On a BMS trend from a room specified at 20 ± 0.5 °C, the chilled-water valve and the reheat valve are sometimes open in the same minute.
The cause is usually in the control sequence rather than the tolerance. Cooling and reheat often run off one room temperature loop, with a gap between the two setpoints: the controller deadband. Set that gap to zero, or let the two sequences overlap, and a small overshoot on cooling brings on reheat, which overshoots and opens the cooling valve again. Nothing alarms and the room stays in spec; only the chiller and the reheat pay.
The product tolerance limits how wide the gap can be. Inside ±0.5 °C it is a few tenths of a degree, and sensor accuracy and valve resolution use up most of it. Where the product data supports 20 ± 2 °C, cooling can act above 21 °C and reheat below 19 °C, and between them the room drifts with both valves shut.
A wider band has to come from the product data and the URS, and it goes through change control. In a tropical plant there is a second limit: the coil often cools to dew point for humidity, so reheat after it stays whatever the temperature band is.
Before widening anything, overlay the two valve outputs on one trend. A zero gap can be fixed on the tight band too.
At Humiley the URS is written alongside the design, so the temperature band has a source before the controls are configured.
What temperature band does your URS state, and where did that number come from?




