Plant Efficiency Summary & Energy Waste Quantification

Abriliam Consulting — Industrial Energy Management

This final notebook brings together the findings from the diagnostic series to quantify the total energy waste attributable to the identified operational faults. We classify inefficient hours by their dominant cause — chiller performance, pumping excess, or tower fan issues — and estimate the energy (MWh) tied to each fault category.

This is the "so what?" notebook — translating diagnostic findings into actionable business metrics.

Index(['oat_C', 'wb_C', 'occ', 'tons', 'chw_sup_C', 'chw_ret_C', 'chw_dT_C',
       'chw_flow_m3h', 'cw_sup_C', 'cw_ret_C', 'cw_dT_C', 'cw_flow_m3h',
       'approach_C', 'dp_kpa', 'chiller_kw', 'tower_fan_kw', 'chw_pump_kw',
       'cw_pump_kw', 'plant_kw', 'kw_per_ton', 'plant_kw_per_ton',
       'tower_fan_kw_per_ton', 'pumping_kw_per_ton', 'kw_per_ton_15_sma',
       'kw_per_ton_5_sma', 'kw_per_ton_24_sma', 'kw_per_ton_240_sma',
       'chiller_kw_per_ton', 'free_cooling_candidate', 'wb_margin_C',
       'free_cooling', 'freecool_possible_oat', 'freecool_threshold_C',
       'calculated_tons'],
      dtype='object')

=== Energy tied to inefficient operation ===
Timestep assumed: 1.000 h per row
Rows analyzed (after cleaning + load gate): 1,344
Bad rows (plant kW/ton > 4.0): 233 (17.3% of hours)
Energy during bad rows: 99.1 MWh (10.2% of total 971.5 MWh)

=== Subsystem dominance among bad hours ===
 chiller-dominated: 233 hours (100.0%)

Average share of plant kW/ton during bad hours (by component):
      chiller_kw_per_ton: 88.0%
      pumping_kw_per_ton: 6.1%
    tower_fan_kw_per_ton: 5.9%

Energy Waste Quantification

Flagging all hours where plant kW/ton exceeds 4.0 and summing their energy consumption provides a conservative estimate of the energy tied to inefficient operation. The subsystem "blame assignment" uses dominance analysis — if a single subsystem (chiller, pumping, or tower fans) accounts for more than 50% of the plant's kW/ton during a bad hour, that hour is attributed to that subsystem.

The bar charts show:

Hours breakdown: Which subsystem is most often responsible for inefficiency
Average component share: How the plant's per-ton energy consumption splits across subsystems during inefficient periods

This analysis directly informs the priority of corrective actions — addressing the dominant source of waste first yields the largest savings.

CW Flow During Inefficient Hours

Examining condenser water flow during inefficient hours reveals a potential minimum flow constraint. The horizontal dashed line at 40 m³/h suggests a CW pump minimum — even at very low loads, the condenser water flow doesn't drop below this floor. This fixed flow at low loads contributes to the pumping-dominated inefficiency seen in the classification above.

Summary of Findings

Across this seven-notebook diagnostic series, we identified three distinct operational faults in the chiller plant:

Low delta-T syndrome — CHW temperature differential collapsed from ~6°C to ~4.2°C around early July, forcing higher pump flows and degrading chiller evaporator performance
Cooling tower degradation — Approach temperature drifted upward by approximately 1.5°C over the 8-week period, increasing condenser pressure and chiller power consumption
Night-time fan control bias — Tower fans operate at elevated power during overnight hours despite minimal cooling loads

Recommended Actions

Priority	Action	Expected Impact
1	Investigate and correct the cause of low CHW delta-T (likely a control valve or bypass issue)	Reduce pump energy by 20-30%, improve chiller efficiency
2	Inspect and clean cooling tower fill; check fan belt tension and blade pitch	Recover 1-1.5°C of approach, reducing chiller power
3	Implement load-based fan speed control with overnight setback	Eliminate unnecessary fan energy during low-load hours
4	Review CW pump minimum flow setpoint	Right-size the flow floor to actual minimum condenser requirements