EXPERIMENT 01 / COOLING & RESILIENCE
Heat rises.
What happens next?
Three cooling units. A changing heat load. One decision at a time. See why spare capacity and a good response both matter.
CHOOSE YOUR EXPERIENCE
PREDICT → OBSERVE → INTERVENE
One unit down.
What would you change?
At 60 seconds, heat load rises to 440 kW. At 120 seconds, one of three 200 kW cooling units fails. Will the zone cross 27°C within ten minutes?
Choose a prediction to begin.
- Heat arriving
- 360kW
- Heat removed
- 363kW
- Available capacity
- 400kW
Temperature through the event
60 s: heat load changes. 120 s: selected units remain available.
- Peak temperature
- 24.75°C
- Time above 27°C
- 0 s
- Capacity minus load
- 40 kW
- Recovery after 120 s
- 61 s
Results cover the full 600 s run. Recovery means within ±0.5°C of 24°C for the rest of the run, with at least 30 s remaining. Capacity reserve is not a redundancy certification.
Inspect the numeric results
| Time (s) | Temperature (°C) | Heat load (kW) | Cooling (kW) | Capacity (kW) |
|---|---|---|---|---|
| 0 | 24.00 | 240 | 240.0 | 600 |
| 60 | 24.00 | 360 | 240.0 | 600 |
| 120 | 24.75 | 360 | 346.5 | 400 |
| 180 | 24.50 | 360 | 399.2 | 400 |
| 240 | 24.11 | 360 | 396.2 | 400 |
| 300 | 23.87 | 360 | 370.3 | 400 |
| 360 | 23.88 | 360 | 350.4 | 400 |
| 420 | 23.99 | 360 | 350.2 | 400 |
| 480 | 24.05 | 360 | 358.6 | 400 |
| 540 | 24.03 | 360 | 363.4 | 400 |
| 600 | 24.00 | 360 | 362.5 | 400 |
OPEN THE ENGINEERING NOTEBOOK
A heat balance you can inspect.
This independent experiment uses a single, well-mixed thermal zone with an effective heat capacity of 6,000 kJ/K. It begins at 24°C with a 240 kW heat load and matching cooling. Each of three idealized units can remove up to 200 kW.
C · dT/dt = Qload − Qcooling
A PI controller requests cooling from temperature error. Equipment responds with a first-order lag. Output is limited to the available units, with conditional integration to avoid windup. The 27°C limit is an illustrative scenario threshold, not an operating recommendation.
Sequence of operations & model assumptions
- Start at thermal equilibrium, with all three units available.
- At 60 s, change heat generation to the selected load.
- At 120 s, restrict available cooling to the selected unit count. Removed units contribute no further cooling.
- Continue to 600 s. Compare with the same controller and load while retaining all three units.
Responsive tuning: Kp = 100 kW/K, Ki = 2 kW/(K·s). Gradual tuning: Kp = 45, Ki = 0.4. Equipment lag is selected independently: 12 s or 45 s. Changing tuning leaves the equipment unchanged. Version 1 links retain their original paired settings. Controller updates every 0.5 s; cooling lag and its heat removal integral are evaluated analytically within each step.
Points list & acceptance checks
Points: zone temperature (°C), heat load (kW), requested cooling (kW), delivered cooling (kW), available capacity (kW), and elapsed time (s).
Automated model checks cover equilibrium, saturated heat balance, event boundaries, capacity constraints, numerical convergence and exact reproduction from a versioned link. Browser checks exercise scenario changes, numeric inspection, sharing and keyboard access.
What this model leaves out
No spatial hotspots, humidity, airflow, hydraulic network, heat-exchanger curves, ambient conditions, electrical failures or equipment-specific protection logic are modeled. There is no calibrated facility or operating dataset behind it. Cooling capacity is held constant within each equipment state. Use the experiment to explore a trade-off, not to size equipment or establish a redundancy rating.
Why this runs in your browser
A dedicated Web Worker computes the simulation while the page remains available for interaction. The initial result is rendered as HTML and SVG. Scenario links contain only versioned settings; running or sharing an experiment requires no account or cloud save.