive prices at their 1 August 2026 level. T01

l0g / ELECTRICITY / LOCAL CALCULATOR

Measure the weight of red peak hours

Fictional starting point: 8 peak kWh and 4 off-peak kWh every day, for 365 days. EDF prices stay fixed at 1 August 2026, including taxes.

Initial example : €984.39 incl. tax / year · 6 kVA

Change annual volumes and shift red peak use

Annual consumption, in kWh

Blue peak€0.1654/kWh incl. taxBlue off-peak€0.1356/kWh incl. taxWhite peak€0.1921/kWh incl. taxWhite off-peak€0.1536/kWh incl. taxRed peak€0.7295/kWh incl. taxRed off-peak€0.1615/kWh incl. tax

Subscribed capacity6 kVA9 kVAShift red peak to red off-peak (kWh)

Restore fictional example

Each volume: 0 to 200,000 kWh, up to three decimal places. Use a decimal point or comma, without thousands separators. The shift must fit within entered red peak consumption.

Tempo before shift

€984.39

Tempo after shift

€984.39

Saving from shift

€0.00

Energy charge before shift

€794.79

Subscription, 12 months

€189.60

Total energy, conserved

4,380 kWh

Red peak charge before shift

€128.39

Red peak share of energy

4.0 %

Red peak share of energy charge

16.2 %

Red peak after shift

176 kWh

Red off-peak after shift

88 kWh

Base 6 kVA comparison

€1,066.76

Base less Tempo after shift

€82.37

Base comparison is available at 6 kVA. Base at 9 kVA is closed to new subscriptions.

Method and scope

Six annual volumes × their energy prices, plus twelve published rounded monthly subscriptions. The Base comparison uses the same total kWh, €0.2001/kWh and 12 × €15.86. A shift conserves energy and the subscription; equipment losses, comfort, practical feasibility, special surcharges and future price changes fall outside this calculation.

Energy-charge shares exclude the subscription. Zero denominators display “Unavailable”. Each monetary result is calculated exactly, then rounded to cents independently.

Calculations stay in this page. Values remain in memory while it is open.

EDF price schedule, effective 1 August 2026

Christmas Eve and New Year’s Eve expose the political choice

The exclusion of 24 and 31 December has a history. 31 December 2014, during the first winter of RTE’s new role, was red. In its July 2015 decision, the CRE considered the selection justified by net demand and consistent with the criteria. At the time, it opposed additional constraints that could weaken Tempo’s effectiveness. Consumer organisations, the energy ombudsman and EDF had challenged the choice. G06

31 December 2025 was selected again, once more with high residual demand according to the CRE. This differs from the March clearance episode: the debate concerns a day that was relevant for the electricity system but socially difficult to select. The 2026 decision places the acceptability of the signal at the centre of the argument, alongside the risk of discouraging customers from joining or remaining on an offer that requires demand reductions. G01

The shift in priorities is clear when the texts are compared. Preserving customer participation becomes a reason to restrict eligible dates. There is an economic case for that approach: a flexibility scheme depends on users’ ability to stay engaged. The documents examined, however, do not quantify a causal effect of excluding the two holiday eves on departures, subscriptions or reductions in consumption. Acceptability is an explicit objective of the decision, whose effectiveness remains to be observed. G06 G01

The consultation also records the counterargument. Some participants prefer the shared signal to target electricity-system conditions as objectively as possible, leaving each supplier to manage its customer relationships. The CRE chooses to change the common calendar. Any system stress during the holidays can be signalled through Ecowatt, with a call to reduce consumption. The instruments then ask users to respond in different ways. G01

The legal scope requires one further distinction. The final provisions exclude 24 and 31 December from red Tempo days from November 2026. In its reasoning, the CRE says those dates should also be excluded from PP days to preserve consistency. This investigation did not establish the corresponding amendment to the PP rules. We therefore retain the Tempo exclusion as the verified decision, without automatically extending its scope to every use of the capacity signal. G01

Earlier notice takes effect in November 2027

A well-chosen day can still be difficult to respond to if the information arrives too late. RTE’s official page currently specifies 10:30 on the previous day as the binding reference for the Tempo colour. When sufficient information is available, a preliminary signal may appear between 08:00 and 10:30; its status is advisory. EDF then says it relays the colour from 11:00. These times describe a chain of publication and transmission. R03 T07

The PP calendar already allows earlier notice: under normal conditions, a day may be signalled two days in advance, before noon, or by 09:30 on the previous day at the latest. The method includes an additional margin for decisions made two days ahead, because forecasts are less certain at that horizon. Once a PP day has been published, RTE retains the decision. If information is missing at an early stage, the decision should still be understood as open. R02 R04

The reform provides for aligning PP and Tempo publication times. From 1 November 2027, RTE must publish the red Tempo day within one hour of the PP signal. An early PP selection will therefore be able to give subscribers more time to organise their activities. The possibility of two days’ notice will still depend on when the PP day is actually decided. G01

For the winter beginning in November 2026, the official Tempo reference therefore remains publication at 10:30 on the previous day. The CRE explains the additional year by the need to adapt processes and information systems. Promising systematic official Tempo notice two days ahead this winter would turn a possibility into a guarantee and move a 2027 implementation date into 2026. G01

A safety exception also prevents PP and red Tempo days from being treated as absolute synonyms. If Ecowatt changes from green to orange or red after 10:30 on the previous day, a PP day can be imposed without a red Tempo day being selected. The decision states this explicitly in a footnote. The case illustrates the difference between a tariff signal announced within a contractual framework and an operational response to changing conditions on the electricity system. G01

Useful transparency therefore includes both the timing and the authority of each piece of information. A forecast, a preliminary indication, a PP day and the official Tempo colour each have a defined status. The tool accompanying this investigation is devoted to prices and consumption volumes entered by the reader. To check an actual colour, readers can follow the link to RTE’s reference publication in the dossier’s sources. R03

The data needed to evaluate the reform

The reform introduces features that can be checked: a documented hourly criterion, a seasonal path with bounds, thresholds announced as public and a specified notice timetable. It also retains an inherent tension in the product: twenty-two days must be selected before the whole winter is known. Clearance days can persist within the monthly constraints, although the evaluations presented show a substantial reduction in their frequency. R02 G01

A useful review will match each selection with the forecast available at the time, its threshold, the remaining stock and the calendar constraints. It will then distinguish realised consumption from what was expected. Assessing economic effectiveness for customers will also require observing volumes shifted or avoided, the burden of organising a response and changes in bills across comparable consumption profiles. Those dimensions complement the electrical ranking of days.

Governance of the signal becomes a matter of explicit choices. The level of demand to target, how efforts are distributed, the exclusion of certain dates and the time available to adapt all influence the scheme’s value. The 2015 and 2026 decisions show that these priorities can change. Making them legible lets readers understand why a day turns red and the basis on which that decision can be examined. G06 G01

The coming winter will provide the first observations of the new calendar. The test will be whether the selected days remain well targeted, whether the sequences become easier to anticipate and how customers actually adapt their use. The price of a red day fits in one cell of a tariff table. Its usefulness depends on the full relationship between electricity-system needs, the selection rule and customers’ ability to respond.

Further reading

• Blackouts: financing service continuity
• The cost of cooling and adaptation
• Electricity’s water constraint in Europe

Sources and references

Documents checked on 29 October 2026. Document dates and effective dates are kept separate.

1. G01 · Commission de régulation de l'énergie / Journal officiel via Légifrance · Document: 2026-10-21 · Published: 2026-10-24Decision No. 2026-189 of 21 October 2026 on electricity network operators’ duties concerning Tempo-type tariffs

   The decision defines the reform and its timetable. It reports the thirteen consecutive red working days in March 2026, consultation trade-offs and RTE’s evaluation results.

   Passages and scope

   Experience: Section 1, p. 3; method and evaluations: Section 2, pp. 4–6; notice: Section 2, p. 7; holidays: Section 3, pp. 7–8; operative decision: p. 9; footnote 8: p. 7.

   The simulations are RTE’s, as reported by the CRE; this investigation does not independently reproduce them. The text does not enumerate March dates. Notice alignment takes effect only in November 2027.

2. G03 · Commission de régulation de l'énergie · Document: 2026-05-22 · Published: 2026-06-01Public consultations Nos. 2026-10 and 2026-11 of 22 May 2026 on the construction of regulated electricity tariffs

   The document tracks the questions put to consultation and the reasoning preceding the October decision. It distinguishes two consultation streams and response deadlines.

   Passages and scope

   Opening: two consultation streams; Tempo discussion around printed pp. 22–24 and Figure 15; landing pages: upload date and deadlines.

   This is a proposal. Figure 15 measures overlap between selected dates, distinct from G01’s residual-demand ratio. The document is dated 22 May; landing pages show publication on 1 June. G01 describes the consultation as running from 22 May to 24 July.

3. G04 · RTE · Document: 2025-01-07Tempo day-selection method : revision 2

   The method documents forecast demand, wind and solar generation, remaining day stocks and progress through the season. Stock constraints can require days to be selected at the end of the period.

   Passages and scope

   p. 1: calendar; pp. 2–3: net demand and normalisation; pp. 3–4: thresholds and stocks; pp. 5–6: illustrative example.

   The simplified formula using 46,050 and 2,160 is illustrative and based on 2014–2015. Published coefficients alone do not reproduce every operational selection.

4. G05 · Commission de régulation de l'énergie / Journal officiel via Légifrance · Document: 2014-10-30 · Published: 2014-11-01Decision of 30 October 2014 on network operators’ duties concerning Tempo-type tariffs

   The transfer of colour selection to RTE from 1 November 2014 explains the separation between a shared calendar and commercial offers. The text sets out the aim of opening access to multiple suppliers.

   Passages and scope

   §2.1: criterion and shared signal; §2.2: publication and notification; operative section: transfer to RTE.

   The 2014 tariffs, capacity bands and technical details are historical. Proposals discussed must be distinguished from adopted provisions.

5. G06 · Commission de régulation de l'énergie / Journal officiel via Légifrance · Document: 2015-07-16 · Published: 2015-07-30Decision of 16 July 2015 on network operators’ duties concerning Tempo-type tariffs

   The 2015 precedent shows the earlier treatment of 31 December, algorithm-transparency requirements and efficiency trade-offs. It places the 2026 decision in a longer regulatory history.

   Passages and scope

   §2.1: 31 December; §2.5: transparency; §2.6: degraded operation; §2.7: structural changes; §2.9: capacity-mechanism convergence.

   Mechanisms and timings described are those of 2015. Comparing them with 2026’s expressly exceptional implementation timetable is not a legal finding of irregularity.

6. R02 · RTE · Document: 2026-09-30Selection method for peak-period days (PP days)

   The criterion uses the forecast hourly peak of residual demand. Monthly cumulative bounds and a restoring mechanism shape the allocation of 22 days. The guide curve is a distribution target, distinct from mandatory bounds.

   Passages and scope

   p. 3: days, hours and notice; p. 4: criterion, thresholds, Bellman optimisation and scenarios; p. 5: cumulative bands, guide curve and advance notice.

   The note predates the 21 October decision excluding 24 and 31 December. Operational MW threshold tables and numerical restoring coefficients were not retrieved in this investigation.

7. R03 · RTECalendar for Tempo-type electricity supply offers

   The next-day colour published at 10:30 is the binding reference. The preliminary morning signal is indicative and can change. Suppliers set the prices associated with each colour.

   Passages and scope

   Introductory text: official 10:30 publication, 08:00–10:30 preliminary signal, supplier roles; 2019–2020 exception.

   The dynamic calendar does not expose its daily series in extracted text. Search-cache counters can be stale.

8. R04 · RTEPP signals

   A PP signal can be published by noon on D−2 and, under the normal procedure, by 09:30 on D−1. An exceptional route uses orange or red EcoWatt warnings. Once published, a PP signal is irreversible.

   Passages and scope

   Signal statuses and normal and exceptional publication windows.

   These times apply to PP signals. The CRE schedules full alignment of Tempo notice for 1 November 2027.

9. R05 · RTEParticipating in the new capacity mechanism : delivery periods 2026–2036

   The mechanism remunerates available capacity and allocates its funding using consumption during peak periods. PP signals therefore connect capacity availability and participants’ contributions.

   Passages and scope

   Mechanism overview and supporting documents, including the 30 September PP method.

   The page collects documents with separate version dates. It has no single established publication date.

10. R06 · RTE / EcoWattHow do red Tempo days differ from a red EcoWatt warning?

    Tempo provides a tariff incentive for subscribers. A red EcoWatt warning signals a situation in which demand reduction helps avoid or limit power cuts.

    Passages and scope

    FAQ answer.

    The FAQ is undated. Its historical claim about red warnings is not used. The reform creates operational links between PP, Tempo and EcoWatt.

11. R07 · RTE / EcoWattHow are EcoWatt signals calculated?

    EcoWatt considers system margins: forecast demand, available production and possible imports. This differs from the residual-demand measure used to rank days.

    Passages and scope

    Descriptions of system margins and the three colours.

    The FAQ offers a general explanation rather than the parameters of a complete adequacy simulation.

12. T01 · EDFTarif Bleu price schedule : effective 1 August 2026

    The six VAT-inclusive Tempo energy rates and monthly subscriptions share a schedule with Base and peak/off-peak comparators.

    Passages and scope

    PDF p.1: Base, Heures Creuses and Tempo, 6/9 kVA rows; p.2: price components and taxes.

    The displayed date is an effective date; a separate upload date was not established. Subscriptions are rounded monthly amounts. Multiplication by twelve is a fixed-rate calculation, not a forecast. Special surcharges are outside the scenarios.

13. T02 · CRE · Document: 2026-07-15 · Published: 2026-07-1615 July 2026 deliberation, Annex B : pre-tax schedules

    The pre-tax schedule cross-checks EDF energy rates after excise and VAT, with rounding.

    Passages and scope

    Annex B1 p.2: pre-tax/network scope; p.3: Tempo periods; p.6: Base/peak-off-peak; p.7: standard Tempo, outside collective self-consumption.

    The prices include standard network services and do not represent spot generation costs. Self-consumption versions have a different scope.

14. T03 · Journal officiel / Légifrance · Document: 2026-07-29 · Published: 2026-07-3129 July 2026 decision on mainland French regulated residential electricity tariffs

    The decision adopts CRE schedules and makes them effective on 1 August. The annex specifies tariff periods and capacity levels.

    Passages and scope

    Articles 1 and 2; Annex B1 definitions and residential Tempo option.

    This text sets pre-tax rates and the option framework; it does not itself describe RTE’s daily selection algorithm.

15. T04 · CREUnderstanding regulated tariffs : VAT-inclusive examples for the 1 February 2026 change

    This page retains February’s six VAT-inclusive rates and annual subscriptions, identifying the previous schedule.

    Passages and scope

    Table headed “Exemples de grilles … mouvement du 1er février 2026”; Tempo 6/9 kVA rows.

    The introduction also uses “in force”. The tables explicitly refer to February; use August’s T01/T02 for current rates.

16. T05 · CRE · Document: 2026-01-14Deliberation 2026-06 of 14 January 2026 : aligning Tempo with costs

    CRE explains continued convergence of Tempo’s overall level toward stacked costs while preserving attractiveness for flexible consumers.

    Passages and scope

    P.17: structure and minimum ratio; p.18: “Cas de l’option Tempo”, proposed 6.2% VAT-inclusive level increase and announced 2027 convergence.

    The 6.2% is an overall level change, not the rate for every cell. The 2027 continuation is a January announcement, not a future price established here.

17. T06 · CRE · Document: 2026-07-16 · Published: 2026-07-16Proposed average 2.5% VAT-inclusive regulated-tariff rise on 1 August 2026

    The release describes August’s average change, the excise reduction from EUR30.85 to EUR30.62/MWh and a separate 24 kVA Tempo subscription.

    Passages and scope

    “Evolution des tarifs” and “Evolution de la structure”; excise, network tariff and capacity.

    This explains a proposal published on 16 July and should be read with adopted decision T03. An average change does not predict every bill’s variation.

18. T07 · EDFTempo option: operation and conditions

    The Tempo day runs from 6am to 6am, with off-peak hours from 10pm to 6am. EDF states a 6 kVA minimum and a communicating Linky meter for new subscriptions.

    Passages and scope

    FAQs on the Tempo day, meter compatibility and red days; relay from 11am.

    This describes EDF’s offer and information relay, not every offer using the calendar. Promotional percentages have specific comparators and rounding; prefer T01 amounts.

19. T09 · Médiateur national de l’énergie / énergie-infoHow Tempo-type electricity offers work

    The factsheet distinguishes RTE’s shared colour from supplier-specific prices and conditions, referring users to the public comparison service.

    Passages and scope

    Overview; RTE publication; comparing offers.

    The factsheet does not establish universal savings across household consumption patterns.

20. W01 · Météo-France · Document: 2026-03-04 · Published: 2026-03-04Winter 2025–2026: Météo-France climate review

    The climate review independently confirms the mild December-to-February period, ranked the fourth-warmest winter since 1900 by Météo-France. It establishes weather context without reconstructing daily electricity-system conditions.

    Passages and scope

    Opening paragraph and the section on the particularly wet and mild winter; the opening defines the meteorological season.

    December–February differs from the November–March electricity selection period. Average national temperature alone cannot explain individual Tempo selections; demand and renewable-generation forecasts also matter.

21. T10 · French Energy Code · Version effective since 3 May 2025Article L271-1: electricity demand reduction

    The Code defines a temporary reduction relative to a forecast schedule or estimated consumption. The action may be accompanied by higher consumption before or after the reduction period.

    Passages and scope

    First paragraph: temporary reduction and baseline; second paragraph: consumption shifted in time. Measuring demand response and operating a supply-tariff calendar serve distinct purposes.