An ideal chiller gains about 3.3% in COP for each kelvin its evaporating temperature rises, for example from 5 °C to 6 °C.
The arithmetic is the Carnot limit: evaporating temperature divided by the lift, both in kelvin. With condensing at 40 °C, 5 °C evaporating gives 278.15 / 35 = 7.95. At 6 °C it is 279.15 / 34 = 8.21. That is roughly 3.3% per kelvin.
Real chillers sit well below Carnot and each has its own curve, so 3.3% is not a saving to quote. The direction holds: every kelvin of lift removed eases the compressor.
In a GMP plant in the tropics, the chilled-water temperature is usually set by the one coil that needs it cold: the fresh-air coil doing the dehumidification. The recirculation coils, which only handle sensible load, could often work on much warmer water. Run the whole plant at the fresh-air coil's temperature and they pay for cold they do not use.
There are two ways out. Split the plant so a small low-temperature circuit serves the fresh-air coils and the rest runs warmer, or dry the fresh air by another route and raise the whole loop. Which one pays depends on the ratio of fresh air to recirculated air and on how much reheat the design already carries.
In our designs, dehumidification staging and the reheat penalty are primary design decisions, and the chilled-water temperature follows from them.
What chilled-water supply temperature does your plant run, and which coil sets it?
Industries




