A home does not consume the sum of its nameplates: it consumes its worst quarter-hour

Add up the nameplates of every appliance in the house and you get a frightening number — twelve, fifteen, twenty kilowatts — that never happens: the oven does not bake while the washing machine spins, the boiler heats and the air conditioning pushes all in the same instant. But the supply contract is not sized on the sum nor on the average: it is sized on the worst quarter-hour — that December Saturday when the loads line up. This calculator estimates that peak with two competing forces: coincidence, which squeezes the sum, and the largest single load, below which no statistic can go.

The formula: the maximum between the statistic and the single

The model is a duel: P_peak = max(S·f/100, M). On one side the sum of loads S reduced by the coincidence factor f — in the ordinary home 30-40% of installed loads work together at the worst moment, in the electrified home that cooks, heats and charges it climbs to 50-70%. On the other the largest single load M: the 7 kW induction hob switched on is 7 kW, and no statistical factor discounts it. The realistic peak is the greater of the two — and it is the number to compare with the available power: the coverage δ = P_c/P_peak says whether the contract holds.

The scale of the sum: the small home is dictated by the single appliance

The table grows the installation with everything else held equal (coincidence 40%, contract 6 kW, largest load 4 kW):

The first two rows tell the hidden regime: at 6 and at 9 kW of sum the peak is identical — 4 kW — because the single load dictates it, not the statistic. Only from the third row does coincidence take command, and from there the peak grows with the installation until it breaks through the contract: at 18 kW of sum, 1.2 kW are already missing from the roll call.

A complete numerical example

Typical home: 12 kW of summed nameplates — oven, dishwasher, washing machine, 4 kW boiler, AC, the rest — coincidence at 40%, a 6 kW contract. The statistic says 12 × 0.4 = 4.8 kW; the largest load says 4 kW: the statistic wins, and the realistic peak is 4.8 kW — about 20.9 amps at 230 volts. The coverage is 6 / 4.8 = 1.25: the contract covers the peak with 1.2 kW of reserve — the right room for surprises and for the first new load. It is the count to redo with every major appliance that enters the house: the house decides the contract, not the other way round.

How to read the coverage

The four bands tell four contracts. Below 0.85 the contract is below the peak: trips are not bad luck, they are arithmetic. Between 0.85 and 1.05 you live on the break-even line: the normal day holds, the full one — oven, washer and AC together — is the test that comes. Between 1.05 and 1.35 there is the right margin: peak covered, reserve for the future. Above 1.35 the power is overabundant — and here the direction surprises: where contracted power is paid on the bill, the margin sleeps but the fee does not. Excess is not neutral: it is a recurring cost that must be justified by a plan.

Coincidence: the number you can govern

The variable f is the only one in the count you can act on for free. It is not a physical constant: it is a portrait of habits — who cooks while washing, who charges at night, who gives the boiler its own hour. The ordinary home lives at 30-40%; electrification pushes it up, because heat pump and wallbox run long and overlap everything. But the arrow works backwards too: load choreography — who starts when — lowers f without switching anything off. Two big appliances never together are worth half a kilowatt of contract: it is the efficiency you do not buy, you organise.

The single load: the floor below which the statistic cannot go

The variable M is the correction hasty counts forget: the coincidence factor is an average, but the induction hob switched on is not an average — it is all there, in that instant. That is why the model takes the maximum: when M exceeds S·f/100, the peak is dictated by the single appliance, and switching off the rest of the house does not lower it by a watt. It is the typical regime of the small home with one big load — the studio flat with induction, the gas-heated house with a new wallbox — and it changes the strategy: choreography is useless, you act on that load or give it a contract that respects it.

International perspective: kilowatts, amps and the same quarter-hour

The domestic peak is universal, the units are not. The Europe of kW contracts reasons like this calculator — power subscribed in steps, disconnection above the threshold; North America reasons in panel amps — the classic 100 or 200 A services at 240 volts — and the codes use demand factors to reach the same number. The fourth output is the bridge: the peak in kW becomes an indicative current, and the logic does not change with the continent — sum, coincidence, single load, margin. Only the line of the bill where the number lands changes.

Limits of the model and correct use

The model estimates the single-phase active power peak: it does not cover motor starting surges (a few seconds meters tolerate), three-phase installations with unbalanced loads, nor power factor correction — the designer's trades. The factor f is a lifestyle estimate: the smart meter that measures the real quarter-hour is worth more than any desk refinement. And the peak is a photograph that ages: the EV charger installation calculator accounts for the heaviest new load that can enter the garage, and the solar panels needed calculator sizes the production that lightens the draw in the right hours — because the electrifying home redoes this count at every step, not once and for all.

The scale of coincidence: from scattered habits to the all-at-once home

The table changes only the coincidence factor, installation and contract held still (sum 12 kW, contract 6 kW, largest load 4 kW):

The first two rows show the floor in action: below a certain threshold coincidence stops counting, because the 4 kW load dictates the peak — lowering f from 30 to 25 does not change a watt. From the third row on, the statistic takes back command, and in two steps the 6 kW contract goes from comfortable to insufficient: the same installation, four different contracts, just for how the house is lived in.

Reading the floor: when choreography pays and when it does not

The operational reading is worth a contract. In the table's upper zone — f at 55-70% — load choreography pays: every avoided overlap lowers the true peak, and the washing machine's timer is worth money. In the lower zone — f at 25-30% — the choreography is already done: the peak sits on the single load's floor, and the only lever left is that load. It is the criterion for choosing the battle: first look at which of the two terms wins the max, then decide where to invest — in habits if the sum wins, in the appliance if it does.

The scale of the contract: the peak does not move, the judgement does

The table changes only the available power, peak held still (sum 12 kW, coincidence 40%, largest load 4 kW):

The peak column does not move — 4.8 kW on every row, because the peak belongs to the house, not the contract — yet the five rows describe five different lives: from the arithmetic trips of the first to the sleeping fee of the last. It is the calculator's most practical message: the contract does not change how much you consume — it only changes how much it costs you and when the lights go out.

The scale's two tails: deficit and excess both cost

The extreme rows deserve a double reading. At the bottom, the deficit: at 0.73 coverage the house lives on trips, and the solution is not necessarily the contract — half a point of coincidence gained through choreography is worth a step of contract, and costs zero. At the top, the excess: at 2.08 coverage the margin is bigger than the peak itself, and where contracted power is paid, those sleeping kilowatts are a perpetual fee for a service never used. The right coverage — 1.05-1.35 — is not timid prudence: it is the point where every paid kilowatt works at least a few quarter-hours a year.

From the count to the contract: the sequence that protects the bill

The right order is a defence of the fee. First, the inventory: nameplates are read, not guessed — and the largest load is identified by name. Second, the true quarter-hour: a smart plug or the meter's display for a week is worth more than any desk factor. Third, the max: statistic against single load — the winner decides the strategy. Last, the contract: you choose the step that gives 1.05-1.35 coverage on the measured peak, and note to redo the count at the next major appliance — because the peak is a photograph, and houses change.

Typical mistakes and how to avoid them

Four recurring traps. Sizing on the sum of nameplates: no house switches everything on together — you size on the worst quarter-hour, not the inventory. Ignoring the single load: the average factor does not discount the induction hob switched on — below its floor the statistic cannot go. Raising the contract at the first trip: first look at what was on — half a point of choreography often solves it for free. Forgetting the fee: where power is paid, the generous contract «for peace of mind» is peace of mind on subscription — coverage above 1.35 must be justified by a plan, not by a fear.

Frequently asked questions

How much power does a home need?

The realistic peak is max(S·f/100, M): with 12 kW of summed nameplates, coincidence at 40% and a 4 kW largest load, the peak is 4.8 kW — a 6 kW contract covers it with 1.2 kW of reserve (coverage 1.25). The count is redone with every major appliance that enters.

What is the coincidence factor?

The share of installed loads that really works together in the worst quarter-hour: 30-40% in the ordinary home, 50-70% in the electrified one that cooks, heats and charges. It is not a constant: it is a portrait of habits — and load choreography lowers it for free.

Why does the calculator ask for the largest single load?

Because coincidence is an average, but the induction hob switched on is not an average: it is all there. The realistic peak is the maximum between sum×factor and largest load — when the latter wins, switching off the rest of the house does not lower the peak by a watt.

What does the coverage δ mean?

The ratio between available power and realistic peak. Below 0.85 trips are arithmetic; between 0.85 and 1.05 you live at break-even; between 1.05 and 1.35 there is the right margin; above 1.35 the power is overabundant — and where power is paid on the bill, the excess is a sleeping fee.

How do I calculate the sum of installed loads?

By adding the nameplates: oven 2-3 kW, induction 3-7.4, washer and dishwasher 2 each, boiler 1.2-4, AC 1-3 per unit, wallbox 3.7-11. Nameplates count, not usage: coincidence and the largest load turn the sum into a peak.

Does the meter trip as soon as I exceed the contracted power?

Almost never instantly: limiters tolerate brief, moderate overruns — it is the prolonged overrun that trips. That is why the model reasons on the quarter-hour, not the one-second spike: motor starting surges are not the problem, the oven plus the AC for half an hour is.

Should I lower the contract if the coverage is high?

Where contracted power is paid, yes — if no new load is in sight: it is a recurring zero-risk saving. If the house is electrifying (induction, heat pump, wallbox), the margin is instead foresight: the difference is made by the plan, not the number.

Does a wallbox force me to raise the contract?

Not always: many wallboxes modulate their power and charge at night, when the rest of the house sleeps — near-zero coincidence. The right question is whether charging coexists with the evening peaks: if so, the contract is redone; if it charges at dawn, what you have is often enough.

Does the calculator work for 100 or 200 amp panels?

Yes: the logic is the same, the unit changes. The fourth output converts the peak into indicative amps at 230 V; for North American 240 V services the ratio is similar. Sum, coincidence and single load are universal — local codes refine the factors, not the structure.

What is the difference between active power and the meter's kVA?

Active power (kW) is what loads consume; apparent power (kVA) includes the reactive part. At home the difference is small — domestic loads have a good power factor — which is why the model reasons in kW: in workshops full of motors the story changes, and the designer is needed.

Does solar power lower the contract peak?

It helps when it produces: at noon the draw falls even with loads on. But the home's worst quarter-hour is often in the evening or winter, when panels sleep: the contract is sized on the sunless draw peak, and solar is enjoyed as a discount, not as a guarantee.

What does this calculation NOT consider?

Starting surges (seconds, tolerated by limiters), three-phase installations with unbalanced phases, power factor correction and the detailed rules of local codes. The model estimates the realistic single-phase peak: the smart meter measuring the true quarter-hour is the verification worth more than any estimate.

Updated July 2026