Electricity

Parameters

Residential kwh per unit day

30

Electricity one household unit draws per day. EIA RECS 2023 puts the average US home at ~10,800 kWh/year ≈ 29.6 kWh/day, rounded to 30. A 12-unit apartment therefore draws 360 kWh/day.

Source: US EIA — Residential Energy Consumption Survey (RECS) 2023

Worker kwh per seat day

20

Electricity one workplace seat draws per day, whether or not anyone is sitting in it — the building is lit, heated and cooled regardless. EIA CBECS puts commercial office intensity around 9-12 kWh/worker/day; rounded up to 20 to cover the equipment-heavy uses (kitchens, refrigeration, workshops) that share this term.

Source: US EIA — Commercial Buildings Energy Consumption Survey (CBECS)

Patient bed kwh day

200

Electricity one hospital bed draws per day. CBECS puts inpatient healthcare at roughly 3× ordinary commercial energy intensity and it runs 24/7 (imaging, sterilisation, ventilation, redundant plant), so a bed is priced an order of magnitude above an office seat.

Source: US EIA — CBECS, inpatient health care energy intensity

Student kwh day

5

Electricity one school place draws per day. CBECS education intensity is the lowest of the occupied categories — classrooms are lit and conditioned on a school day only — game-rounded to 5.

Source: US EIA — CBECS, education energy intensity

Inmate kwh day

15

Electricity one prison place draws per day. Between a school place and an office seat: a cell is small but occupied around the clock, with perimeter lighting and security systems running continuously.

Source: US EIA — CBECS, public order and safety energy intensity

Floor kwh per tile day

5

Floor under derived demand, per footprint tile (10m × 10m) per day. Catches blueprints with no units, beds, students or inmates at all — monuments, squares, civic buildings that declare no staff — so nothing in the city is literally free to run.

⚠️ Source pending

Wind output multiplier by regime

Static per-map multiplier on wind farm output, keyed off the map’s wind regime. A windy map (Velaria) is genuinely the wind-power map — siting a city there is a real energy decision — while output stays deterministic day to day: the rated figure already bakes in a ~35% capacity factor, and letting the weather simulation move it would split the grid into two contradicting on-screen numbers and turn the clean-grid achievement into weather RNG.

⚠️ Source pending

Carbon g per kwh

Carbon intensity of each source of the city’s electricity. The four it generates itself are IPCC AR5 WGIII Annex III lifecycle medians (coal 820, utility solar PV 48→45, nuclear 12, onshore wind 11). The import figure is a FALLBACK: what a tower actually carries is its neighbour’s, and the four differ (city/utils/neighboring_cities.ts). 900 is the worst of them, Richton’s lignite. Buying electricity is not met by a neighbour’s average mix but by their MARGINAL unit — whatever they spin up to cover the extra demand, firmed for the hours their own weather does not cover — and then ~8% of it is lost in the wires on the way over. That firming is why importing ‘solar’ is not clean power: you draw round the clock and their panels do not. This is consumption-based accounting (the Our World in Data framing the Environment charts use): buying power cannot move its emissions off the city’s chart, while domestic sources can displace imports through normal dispatch. The bare value is the import figure, which is also what the regional cooperative feed every city starts with emits.

Source: IPCC AR5 WGIII Annex III (lifecycle medians); Our World in Data — carbon intensity of electricity

Coal fuel cost at full dispatch

$750,000

Monthly coal fuel bill when the plant runs flat out, on top of its fixed operating cost, at a regional price index of 1.00. Fuel dominates real coal plant operating expense, and unlike every other source this line scales with how much the plant is actually dispatched — merit order idles coal behind cleaner sources, so building renewables cuts the bill the same month, without demolishing anything. The index that multiplies it is the Commodities Desk’s (see simulation/fuel_market.ts).

⚠️ Source pending

Seasonal demand

Multiplier on every class’s electricity demand, by calendar month. Demand breathes: heating peaks in January and cooling in August, so a city sized to its average browns out twice a year and a city sized to its January never does. Deterministic, because the point is that the squeeze can be planned for — the paper prints the January forecast every September. EIA hourly load puts US summer peaks 10-20% above the annual average and RTE winter peaks around +15%; ±8-10% is the tame end of that, a planning nudge rather than a treadmill. The bare value is January, the peak the placement cursor and the founding gate always quote.

Source: US EIA — Hourly Electric Grid Monitor; RTE — French winter peak load

Import rate brackets

What imported electricity costs per kWh, by how deeply the city leans on it. The first 15% of the city’s dispatch bills at wholesale; the slice from 15% to 30% carries a scarcity premium; everything past 30% is priced as the neighbour’s worst marginal unit. Billed per slice rather than at one blended rate, so crossing a bracket never produces a cliff worth micro-managing. EIA wholesale power trades at $30-60/MWh; the upper brackets are the multiples scarcity events actually reach.

Source: US EIA — Wholesale Electricity Market Data

Import lifeline kwh daily

30000

Imported volume billed at bracket 1 before the share slicing starts. Share sets the RATE and city size sets the VOLUME, so without this a forty-home hamlet on one tower — necessarily 100% import share — would pay the scarcity rate on every kWh it has. The lifeline protects small cities by construction rather than by a special case, and is excluded from the Ledger’s FUEL BINDS test so the verdict names the month’s real story.

⚠️ Source pending

Import heatwave multiplier

2

How far a regional heatwave moves every import bracket, for 60-90 days, summers only. Booked through the Regional Events Calendar like every other shock, so it is announced in the business pages before it lands — and a city that got off the towers reads that story with satisfaction instead of dread. ERCOT and CAISO scarcity events routinely double or worse.

⚠️ Source pending

Rate lock

San Gigante Power & Light’s rate lock: every bracket rate frozen at 1.15x today’s for five years, immune to heatwaves. The lock is on the TABLE, not on the blended effective rate — locking the blend would let a player park their import share low for one month, sign, and then lean hard for five years at a discount. Freezing the table keeps the convexity, which is the mechanic. Early exit costs three months’ billing, like every ContractLock.

⚠️ Source pending

Coal index bounds

The regional coal price index that multiplies the fuel bill, and the band it may move in. Newcastle thermal coal moved roughly 4x across 2021-22; the game caps a civilized 0.6-2.2 so a fuel market is dramatic without ever being terminal. Its long-run mean must sit at 1.00, because that is what makes the published $0.0125/kWh honest for a player who never opens the Commodities Desk.

Source: Newcastle thermal coal futures, 2021-2022

Coal index drift max

0.02

How far the coal index may drift on its own in a year, absent any announced event. Mean-reverting breathing so quiet periods aren’t frozen — never a squeeze. Every LARGE move is event-driven and published two quarters ahead; a caused-by-nothing random walk is the exact shape the wind-AR(1) rejection condemned.

⚠️ Source pending

Coal reserve max months

3

How much coal the city may stockpile, in months of full-fleet burn. Power Grid lets a player hold twice their plants’ appetite; three months is that idea resized for a monthly billing cadence. Insurance and speculation — buy at 0.6, coast through 2.2 — never denial and never an outage: the reserve running dry costs money at the index, not darkness.

⚠️ Source pending

Coal supply agreement

The Commodities Desk’s supply agreement: the index locked at 1.10x today’s for three years, exit at three months’ fuel. Deliberately a real either/or against the reserve — lock the price or hoard the tons; holding both is paying twice for the same insurance.

⚠️ Source pending

Catalog arrival window years

How wide the announced arrival window of a new plant generation is. The catalog prints the window (‘Expected 2031-2033’); the day inside it is rolled once per city and never shown. Power Grid’s Step 3 card translated: a clock you can see coming and cannot precisely time.

⚠️ Source pending

Catalog discontinue after years

1

How long the previous generation stays on sale after its successor arrives. This is the obsolescence march — built plants run forever, but the cheap old model eventually stops being for sale. One year keeps at most two generations of a source buyable at once, and guarantees a source is never unbuyable (§2.4).

⚠️ Source pending

Carbon levy schedule

The Clean Air Act’s carbon levy: $10 per tonne at enactment, rising $5 each year to a $60 ceiling, printed in full to the horizon the day the Act is drafted. This is what completes the clean transition’s cash story — a coal city reads its own future in a table and starts planning the exit years before the first bill. Canada’s federal carbon price runs $10 to $170/t on a published schedule; EU ETS trades near €80.

Source: Government of Canada — federal carbon pollution pricing schedule

Act trigger cumulative co2 tons

2000000

Cumulative dispatched carbon that summons the Regional Assembly. The reckoning is self-made, exactly like Power Grid’s player-made scarcity: run dirty and you meet the Assembly sooner. One coal plant at full dispatch accrues 49.2 kt/month and gets there in about 40 months; a two-plant fleet in 20; four towers leaning on Richton’s lignite in 31.

⚠️ Source pending

Act enactment roll months

How far ahead the Act’s enactment day is rolled once it is drafted. The countdown WINDOW is printed on the front page; the exact day is not, and never reaches the client. Power Grid’s Step 3 card at unknown deck depth.

⚠️ Source pending

Era2 population backstop

15000

Population that summons the Assembly regardless of carbon, so a city that ran clean from day one still eventually gets its Clean Air Act — as a formality it can read with pride rather than a bill it has to pay.

⚠️ Source pending

Era3 delay years

How long after the first Act’s enactment the Second Act bans new coal — pulled forward to five years if the grid is still dirtier than ERA3_INTENSITY_GATE at that mark. Two catalog cycles either way, with the levy sitting mid-schedule by then. Existing plants are never confiscated; they bleed.

⚠️ Source pending

Era3 intensity gate

400

Grid carbon intensity above which the Second Act arrives early. 400 gCO₂/kWh is roughly the current world average — a city still above it eight years into a carbon levy is not transitioning, and the Assembly notices.

⚠️ Source pending

Fuel binds cost share

0.7

Share of total grid cost that must be VARIABLE (coal fuel plus metered imports above the lifeline slice) before the Grid Ledger’s verdict reads FUEL BINDS rather than CAPACITY or DEMAND. The lifeline is excluded so the verdict names the month’s real story instead of the mere fact that imports now cost money.

⚠️ Source pending

Reserve margin bands

Headroom over PEAK demand, banded for the Grid Ledger: at or above 15% comfortable, 0-15% tight, below zero a brownout season. NERC plans real grids to roughly a 15% reserve margin; the bands are that reference, arcade-rounded.

Source: NERC — Long-Term Reliability Assessment (planning reserve margins)

Cup clean reserve margin

0.05

Reserve margin a peak month (January, August) must also clear for the month to count toward the Reliability Cup. Zero dark homes alone is not enough in the months that test a grid — the Cup is for cities that had headroom when it mattered, not cities that squeaked through.

⚠️ Source pending

Cup population gate

10000

Population needed for the Cup to also count as the PERFECT_ATTENDANCE achievement. Power Grid’s winner powers the MOST cities, not merely all of their own: a forty-home hamlet cannot farm a perfect record.

⚠️ Source pending

Poach streak months

3

Consecutive monthly passes with industry short of power before San Gigante’s Development Agency makes an offer to one of your factories. Three months ties the offer to a sustained shortage rather than to one difficult month.

⚠️ Source pending

Poach clock days

90

How long a rival offer remains open before the company leaves. Changing the grid conditions, paying a retention grant, or letting the offer expire have different consequences. At most one active poach per quarter — pressure, never a death spiral.

⚠️ Source pending

Retention grant months

The retention grant is the larger of six months of the company’s corporate tax and three months of its payroll. The payroll floor is load-bearing: corporate tax defaults to 0%, which would otherwise make every grant free. For a largest-payroll target the grant deliberately exceeds the $3M transmission tower, putting immediate certainty against infrastructure that also changes future grid conditions. US states really do price retention per job in interstate incentive wars.

⚠️ Source pending

Export tiers

What the neighbours pay for surplus power, in concave 500,000 kWh/day tiers — Power Grid’s payment table, which pays well for the first cities and poorly for the last. EIA wholesale sits near $36/MWh: the top tier is a scarcity premium, the tail is dumping. Exports curtail themselves before anything in the city dims.

Source: US EIA — Wholesale Electricity Market Data

Export neighbor cap

How much the region will buy per day, and how fast that shrinks. The neighbours are building their own plants, so the cap melts 5% every January: export gravy is a melting asset, never an annuity. The Waste Compact’s ratchet, mirrored downward.

⚠️ Source pending

Negawatt ranks

Class demand multipliers from the three efficiency policy ranks: LED street and civic lighting, home weatherization, commercial equipment standards. LED retrofits really do cut lighting load 50-65%; the US Weatherization Assistance Program averages about 10% per home; appliance and code standards land in the 8-12% range. A saved kWh/day is worth more than its rating suggests, because supply must be built to the PEAK.

Source: US DOE — Weatherization Assistance Program impact evaluation

Negawatt unit costs

Capex per unit TREATED, never a flat price: a flat $12M weatherization programme needs roughly 40,000 households before it beats solar, which means it is dead on arrival in every city that would want it. Priced per unit, all three ranks land near $90-95 per kWh/day saved at any city size — slightly above catalog-era solar on capex, and better on everything else: no land, no build time, no auras, and they shrink the January peak itself.

⚠️ Source pending