Sanitation networks under tension: how to limit overflows using artificial intelligence

Urban sanitation networks are now facing double pressure. On the one hand, intense meteorological episodes (intense rains, flash floods) are multiplying under the effect of climate change. On the other hand, existing infrastructures are often aging, undersized and energy-intensive, making their operation expensive and difficult to adapt.

This combination of factors weakens the resilience of sanitation systems: wastewater overflows, pollution of natural environments, non-compliance, increased operating costs... The consequences are serious, both in terms of health and financial.

How to improve the efficiency of sanitation networks in this context of instability?

Spoiler: the dynamic management of wastewater networks provides an effective response, capable of optimizing pumping operations and reducing the risk of overflow.

‍

Sanitation networks: a fragile balance between capacity, flow and performance

Sanitation networks have to deal with increasingly fluctuating volumes of wastewater. Intense precipitation quickly saturates storage and treatment capacities, threatening the balance of the network:

• Aging infrastructures : many networks have been designed for climatic and demographic contexts that are no longer current. This aging also leads to overconsumption of energy.
• Undersized networks : urban growth and weather changes make storage capacities limited.
• Regulatory compliance : the standards for discharges into the natural environment are being strengthened, forcing operators to reduce their environmental impact.

These networks represent significant energy consumption for communities. Pumping raw wastewater can account for up to 18% of the total energy consumption of wastewater collection and treatment systems, according to the comparative analysis by Soares et al. (2017) published in International Journal of Architecture, Arts and Applications.

‍

At the same time, non-optimized management pumping cycles lead to overconsumption and high operating costs. Reducing this energy consumption while maintaining efficient network management is a strategic challenge for operators.

‍

Today, the management of sanitation infrastructures is mainly based on a reactive approach:

• Operators often intervene after the fact, when an incident occurs (overflow, overloading of treatment plants).
• The lack of real-time decision support tools makes it impossible to anticipate load variations and optimize pumping operations.
• Because it is difficult to manage a network as a whole, analysis and actions are most often done position by position, without a global understanding of the network.

The transition to proactive and dynamic management is the solution to improve the resilience of networks and ensure their proper functioning, without going through an expensive investment plan to renew hardware.

‍

Dynamic management of wastewater networks to meet energy and climate challenges

The use of a data-driven approach combined with predictive control of pump stations makes it possible to reduce energy consumption, while reducing overflows during rainy events.

Concretely, the operations on the sanitation network are dynamically adapted to meteorological events and variations in energy prices. To achieve these results, Purecontrol's solution for dynamic management of wastewater networks activates three main optimization levers:

‍

1/ Anticipate rainy periods to reduce overflows in the natural environment

Traditionally, the pumping cycles in a lift station are regulated on the basis of predetermined tidal thresholds. Conversely, Purecontrol uses dynamic regulation: the levels of triggering the pumping vary according to storage capacity and weather conditions.

To reduce overflows, the water level in the pump stations and buffer basins is adjusted according to the expected volumes in the collection basin. These volumes are predicted using mathematical learning models based on user habits, groundwater levels and the history of rainy events.

When a significant rain event approaches, the network is completely emptied. During the rainy event, the pumping configuration of each station is adapted according to the situation to delay or cancel overflows.

‍

‍

AI algorithms make it possible to go beyond monitoring station by station: the entire sanitation network can be analyzed and managed holistically. For example, if a downstream station that is already saturated is at risk of overflow, the control takes into account the situation by stopping the pumping on the upstream station so that it can take over the storage. This strategy makes it possible to reduce the risk of overflows, by 30 to 70%.

‍

‍

2/ Optimize pumping cycles to reduce energy consumption and improve network performance

Outside of periods of bad weather, management focuses on energy performance to reduce costs in kWh and euros. The solution finds the optimal operating point for each pump to improve its specific energy* Performance differences between pumps are also highlighted automatically, without an electrical sub-meter, and the best pumps are thus prioritized.

Pump cycles are also adjusted to maximize consumption during the most advantageous pricing periods (electrical off-peak hours).

‍

Prioritizing the most efficient pump or combinations of pumps, exploiting structures at optimum operating points and adapting pump cycles to fluctuations in energy prices allows an average saving of 10% on average on operating costs.

* The specific pumping energy corresponds to the energy required to pump 1 m3 of water, expressed in kWh/m3

‍

3/ Better supervision to better exploit sanitation networks

Another key performance driver is based on improving supervision for better responsiveness in the field. To facilitate the daily work of operators, Purecontrol provides them with a complete and intuitive hypervision web platform: interactive mapping, performance indicators, customizable dashboards... at the level of the network, collection basins or even station by station.

This hypervision makes it possible to identify equipment performance drifts. For example, overconsumption of energy can be detected more quickly thanks to the solution's prediction tools. The aim is to Detect signals that may indicate an upcoming malfunction. This makes it possible to target interventions, avoid critical breakdowns and extend the life of equipment.

Technical teams save considerable time inIdentifying priorities, reduce unnecessary travel and facilitate their diagnosis. In summary: better exploitation of data for better exploitation of networks.

‍

‍

Towards a proactive, sober and resilient management of sanitation networks

The observation is shared by the entire sector: sanitation networks are increasingly being put to the test. And yet, many decisions are still taken “blindly” or urgently, for lack of appropriate tools to anticipate, manage and prioritize.

With the contribution of real time, the crossing of data (weather, energy, equipment performance) and intelligent management, it is finally possible to take back control. Not by adding a layer of complexity, but by making systems more readable, more predictable — and therefore more manageable.

This change in operation is already showing concrete results for operators who have adopted the solution: less overflows, less energy waste, fewer unnecessary alerts... and above all, more serenity in daily management.

Are you looking for a solution to help you operate a wastewater collection network with peace of mind? So check out this fact sheet which details our approach to the dynamic management of sanitation networks: Download the product sheet

‍