Abstract
Most operating chiller plants run well above their real part-load potential. The causes are structural, measurable, and bound by physical laws—none of which are visible when relying solely on full-load COP metrics. This paper deconstructs the thermodynamic realities of part-load operation, condenser lift, and heat exchanger degradation.
1. The Carnot Limit and Part-Load Reality
The theoretical maximum efficiency of any chiller is dictated by the Carnot limit. In reality, industrial and commercial chillers spend >85% of their operating hours between 40% and 80% of nameplate capacity. A machine optimized mechanically for full-load efficiency often sacrifices part-load performance due to fixed port geometries or aerodynamic stall in centrifugal impellers. Assessing true efficiency requires mapping the machine’s specific part-load curve against the facility's actual historical load-duration and condenser-water temperature profiles.
COP_ideal = T_evap / (T_cond - T_evap)2. Condenser Lift and Wet-Bulb Tracking
Compressor work is a direct function of the pressure differential (lift) between the evaporator and condenser. Fixed condenser-water setpoints (e.g., a static 30°C entering) inherently ignore ambient wet-bulb fluctuations. By implementing a reset strategy that actively tracks ambient wet-bulb temperature and manages cooling tower approach, facilities can safely reduce condenser entering temperatures. Thermodynamically, every 1°C reduction in lift typically yields a 2% to 3% reduction in compressor power (kW).
3. Sequencing via Marginal kW
Standard Plant Room Automation (PRA) logic sequences chillers based on total plant thermal load (TR) or simple return water temperatures. This is thermodynamically inefficient. Optimal sequencing requires staging based on marginal kW/TR. Depending on the isentropic efficiency curves of the specific compressors, running two variable-speed magnetic bearing machines at 50% load often requires less total electrical power than running a single machine at 100%, largely due to the massive reduction in heat exchanger approach temperatures at partial flow.
4. The Affinity Laws in Auxiliary Systems
Primary and secondary chilled water pumps, alongside cooling tower fans, often represent 15-25% of total plant energy. Operating these at constant speed during part-load conditions violates fundamental fluid mechanics. According to the pump affinity laws, power consumption is proportional to the cube of the shaft speed. A 20% reduction in flow requirements allows a nearly 50% reduction in auxiliary power, provided VFDs are actively modulated by differential pressure or thermal demand.
P1 / P2 = (N1 / N2)^35. Thermal Degradation and LMTD
Heat exchanger performance decays due to waterside fouling, scaling, and non-condensable gas accumulation. As the fouling factor (Rf) increases, the overall heat transfer coefficient (U) decreases, forcing the compressor to increase lift to maintain the required heat transfer rate. Continuous calculation of LMTD (Log Mean Temperature Difference) against a clean-state baseline is the only reliable method to predict tube degradation before a high-pressure trip occurs.
Q = U * A * ΔT_LMTDTable 1: Theoretical kW/TR vs. Condenser Entering Water Temp (CEWT)
| Load % | CEWT 32°C | CEWT 29°C | CEWT 26°C | CEWT 23°C |
|---|---|---|---|---|
| 100% | 0.58 | 0.52 | 0.47 | 0.43 |
| 75% | 0.54 | 0.48 | 0.41 | 0.36 |
| 50% | 0.56 | 0.49 | 0.40 | 0.32 |
| 25% | 0.68 | 0.58 | 0.49 | 0.41 |