Jul 2 baseline condenser-water pump frequency
Redrawn from the source figure: fixed-frequency operation around 48Hz, 47.6Hz, and 45.4Hz.
CwPump #1CwPump #2CwPump #4
Source chart shows stable fixed operation rather than active search.
Anonymized manufacturing case study
A pharmaceutical cooling plant improved reported average COP from 4.75 to 5.41.
This page preserves the report detail: equipment brands and sizes, BAS compatibility, control bands, daily COP and weather data, redrawn frequency charts, and the published COP-uplift calculation.
- Published COP uplift
- 12.11%
- AI operating window
- 37 days
- Average COP
- 4.75 -> 5.41
Published by the source report for the AI operating period.
Sep 1 to Oct 7 AI control compared with Jun 1 to Jul 7 autonomous control.
Baseline average to AI-period average.
Plant inventory
Equipment brands, capacities, and active plant topology
The report describes a secondary-pump variable-flow chilled-water system. The central plant includes four chillers with one standby unit, four chilled-water pumps, four condenser-water pumps, three air-side chilled-water pumps, and four process-side chilled-water pumps. Normal operation is one chiller paired with one pump.
| Equipment | Brand | Specification | Status |
|---|---|---|---|
| Centrifugal chiller | Trane | Cooling capacity 650RT / Power 402kW | Running |
| Chilled-water pump | Grundfos | Head 20m / Power 30kW / Flow 350m3/h | Running |
| Condenser-water pump | Xylem | Head 22m / Power 45kW / Flow 190m3/h | Running |
| Cooling tower | BAC | Flow 500m3/h / Power 22kW | Running |
Control and BAS compatibility
The AI layer stayed compatible with existing group control.
The platform supports one-click switching between existing group control and AI control. Operators can also apply AI recommendations to group-control setpoints for condenser-water pump frequency and cooling-tower frequency.
- Condenser-water pump band
- 42Hz to 50Hz, constrained by low-flow protection.
- Cooling-tower fan band
- 22Hz to 50Hz, optimized against plant energy and indoor-temperature response.
- Current optimization scope
- Condenser-water pumps and cooling towers.
- Reserved future interfaces
- Chilled-water supply temperature and chilled-water pump frequency.
- Data source
- BAS operating records, daily system COP, outdoor dry-bulb temperature, and outdoor wet-bulb temperature.
Redrawn Figure 2
Cooling-water pump and cooling-tower optimization loop
The source report frames the control as reinforcement learning: read operating state, check load and action limits, update strategy, and execute pump/tower actions.
- The reward combines overall operating energy and indoor-temperature change.
- The strategy directly targets condenser-side energy instead of controlling only an intermediate temperature-difference variable.
- 01
Optimization cycle starts
- 02
Read state parameters
- 03
Load above start threshold?
- 04
State parameters inside allowed action range?
- 05
Discretize state parameters
- 06
Update Q(s,a) with reward feedback
- 07
Use epsilon-greedy strategy for pump/tower command
- 08
Execute action
- 09
Optimization cycle ends
Sep 28 AI condenser-water pump frequency
Redrawn from the source figure: two running pumps moved within the 42Hz-50Hz control band.
CwPump #1CwPump #3
The AI period includes deliberate drops near mid-day and late evening, then returns to higher frequency when conditions require it.
Jul 2 baseline cooling-tower fan frequency
Redrawn from the source figure: tower fans held near fixed frequency levels.
Tower #1Tower #2Tower #3Tower #4
The baseline chart is nearly flat, with one visible transient dip.
Sep 28 AI cooling-tower fan frequency
Redrawn from the source figure: tower fans stepped down and up across the day.
Tower #4Tower #1Tower #3
The report text states AI tower operation reduced tower running energy by about 30%.
Calculation detail
COP uplift and operating-period calculation
- 1. Delta COP = AI average COP - baseline average COP = 5.41 - 4.75 = 0.66.
- 2. Published report uplift = 12.11%.
- 3. Displayed rounded averages reproduce approximately (5.41 - 4.75) / 5.41 = 12.2%; this page keeps the report's published 12.11% value.
- 4. The report's savings formula is: energy saved = AI-period total energy * COP uplift / (1 - COP uplift).
- Baseline period
- Jun 1 to Jul 7 autonomous-control operation; 27 valid daily records after missing-power-data exclusions.
- Baseline averages
- COP 4.75, outdoor dry-bulb 22.66C, outdoor wet-bulb 19.36C.
- AI period
- Sep 1 to Oct 7 AI control; 37 daily records.
- AI averages
- COP 5.41, outdoor dry-bulb 25.64C, outdoor wet-bulb 22.43C.
- AI-period minimum COP
- 4.84, from the daily appendix table.
Redrawn Figure 7
High-efficiency cooling-plant benchmark table
The source report compares the AI-period average COP of 5.41 with this benchmark table and states that the result has exceeded the level-2 high-efficiency plant threshold and is close to level 1.
| Climate | Level 3 | Level 2 | Level 1 |
|---|---|---|---|
| Severe cold | >=4.2 | >=4.6 | >=5.3 |
| Cold | >=4.2 | >=4.6 | >=5.3 |
| Hot summer, cold winter | >=4.6 | >=5.1 | >=5.5 |
| Hot summer, warm winter | >=4.7 | >=5.2 | >=5.5 |
| Temperate | >=4.2 | >=4.7 | >=5.3 |
Appendix data and redrawn Figures 8-9
Daily COP, dry-bulb, and wet-bulb records used for the scatter plots
All daily points extracted by MarkItDown from the report appendix are preserved below, and the two COP-weather scatter plots are redrawn from these points.
COP vs wet-bulb temperature
Autonomous controlAI control
COP vs outdoor dry-bulb temperature
Autonomous controlAI control
| Mode | Date | COP | Dry-bulb C | Wet-bulb C |
|---|---|---|---|---|
| Autonomous control | Jun 1 | 4.95 | 21.27 | 18.27 |
| Autonomous control | Jun 2 | 4.79 | 22.46 | 17.78 |
| Autonomous control | Jun 3 | 5.22 | 20.36 | 15.11 |
| Autonomous control | Jun 8 | 5.12 | 21.29 | 17.67 |
| Autonomous control | Jun 9 | 5.00 | 23.23 | 18.43 |
| Autonomous control | Jun 10 | 4.95 | 23.06 | 17.79 |
| Autonomous control | Jun 11 | 5.09 | 21.24 | 17.54 |
| Autonomous control | Jun 14 | 5.07 | 20.80 | 17.49 |
| Autonomous control | Jun 15 | 5.11 | 21.57 | 16.96 |
| Autonomous control | Jun 16 | 5.14 | 20.71 | 16.87 |
| Autonomous control | Jun 17 | 5.07 | 18.25 | 17.24 |
| Autonomous control | Jun 18 | 4.85 | 19.30 | 18.77 |
| Autonomous control | Jun 19 | 4.61 | 20.52 | 19.80 |
| Autonomous control | Jun 20 | 4.57 | 22.31 | 19.73 |
| Autonomous control | Jun 21 | 4.71 | 23.58 | 18.96 |
| Autonomous control | Jun 22 | 4.87 | 22.69 | 17.77 |
| Autonomous control | Jun 23 | 5.06 | 20.24 | 17.18 |
| Autonomous control | Jun 24 | 4.94 | 18.11 | 17.63 |
| Autonomous control | Jun 25 | 4.64 | 21.17 | 19.72 |
| Autonomous control | Jun 26 | 4.60 | 21.62 | 19.60 |
| Autonomous control | Jul 1 | 4.45 | 23.86 | 20.88 |
| Autonomous control | Jul 2 | 4.30 | 24.27 | 21.66 |
| Autonomous control | Jul 3 | 4.28 | 24.67 | 21.84 |
| Autonomous control | Jul 4 | 4.31 | 26.12 | 21.82 |
| Autonomous control | Jul 5 | 4.34 | 25.11 | 21.78 |
| Autonomous control | Jul 6 | 4.14 | 31.39 | 27.41 |
| Autonomous control | Jul 7 | 4.14 | 32.54 | 26.95 |
| AI control | Sep 1 | 5.48 | 26.10 | 21.77 |
| AI control | Sep 2 | 5.55 | 27.06 | 22.70 |
| AI control | Sep 3 | 5.49 | 26.96 | 23.74 |
| AI control | Sep 4 | 5.27 | 28.69 | 25.27 |
| AI control | Sep 5 | 5.03 | 28.18 | 25.15 |
| AI control | Sep 6 | 5.22 | 28.71 | 24.63 |
| AI control | Sep 7 | 5.48 | 28.15 | 24.29 |
| AI control | Sep 8 | 5.59 | 28.17 | 23.43 |
| AI control | Sep 9 | 5.62 | 27.57 | 22.85 |
| AI control | Sep 10 | 5.67 | 27.78 | 22.56 |
| AI control | Sep 11 | 5.52 | 28.80 | 24.47 |
| AI control | Sep 12 | 5.28 | 27.48 | 24.90 |
| AI control | Sep 13 | 5.34 | 25.50 | 23.95 |
| AI control | Sep 14 | 5.55 | 24.65 | 23.29 |
| AI control | Sep 15 | 5.56 | 24.42 | 23.51 |
| AI control | Sep 16 | 5.44 | 25.30 | 23.86 |
| AI control | Sep 17 | 5.44 | 27.73 | 23.87 |
| AI control | Sep 18 | 5.11 | 29.24 | 25.74 |
| AI control | Sep 19 | 4.88 | 30.19 | 26.60 |
| AI control | Sep 20 | 5.13 | 26.86 | 24.41 |
| AI control | Sep 21 | 5.54 | 23.20 | 20.68 |
| AI control | Sep 22 | 5.76 | 21.04 | 19.38 |
| AI control | Sep 23 | 5.71 | 20.42 | 18.89 |
| AI control | Sep 24 | 5.53 | 23.11 | 22.05 |
| AI control | Sep 25 | 5.38 | 26.63 | 23.95 |
| AI control | Sep 26 | 5.30 | 25.63 | 23.42 |
| AI control | Sep 27 | 4.91 | 27.62 | 24.71 |
| AI control | Sep 28 | 4.88 | 25.80 | 24.58 |
| AI control | Sep 29 | 4.84 | 27.40 | 24.55 |
| AI control | Sep 30 | 4.94 | 24.15 | 22.09 |
| AI control | Oct 1 | 5.23 | 23.62 | 19.32 |
| AI control | Oct 2 | 5.98 | 22.95 | 17.60 |
| AI control | Oct 3 | 5.68 | 22.63 | 18.51 |
| AI control | Oct 4 | 5.33 | 23.54 | 18.88 |
| AI control | Oct 5 | 5.81 | 22.26 | 16.79 |
| AI control | Oct 6 | 5.83 | 21.45 | 16.93 |
| AI control | Oct 7 | 5.81 | 19.55 | 16.60 |
Run the same measurement discipline on your cooling plant.
We can review BAS points, operating history, equipment sizes, and metering coverage to determine whether a similar AI supervisory layer is practical.