The heating goes quiet.
Switch the heater off on a cold day. Does the indoor temperature instantly follow the outdoor temperature? Watch the slope of the indoor trace.
FIELD NOTE 08 / HOUSE HEAT LAB
A house holds the heat.
Follow the warmth. Find where it escapes. Give it a reason to stay.
0.0 hours
Outside −19.2 °CDay 1 · 06:00
Heat leavingHeat enteringSolar gain
Arrows show the direction of heat transfer; labels give the rates. The illustration is schematic. Living space and basement share one model temperature.
- Indoor temperature
- 20.0 °C
- Heater output now
- 8.00 kW
- Heater energy used
- 0.0 kWh
FOLLOW ONE PATH
Temperature through time
°C · elapsed hoursIndoorsOutdoorsThermostat targetSaved indoor trace
The whole house begins at 20 °C. Its stored warmth changes gradually.
Energy through time
kWh · elapsed hoursHeater electricitySolar heat admittedSaved heater trace
The heater supplies 1 kWh of heat per kWh of electricity. Solar energy is a separate gain.
Keep a run, change one thing, then Reset to compare from the same 20 °C start.
Where the energy goes now
Positive = into the houseSpace heater—
Sunlight—
People & appliances—
Walls—
Roof / ceiling—
Windows—
Air exchange—
Basement / ground—
Net stored in the house—
The net input controls how quickly the temperature changes.
Try a change
Ready. Press Play, then change the house while time keeps moving.
Make a prediction. Change one thing. See what happens.
Model notes & references ↗FOLLOW THE SCIENCE
A warmer inside.
An energy story.
On a cold evening, a warm house is continually transferring energy to its surroundings. Its heater replenishes that energy. Insulation slows the transfer, while the materials inside store warmth. Turn the heating off and those two properties help determine how quickly the temperature falls.
On a sunny day, the windows add another path for energy to arrive. The same insulation that slows winter heat loss also slows summer heat gain through the envelope. Shading reduces the sunlight entering through the glass; insulation alone does not remove heat already inside.
A FEW QUESTIONS TO TRY
Let’s see what happens.
Give the warmth more time.
Improve the walls, ceiling, basement, windows and air tightness together. During an outage, cooling slows; with the thermostat on, the heater can do less work.
A warm welcome for the Sun?
Try a warm day with strong sunshine and heating off. Then add exterior shading. Watch how solar gains and the indoor temperature respond at different speeds.
These actions change conditions from the current moment. “Keep this run” saves its plotted traces; Reset starts a new run while retaining the comparison. For a fair comparison, use the same weather, starting temperature, and elapsed duration.
Three ideas that work together.
RESISTANCE
Slower heat transfer.
Thermal resistance describes how strongly an assembly resists heat flow. At the same area and temperature difference, doubling its effective RSI halves its steady heat-transfer rate. Gaps and framing matter, so the controls represent whole assemblies rather than insulation batts alone. Australian Government: insulation ↗
STORAGE
A slower temperature change.
Air, furniture and building materials store energy. A larger heat capacity requires more net energy to change their temperature by one degree. Thermal mass and insulation describe different properties: storage and resistance. Australian Government: thermal mass ↗
A steady temperature can hide a busy exchange.
When the heater and other gains exactly balance heat leaving, the indoor temperature holds steady. Energy is still moving. Watch the flow arrows and the heater counter while the temperature trace is flat.
The air takes energy with it.
Replacing warm indoor air with cold outdoor air creates a heating load. In this model, the air-change rate specifies how much outdoor air enters during normal operation. Insulating a wall does not close an air leak. Real homes also need planned ventilation, and heat recovery can reduce its thermal load; the experiment represents air exchange without recovery. Australian Government: ventilation and airtightness ↗
A window has two different jobs in the balance.
The U-value describes heat transfer caused by a temperature difference; lower values mean less transfer. The solar heat gain coefficient describes the fraction of incident solar energy admitted as heat. Exterior shading reduces the sunlight that reaches the glass. The example windows change both U and solar gain, so compare their effects in darkness first to isolate their insulating performance. Australian Government: glazing ↗
Can the ground warm the house?
Yes, if the house becomes colder than the ground reservoir. The ground arrow then points inward. Here the reservoir stays at 8 °C, with a simple resistance between it and the basement. To explore how the soil itself warms, cools and freezes, visit The ground remembers ↗.
The mathematics, assumptions & references
One temperature, a complete energy balance
C dTin/dt = Q̇heater + Q̇sun + Q̇internal
+ Σ Hj(Tboundary,j − Tin)
+ Σ Hj(Tboundary,j − Tin)
Each heat-flow rate is in watts, positive into the house. The main floor, basement and participating thermal mass share one temperature and a fixed effective heat capacity C = 18 MJ/K (5 kWh/K). This deliberately combines air and stored heat in materials; it does not resolve room temperatures, surface temperatures, humidity, air circulation or delayed transfer inside individual walls. The illustrated attic is unconditioned; its heat path is represented by the ceiling resistance.
Hopaque = A / RSI · Hwindow = U A
Hair = ρ cp V × ACH / 3600
Q̇sun = I Aexposed SHGC (1 − shade)
Hair = ρ cp V × ACH / 3600
Q̇sun = I Aexposed SHGC (1 − shade)
RSI has SI units m²·K/W. The familiar North American R-value is approximately 5.678 × RSI. Whole-assembly resistances include their effective surface films and thermal bridges. The air model uses ρ = 1.2 kg/m³ and cp = 1005 J/(kg·K), with a constant 500 m³ indoor volume and equal incoming and outgoing air flow. ACH means operational air changes per hour, not air changes measured at 50 Pa.
| Part | Model value |
|---|---|
| House footprint | 10 × 10 m; one 2.5 m storey and a 2.5 m basement |
| Opaque outside walls, including doors | 80 m² (100 m² gross walls minus 20 m² windows) |
| Ceiling below attic | 100 m², connected directly to outdoor temperature through effective RSI |
| Windows | 20 m² total; 10 m² receives the prescribed sun |
| Basement walls + slab | 200 m²; user RSI plus a fixed 1.5 m²·K/W soil resistance to 8 °C |
| Single / double / triple examples | U = 5.7 / 2.7 / 1.1 W/(m²·K); SHGC = 0.72 / 0.60 / 0.50 |
These are illustrative values, not ratings for particular products. Window position, the pictured wall thicknesses, furnishings and heating equipment do not define model geometry or material properties. The ground is an imposed reservoir, not a linked soil simulation. Solar heating of opaque outside surfaces, sky radiation, wind-dependent leakage, moisture, hot water demand, mechanical cooling and heat-recovery ventilation are omitted.
Weather, the thermostat, and the energy counter
“Hold steady” maintains the chosen outdoor temperature and window irradiance continuously. “Day & night” starts at 06:00, gives the air a sinusoidal daily swing with a maximum at 15:00, and uses a half-sine solar profile between 06:00 and 18:00. This is controlled forcing, not local weather or a latitude-dependent solar model. Exterior shade is a fractional reduction of window irradiance.
The heater is an ideal electric resistance source, with 100% of its electricity becoming heat inside. Below the target it runs up to the selected capacity; at the target it supplies only the heat needed, if capacity permits. It switches off above the target and provides no cooling. There is no thermostat deadband or startup delay. The meter integrates actual heater output through time, excludes people and appliance gains, and starts at zero on Reset.
Why the curve bends
τ = C / ΣH · T(t) = Teq + [T(0) − Teq] e−t/τ
With constant surroundings and fixed heating input, the temperature approaches an equilibrium exponentially. τ is the time to cover about 63% of the remaining change, not the time to reach equilibrium. The model integrates that solution exactly over substeps of at most two minutes, samples changing weather at their midpoints, and resolves thermostat crossings within each substep. The same temperature integral accounts for energy through every boundary. Weather and control changes preserve the current stored energy; only Reset reinitializes it. Runs stop after seven model days.
The saved comparison contains the earlier run’s actual trace, with any setting changes that happened during it. It does not recompute a matched reference case. All values are simulated; the physics and time-series data are separated from the display so measured sensor data can be compared in a future addition.