Energy Optimization of Drinking Water Networks: When Artificial Intelligence Revolutionizes Pumping

Environmental challenges, growing need to control energy costs. Faced with these challenges, the optimization of drinking water distribution networks is becoming an imperative.
To achieve this, solutions are emerging, offering innovative approaches to reduce the energy cost of pumping drinking water, and ensure the continuity of service. Let's dive into the details of this technological revolution.

Energy at the heart of the challenges of drinking water networks

The water sector is facing significant energy challenges. In particular, pumping, which represents on average more than 90% of electrical consumption in a drinking water production plant.

This activity is one of the most energy-intensive in the water cycle, after sanitation wastewater, causing high costs for operators and a significant carbon footprint. This high consumption is often explained by the mismatch between the capacity of the pump systems and the usual demand. In the absence of fine and adaptive management, pump stations are often oversized in order to be able to meet peak demand.

However, these extreme conditions rarely occur, which leads to a waste of energy during periods of low use. In some cases, system sizing is based on rough estimates in the absence of accurate data on actual water demand. In a context where energy efficiency and sustainability have become priorities, it is essential to rethink these processes to reduce the energy footprint and optimize operating costs.

What if the key was to put the user back at the center? To better size and adapt the service to user demand?

How to reduce the energy consumption of pumping drinking water?

To meet these challenges, several strategies can be implemented. Among them, the real-time adjustment of the operation of pumps in the drinking water network. To achieve optimal energy efficiency, it is first necessary to carefully analyze the functioning of the pumping system specific to each territory: pressure, flow, reservoir levels and any other critical parameter must be monitored and analyzed continuously, before diagnosing optimization options.

Purecontrol has precisely developed a solution to analyze the distribution of drinking water and to predict water demand, for up to 7 days, using artificial intelligence. The solution then sends control commands to constantly adjust the pumping and storage in a manner adapted to demand. The service is sized and adapted to real demand, allowing In fine to reduce energy consumption.

‍

‍

Concretely, the Purecontrol solution will model the shape of a drinking water reservoir (height/volume relationship) using only historical flows and reservoir level variations. This knowledge of the available volume, combined with a prediction of water demand, thus makes it possible to know in real time the autonomy of drinking water storages. The solution thus implements optimized management of pump cycles, while ensuring control of the risk of disrupting service continuity.

Thanks to the prediction of water demand and battery life, the solution will automatically plan pump cycles to maximize operation during off-peak hours.

The technology developed also makes it possible to know the efficiency of the various pumps in real time and without electrical sub-metering, and thus to identify the most efficient pumping scenarios. Which will then be automatically prioritized in order to minimize energy consumption.

‍

A global approach that allows the simultaneous achievement of several objectives (energy, cost, carbon impact, service security)

The solution is also capable of simulating different scenarios to predict the impacts of future incidents or consumption peaks. This gives operators leeway to adapt management strategies quickly, thus ensuring increased infrastructure resilience.

The advantage of this solution that uses Artificial Intelligence is that it can take into consideration different constraints and objectives, it analyzes continuously to be “self-learning”. Thus the prioritization of pumps is automatic, and can therefore evolve automatically according to the maintenance carried out on the pumps.

This new mode of regulation by AI opens the door to very dynamic functioning, for example in connection with hourly electricity pricing, taking into account the carbon intensity of grid electricity, or to maximize the self-consumption of solar electricity produced on site.

‍

‍

This global approach not only allows significantly reduce energy costs (up to 30% in some cases), but also to extend the life of equipment. In fact, by avoiding excessive on/off cycles, pump maintenance is reduced, which results in reduced maintenance costs and improved service reliability. By relying on data analysis to anticipate equipment faults or detect anomalies such as leaks, the solution guarantees essential service continuity for communities.

A solution that has proven its worth, especially at Eau de Valence Romans Agglo, which was able to reduce its energy bill by €1,500 per week, achieving a nearly 15% reduction in CO2, thanks to the intelligent management of the pumping systems.

‍

To find out more, discover the testimony of Eau du Bassin Rennais during a conference at the Carrefour des Gestion de l'Eau: Discover the testimony

‍

Purecontrol is revolutionizing the management of drinking water distribution networks thanks to its solutions based on artificial intelligence. By optimizing pumping systems, reducing energy consumption and making infrastructure more reliable, we enable communities to meet the environmental and economic challenges of the sector. Our solutions not only ensure significant savings, but also contribute to more sustainable and resilient management of water resources.

‍

‍