The Verification Venue · a thing half the house gets wrong

The Dial That Doesn't Hurry

You come home to a cold house and shove the thermostat to 90, figuring it'll warm up faster and you'll dial it back once it's cozy. It won't. A single-stage furnace has exactly one throttle (full), and the number on the dial only tells it when to stop, not how hard to blow.

Here's the physics in one line. A room is a bucket of heat: your furnace pours warmth in at a fixed rate, and the cold outside drains it at a rate set by how leaky the house is and how big the temperature gap is. Written out, the room's temperature T obeys one equation:

C · dT/dt = P_furnace − UA · (T − T_outside)

Read the right-hand side. P_furnace is fixed. UA (leakiness) is fixed. T_outside is fixed. C (thermal mass) is fixed. The target number is nowhere in it. So the warm-up curve, how fast the room actually heats, cannot possibly depend on whether you asked for 71 or 90. The setpoint is just a line where the furnace switches off. Drag it and watch.

Your heating system

Time to reach 70 °F

vs. setting just 71°

Furnace output right now

whenever it's running

Wasted overshoot band

heat you never needed

Drag from 71 to 90. On a single-stage furnace the orange warm-up curve does not move; it sits exactly on the grey "set it to 71" ghost. Only the dashed shut-off line climbs, dragging a red band of run-time you'll pay for and never feel.

The free choices (name them, then change them)

These are example inputs, not universal constants. The conclusion doesn't hinge on any one of them; move every slider and the single-stage curve still won't care about the dial. Default furnace output ≈ a mid-size 60,000 BTU/hr delivered (a ~72,000 BTU/hr input unit at ~83% efficiency).

Where the fear is partly right

Flip the toggle above to Heat pump and the ghost and the live curve split apart: the crank does warm the room faster. This is the honest twist, and pretending otherwise is how a heat-pump owner catches a page lying. Not every system has one throttle:

Heat pumps. A heat pump sips electricity and pumps ~2–3 units of heat for every unit it draws: cheap, but gentle, and it fades as it gets colder outside. When you ask for a temperature more than about 2–3 °F above the room, the thermostat decides the pump can't keep up and fires the auxiliary electric-resistance strips, the glowing-coil kind. Those genuinely add heat faster. They are also the most expensive heat in the house (resistance heat is ~1 unit out per unit in, roughly 2.8× the running cost of the compressor). A big overnight setback is exactly the situation that trips them on recovery, so "set it back at night to save money" and "the strips ran all morning" can both be true.

Two-stage and modulating furnaces. A two-stage furnace starts on low fire and jumps to high when it's far from the target (or after a long low-fire run). A modulating furnace ramps continuously from ~40% to 100% by distance from the setpoint. On both, a bigger gap really does mean more output, so cranking bends the early curve upward. Watch the Modulating curve peel above the ghost as you raise the dial.

So the complete answer isn't "no." It's: identical speed on single-stage on/off gear; faster-but-wasteful on staged, modulating, and heat-pump systems. That system-type flip is the whole story.

The check: every number recomputed in front of you

The curve above is the live solution of C·dT/dt = P − UA·(T−T_out). For a single-stage furnace P is constant, so the equation has a closed form, T(t) = T∞ − (T∞−T₀)·e^(−t/τ) with τ = C/UA and T∞ = T_out + P/UA. Here it is with your current numbers:

Two independent methods must agree: the closed-form time-to-70 above, and a step-by-step RK4 integration of the very same ODE. Press the button to run the integrator live in your browser:

not run yet

Free choices & uncertainty. The four sliders (P, UA, C, T_out) plus start temp are all your inputs; the BTU figures are illustrative, not a spec. The heat-loss law P_loss = UA·ΔT and the single lumped mass C are the standard first-order idealisation (real houses have several coupled masses and a thermostat deadband). The heat-pump aux threshold is modelled at 2.5 °F (sources cite ~2–3°F) and the compressor derates linearly with cold. None of that touches the headline: for single-stage, the curve up to the target is byte-identical for every setpoint. Run it offline: node research/does-cranking-the-thermostat-heat-faster/verify-does-cranking-the-thermostat-heat-faster.mjs.

The complete answer, by system

SystemDoes cranking heat faster?WhyThe cost
Single-stage on/offNo: identical curveOne fixed output; dial only sets shut-offNone; overshoot wastes run-time
Two-stagePartly: high fire when far off2nd stage at ~1–2°F below target / long low-fire runMore gas while on high stage
ModulatingYes: output ∝ gapRamps 40→100% by distance from setpointRuns harder than it needed to
Heat pump + auxYes, but on strip heatGap > ~2–3°F energizes electric-resistance stripsThe priciest heat in the house
What's exactly true, what's idealised, and the cooling case

Exactly true. A single-stage furnace or AC compressor is an on/off device: it delivers the same rated output whenever it runs. A 100,000 BTU/hr furnace makes 100,000 BTU/hr at 2°F outside or 32°F outside; the thermostat only varies the on/off cycle length. Because none of the terms in C·dT/dt = P − UA·ΔT contains the setpoint, the warm-up trajectory is setpoint-independent for such systems. The dial sets where it stops, not how fast it climbs.

Idealised. This is a single lumped thermal mass (one C, one T). A real house is several coupled masses (air responds in minutes, walls and furniture in hours), and the thermostat has a small on/off deadband we've smoothed over. Those refine the exact minutes-to-70 but not the setpoint-independence. UA and C are genuine free parameters that vary house to house; that's why they're sliders you can move.

The cooling case is symmetric, with one caveat. Setting the AC to 60°F doesn't cool to 74°F any faster; the compressor removes heat at its fixed rate regardless. But an air conditioner also does latent work, wringing water out of the air (dehumidification), and comfort depends on humidity as well as temperature. This model tracks only sensible heat (temperature), so it captures the "cranking doesn't help" point but not the full clamminess story.

No cranking taboo. The claim is not "cranking never helps." On a heat pump it demonstrably does, by burning the expensive strips. The honest verdict is the system-type split, which is exactly what the toggle shows.