The load side and the weather side move on their own. Watch how each one degrades a different health signal.
Move cooling load or coil fouling and the ΔT (top) reacts — but efficiency barely moves. Move wet-bulb and the kW/ton (bottom) climbs while ΔT stays put. That's the lesson: a hot, humid day makes the chiller work harder for the same cooling — a real degradation in efficiency, but a normal, weather-driven one, not a fault. A health system has to subtract out the weather before it can tell whether the plant is actually getting sick. Load is the main trigger; weather is the independent second one.