The Challenge

In this case, the challenge is to optimize the operation of 8 large water pumps at a Sewage Treatment Works facility to reduce costs. The key issue is that during unusual weather conditions like typhoons or heavy rain, the operators tend to turn on multiple pumps at high speeds based on experience to avoid flooding, rather than using a methodical approach. As a result, the annual energy cost can reach up to HKD 100,000,000 without proper optimization.

Our Solutions

To address this challenge, Carnot has leveraged your AI capabilities to develop a smart control system that can dynamically optimize the pump operations based on historical data, real-time data, and weather forecasts:

Predictive Modeling

Using multiple prediction algorithms, we have built models that can accurately forecast water inflow volumes based on weather patterns, historical data, and other relevant factors. This allows the system to proactively adjust pump operations to meet demand while minimizing energy consumption.

Automated Pump Scheduling and Speed Control

The AI system can automatically schedule the pumps to run at optimal times and combinations, taking into account factors such as energy costs, water levels, and weather conditions. This ensures the most efficient use of the pumps without compromising the facility's ability to handle heavy rainfall or flooding.

Adaptive Control

The AI system continuously monitors the performance of the pumps and makes adjustments to the control algorithms to adapt to changing conditions, such as equipment degradation or fluctuations in water inflow. This helps maintain optimal energy efficiency over time.

Results

The proof of concept project was completed in Q1 2024, and the results show an annual energy cost savings of 5%.