How Much Air Conditioning Can Cool Nights Replace? A 153-City Answer

National Analysis: All 50 States¶

To build a comprehensive national map of passive cooling potential, we analyze daily weather data ($T_\text{max}$, $T_\text{min}$, and dewpoint) for ~50 major US metros, one or two per state, covering every climate zone.

Commercial buildings solved this decades ago. When outdoor air drops below a setpoint, building management systems open economizer dampers and pull in free cooling — mandated by ASHRAE 90.1 in most US climate zones. Homes have never had an equivalent: their only air pathway is the window, and until now there has been no affordable, retrofit, homeowner-installable automation layer to manage it.

The map clearly shows the opportunity zones for passive cooling: a broad band across the West, from inland California through Nevada, Idaho, Utah, Colorado, and New Mexico, plus less obvious pockets like the northern Plains, where dry air keeps summer nights cool. Every metro clearing 100 usable nights a year sits inside that band.

A second tier runs east of it. Indianapolis, Columbus, Kansas City, Philadelphia, and Charlotte each land between 60 and 76 nights, and the character of those nights is different. Sacramento vents 24 to 30 nights in every summer month, close to nightly. Indianapolis vents 9 to 12, and Atlanta's July count is zero: the humid interior and Southeast give up their usable nights in spring and fall, on either side of the weeks when air conditioners actually run hardest.

For a homeowner in these states the benefit is immediate and personal. For a utility, the same map is a service-territory screening tool: everywhere orange to red, night ventilation can displace real summer load at essentially zero marginal cost.

What's This Worth in Dollar Terms?¶

Automated ventilation doesn't eliminate AC. They reduce total AC runtime by pre-cooling the house overnight. In the target climates, the thermal simulation above puts that reduction at 25-46% of annual AC hours (Sacramento 27%, Denver 46%). Humid metros see essentially none, which is why the economics below focus on the climates where night cooling works.

To convert AC hours avoided into dollars, we use a standard 3-ton central AC system (the median US home size), which draws 3.5 kW but runs at about 50% duty cycle, so effectively 1.75 kW per hour of AC runtime.

A critical detail: when an avoided hour occurs determines what it's worth. Time-of-Use (TOU) utilities charge 2-3× more during the evening peak than off-peak. PG&E's E-TOU-C summer peak (4-9 PM) runs about 42¢/kWh against 33¢ off-peak. The simulation tracks the hour of day of every avoided AC-hour, and the histogram below shows that the avoided hours cluster in the afternoon, with roughly half landing inside the 4-9 PM peak window. Each avoided hour in the table is priced at the applicable TOU rate for that metro's dominant utility. Pre-cooling doesn't just cut AC runtime; it preferentially cuts the most expensive runtime.

The table uses the actual AC hours avoided from the thermal simulation for each city, scales across metro areas using EIA RECS 2020 [5] AC penetration rates, and converts avoided kWh to CO₂ at the US grid-average intensity (EPA eGRID [13]).

The late-afternoon concentration means the avoided hours are the ones that cost households the most and strain the grid the hardest: pre-cooling defers the hour when the house first needs AC until late in the evening peak, or past it entirely. Automated night ventilation delivers both bulk energy savings and peak-hour relief, making it a natural fit for utility rebate and demand-response programs.

California in Detail¶

Fourteen California metros, hour by hour on 2025 weather: the Bay Area and Sacramento, the Central Valley from Redding to Bakersfield, and the inland basins behind the Los Angeles and San Diego coasts. Same thermal model and duty-cycle assumption as the national table, each metro priced on its serving utility's summer time-of-use rates.

California is worth $1.89B a year across 8.7 million air-conditioned households, and about 5,661 GWh of avoided summer generation.

That figure and the global table later in this analysis are not nested. The global set contains three Californian metros — Los Angeles, Sacramento and San Jose — and reaches about $0.8B of the state. The eleven metros simulated here that it does not contain, Riverside and the Central Valley among them, are most of the difference.

The strongest markets are inland. Riverside clears 199 venting nights, more than any metro here, and avoids 435 AC-hours per home. The Central Valley matches it: Modesto avoids 480 AC-hours, Stockton 454 and Fresno 371, against Sacramento's 475 and the Bay Area's 363.

Climate and money rank differently. Modesto avoids more AC-hours than any metro in the state, 480 a year, and returns $185 per household; Stockton, seventy miles north in the same Valley climate, avoids 454 and returns $292, because Stockton is PG&E territory and Modesto's public utility charges a flat rate near half of PG&E's peak. Los Angeles carries the largest household count and the largest absolute figure, $547M, on a per-home result held down by the coastal share of the metro where nights are already mild.

What This Means for the Grid¶

For utilities, coincident peak demand — the late-afternoon window that sets capacity requirements — is an important driver of what they must build and buy. Passive ventilation through smarter windows produces two savings against it:

1. Peak deferral on hot days. A pre-cooled house crosses the AC setpoint about two hours later, and on a fifth to a third of would-be AC days it never crosses it at all; the deferred hours fall precisely in the 2-6 PM window when the grid is climbing toward its daily maximum.
2. Full-evening elimination on moderate days. On the many days when pre-cooling keeps the house below the setpoint entirely, the closed-up house would have been running AC straight through the 4-9 PM peak and the smart house never turns it on.

The table below converts the simulated peak-window avoided hours into the average power shaved during the utility's peak window on vent days, scaled to full adoption across each metro's AC-equipped households (metros with fewer than 30 vent days a year are excluded; they aren't target markets). These figures are a lower bound: an hour-counting model excludes partial duty-cycle relief — a pre-cooled house that still needs some AC at 5 PM runs its compressor less than a warm one — an effect utility pre-cooling studies indicate is comparable or larger.

Two rows of that table are Californian, the Bay Area and Greater Sacramento, and between them they reach about 4.5 million households. That is well short of the state: the Central Valley from Redding to Bakersfield, and the inland basins behind the Los Angeles and San Diego coasts, are the strongest night-cooling climates in the country and none of them appear above. The next section simulates all of them.

The Global Picture¶

Below is the same analysis run across 120 major cities worldwide. Bubble size reflects metro population, a rough proxy for market volume. Color reflects Smart Venting Days, or climate fit.

The opportunity clusters in a specific climate band. Hot enough to need cooling, but with nights that actually cool down. Climatologists recognize five Mediterranean-climate zones on Earth, and all five light up on this map: the Mediterranean basin itself, coastal and inland California, central Chile (Santiago and Valparaíso), South Africa's Western Cape (Cape Town), and southern Australia (Perth, Adelaide, Melbourne). The Atlantic edge of the Iberian Peninsula (Porto and Galicia) shows the same signature of warm, dry days and cool marine nights, as do high-altitude dry cities like Mendoza, Arequipa, and Windhoek, the Mountain West, and parts of the Middle East. Humid tropical cities don't register, because nights there stay warm and sticky year-round, and opening windows wouldn't help. Northwestern European cities barely register either, since they rarely get hot enough in the first place.

Ranking by vent-days × population puts numbers behind the colors. Mexico City tops the list, with 195 usable nights a year and 22 million people in a high-altitude climate practically designed for night cooling, followed by Cairo, Delhi, and São Paulo. Many of the world's fastest-growing air-conditioning markets are also its best passive-cooling markets, which means smarter windows can grow with the global AC boom and reduce its grid and carbon impact.

Market Sizing by Metro¶

Pulling everything together: vent days (how good the climate is), estimated AC-hours avoided (from the thermal model), household counts (market size), and local electricity rates (dollar impact).

For cities outside the 17-city hourly dataset we can't run the thermal simulation directly, so we scale by vent days using the ratio the simulation actually produced across the simulated cities: about 2.7 AC-hours avoided per vent day (computed below, not assumed). That ratio is a conservative one: it averages in shoulder-season vent days that displace only an hour or two of AC. Savings are then billed only to households that already own air conditioning, at country-level ownership rates [22].

These numbers represent the energy savings ceiling for each metro at widespread adoption — market sizing, not a sales forecast. Adoption takes time, and not every home has operable windows.

The Other Half: Fresh Air¶

Everything so far measures one thing — air conditioning displaced. That is the benefit that shows up on a utility bill, and it is why the map above favours the arid West. But it is not the reason most people open a window.

The other benefit is that outdoor air is often pleasant, and a sealed house never gets any of it. Here we analyse a different question: not whether an open window would displace air conditioning, but whether you would want it open at all. We count it the same way, hour by hour, against a comfort band rather than a cooling threshold.

Comfortable open-window criteria. Daytime runs warmer than overnight because people are dressed and active by day and have bedding at night. The dewpoint ceiling is looser than the 60°F used for cooling: drawing in fresh air tolerates more moisture than chilling a house with it. Rain and wind gates reflect conditions a window would be closed for regardless of temperature.

Across 153 cities the median is 1,770 comfortable hours a year, roughly a fifth of all the hours there are, and 142 of the 153 clear a thousand. Nairobi tops the list at 5,475 hours, with Medellín, Lima and Guadalajara close behind — moderate elevation in the tropics does well generally, altitude holding a hot latitude to a mild temperature the year round. Go higher still and it reverses: Bogotá, at 2,600 m, ranks 142nd because it is more often too cool than too warm.

There is a real floor. Jakarta, Manila and Singapore record no comfortable hours at all in 2025 under these criteria: on the equator the temperature never leaves the humid band, day or night, in any month. Bangkok manages 400 hours.

Why it needs a motor¶

Those hours do not arrive in convenient blocks. In the median city they come in 255 separate stretches across the year, which is how many times somebody would have to open a window and remember to close it again. At the extreme, Nairobi's 5,475 hours arrive in 897 stretches, two to three a day with a median length of four hours: highland nights dip below the sleeping band, middays rise above the daytime one, and equatorial rain interrupts what remains.

What the hours are worth¶

Ventilation dilutes what accumulates indoors: CO₂ from breathing, humidity from cooking and washing, and the volatile compounds released by furnishings and cleaning products. A closed bedroom concentrates all three overnight. In a controlled trial, simply opening the window took average overnight CO₂ from about 2,585 ppm to 660 ppm, and measured sleep quality, next-morning sleepiness and performance on a logical-thinking test all improved with it [20].

Temperature matters for sleep in its own right. Core body temperature has to fall for sleep to begin and stay consolidated, and the recommended bedroom range is 60-67°F, with roughly 65°F as the optimum [21]. That is close to where a venting night leaves a house by morning, and several degrees below where a thermostat would hold it.

In most temperate cities the air outside is better than the air inside for close to two thousand hours a year, spread across roughly 255 separate stretches.

Is 2025 a Fair Year?¶

Every figure in this analysis uses calendar year 2025. To confirm that 2025 is a representative year rather than an unusual one, we re-ran the 17 cities with hourly data across 2022 through 2025 under the identical criterion.

The high-volume markets are steady. The nine cities averaging 90 or more usable nights move by 10 to 23 percent of their mean across four years, and 2025 sits inside that band for all of them: Sacramento's 180 is three nights above its four-year average, San Jose and Boise run higher, Albuquerque and Madrid lower. Summed across those nine, 2025 lands within one percent of the four-year average, so nothing in the headline figures depends on the year chosen.

The marginal cities move far more. London, Paris, Miami and Houston each swing by roughly 70 percent of their mean or more, and a single year says little about any of them individually. In 2025 most of them ran above their averages rather than below: London recorded 48 usable nights against a four-year mean of 36, Tel Aviv 50 against 40. Madrid and Atlanta went the other way, at 93 against 110 and 45 against 55.

We report 2025 throughout rather than a four-year average because the US and global datasets behind the maps are single-year, and mixing a four-year basis for 17 cities with a one-year basis for the other 155 would be worse than either alone. Single-city figures below roughly 90 days are better read as a range than a point estimate.

Assumptions & Limitations¶

This analysis makes several simplifying assumptions:

Baseline assumes zero manual ventilation. Our thermal model compares "AC only" against "automated ventilation + AC." In practice, a nationwide survey (Morrison et al., 2022) found that ~44% of US households report opening at least one window on any given day. However, this behavior peaks in the 64–70°F comfort range and drops sharply in hotter weather, which are the days that drive the majority of projected savings. Systematic overnight pre-cooling (opening once the evening air turns cooler than the house, closing before it warms back up) is not a documented behavior pattern, likely because it requires attention to timing that most people don't maintain consistently.

Single-zone thermal model. We model the house as a single thermal mass with τ = 22 hours. Real homes have room-by-room temperature variation, multi-story stratification, and varying insulation quality. The model is directionally correct but should not be treated as an engineering simulation.

Venting is almost entirely overnight. In the simulation 87% of open-window hours fall between 9 PM and 7 AM and 1.2% between 9 AM and 6 PM, because outdoor air is rarely below 70°F during a summer working day. The model does not represent daytime ventilation of an unoccupied house, and the savings reported here do not depend on it.

Fixed AC setpoint of 74°F. Cooling is called when indoor temperature passes 74°F, the top of the 72-74°F band households typically set. Every home is given the same setpoint and no schedule, where real thermostats are often set back overnight or when the house is empty. The DOE-recommended 78°F is reported alongside the results as a sensitivity.

Weather data is for one year (2025). Year-to-year variation exists, though the ERA5 reanalysis data represents typical conditions well.

No humidity discomfort modeling. We use a dewpoint ceiling of 60°F as a binary filter, applied to the mean dewpoint across the venting window. In practice, comfort is a spectrum, and some nights near the threshold may feel borderline.

Flat 1.75 kW per avoided AC-hour. A counted hour is an hour the thermostat calls for cooling, during which the compressor cycles; we convert at a flat 3.5 kW × 50% seasonal-average duty. Real duty cycle varies with load, and many avoided hours are lighter-load morning hours that would draw less than the flat figure. The kWh numbers are therefore approximations, partially offset by deliberately conservative metro household counts.

Appendix: How the Thermal Model Works¶

The thermal simulation used in this analysis models your house as a single object that slowly absorbs heat from outside. Here's an intuitive way to understand it.

The Coffee Cup Analogy¶

Imagine you put a hot cup of coffee in a cold room. It cools off, but not all at once. At first, when the coffee is much hotter than the room, it cools quickly. As the coffee gets closer to room temperature, the cooling slows down. It never instantly reaches room temperature.

Your house works the same way, but in reverse during summer. A pre-cooled house (65°F inside, 95°F outside) slowly absorbs heat through the walls, roof, and windows. The bigger the temperature difference between inside and outside, the faster heat flows in.

What Is τ (Tau)?¶

τ (tau) is the thermal time constant of your house. It captures how well-insulated your home is in a single number.

Practically: τ is the number of hours it takes for the indoor-outdoor temperature gap to shrink by about 63%.

Diurnal effective thermal time constants used in this analysis. Derived as τ = C/H for a 2,000 sq ft wood-frame envelope; τ = 22 h corresponds to an effective heat capacity of about 108 kJ/m²K of floor area, between the "light" and "medium" classes of EN ISO 13790 [15].

Two clarifications, because published figures vary by nearly an order of magnitude.

First, these are diurnal time constants. Smart-thermostat studies of real housing report medians near 51 hours across 60,000 North American homes and 59 hours across 11,000 Canadian homes [16][17], but those fit multi-day free-float decays, which are governed by the whole structural mass. Only mass within the periodic penetration depth — roughly 10 cm of concrete for a 24-hour cycle, per ISO 13786 [18] — takes part in day-to-night cycling. The capacitance that matters for a diurnal model is therefore smaller, and so is τ. As a check, τ = 22 damps the simulated indoor swing to about 27% of the outdoor swing, the expected range for light-to-medium construction; τ = 15 gives 35 to 49%, too fast even for light construction.

Second, none of these are Passive House numbers. A certified Passive House runs at about τ = 90 hours for lightweight timber frame and τ = 200 hours for heavy masonry with exterior insulation [19]. Those buildings are far outside the existing housing stock this analysis is about, and they would show substantially larger benefits than anything reported here.

Why τ = 22 Doesn't Mean "Stays Cool for 22 Hours"¶

τ = 22 means the gap between indoor and outdoor shrinks by 63% in 22 hours. It does not mean the house holds its temperature for 22 hours and then changes. The rate depends on how large the gap is at that moment: each hour, indoor temperature moves toward outdoor by

$$\text{Hourly drift} = \frac{T_\text{outdoor} - T_\text{indoor}}{\tau}$$

So with τ = 22, indoor at 65°F, outdoor at 95°F:

• The gap is 30°F, so the house warms at 30/22 = 1.4°F per hour
• After 3 hours: indoor is ~69°F, gap is 26°F, drift slows to 1.2°F/hr
• After 6 hours: indoor is ~72°F, still just below the 74°F setpoint
• The setpoint is not reached until about 12 hours in

The house warmed about 7°F in 6 hours, not the full 30. That's because the rate is proportional to the current gap. Big gap = fast warming. Small gap = slow warming. This is exponential decay, and it's the same physics that governs everything from coffee cooling to radioactive decay.

What Makes τ Bigger or Smaller?¶

τ depends on two things:

1. Thermal mass (how much "stuff" inside the house absorbs and holds temperature: concrete slabs, furniture, drywall). More mass = bigger τ.
2. Envelope heat loss (how readily heat crosses the walls, roof, windows and air leaks). Less loss = bigger τ.

The formula is $\tau = C / H$, where $C$ is the effective thermal capacitance in joules per kelvin and $H$ is the overall heat-loss coefficient in watts per kelvin, giving a time in seconds. $H$ is the sum of conduction through the walls, roof and windows plus air infiltration, and it is the quantity building standards and energy models tabulate, which is why this analysis works in those terms. The same relationship is sometimes written as the electrical analogy $\tau = R \times C$, with $R = 1/H$ the thermal resistance; the two are identical. A house with thick walls, good windows, and a concrete slab floor has a much higher τ than a poorly insulated wood-frame house with single-pane windows.

The Physics: Newton's Law of Cooling¶

The thermal model is based on Newton's Law of Cooling, which states that the rate of heat loss of a body is proportional to the temperature difference between the body and its surroundings. The differential equation is:

$$\frac{dT}{dt} = -\frac{1}{\tau}\,\bigl(T_\text{indoor} - T_\text{outdoor}\bigr)$$

Where:

• $T_\text{indoor}$ is the current indoor temperature
• $T_\text{outdoor}$ is the current outdoor temperature
• $\tau$ is the thermal time constant (hours)
• $dT/dt$ is how fast the indoor temperature is changing (°F per hour)

The solution to this equation is an exponential decay toward the outdoor temperature:

$$T(t) = T_\text{outdoor} + \bigl(T_\text{initial} - T_\text{outdoor}\bigr)\, e^{-t/\tau}$$

This is the same equation that governs coffee cooling, radioactive decay, and capacitor discharge in electronics. The key insight: the rate of temperature change depends on the current gap between inside and outside, not just time.

How Night Ventilation Delays AC on a Hot Day¶

The key comparison: on a hot summer day, how much longer can you avoid turning on AC if the windows opened themselves overnight?

The plot below simulates a typical Sacramento July day (low 63°F, high 100°F) with typical construction (τ=22). The red line shows a house with no automation: the house sat closed overnight and starts the day at 72°F, having drifted down only a couple of degrees through the closed envelope from the 74°F it was held at the previous evening. The green line shows the same house with an actuator that opened the windows overnight, pre-cooling to 65°F. The gap between when each line crosses the 74°F setpoint is the AC delay.

The Bottom Line¶

Every night, in hundreds of cities, as temperatures drop, a free cooling opportunity is created by nature, but it requires opening and closing windows at the right times, every day. Commercial buildings have mandated economizers that automatically cool building interiors and now theire is an economical solution for residential homes with the Smarter Window.

Aggregated across the cities in this analysis:

© 2026 Y. Dan Rubinstein. Published via Smarter Window. Data retrieved May 2026. All weather data from Open-Meteo ERA5 reanalysis for calendar year 2025.