Water scarcity is becoming a pressing issue worldwide. Italy, nestled in the heart of the Mediterranean, serves as a poignant example. In 2022, the country experienced a severe drought, resulting in a staggering loss of 36 billion cubic meters of water. Such loss not only threatens the environment but also jeopardizes our economy, given that water scarcity can impact the country’s GDP by as much as 18%. While global initiatives like COP28 and the recent meeting of G7 Ministers of Energy, Climate, and Environment have shown progress, addressing this crisis requires urgent action beyond just fighting climate change, and the private sector will play a pivotal role.
Underground aquifers as storage reservoirs
Technological advancements offer a promising path forward in water management. By developing innovative solutions, we can responsibly manage water resources, storing excess water during times of abundance and efficiently distributing it during periods of scarcity.
For instance, underground aquifers serve as natural reservoirs. With the change in rainfall patterns – heavier rainfall after long drought periods – water struggles to infiltrate the subsoil to reach the aquifers and quickly flows downstream to the sea. A more efficient approach relies on adequate knowledge of surface geography and subsoil geology: if we want to store excess water for drought periods, we cannot ignore the use of underground aquifers as storage reservoirs. These are places where there is no evaporation – being them protected from surface events – and require minimal land use for the preparatory works to inject into the aquifer.
Management of the underground resource entails detailed mapping to identify safeguard areas that could be devoted to groundwater recharge. The knowledge of underground flows enables the identification of valuable aquifers for different uses. The Airborne Electromagnetic Method (AEM) allows to create a 3D model of subsoil resistivity relative to the response of materials to the electromagnetic signal generated in the circuit. The huge amount of data collected – resulting from an electromagnetic survey every 30 metres and up to 400 metres deep on parallel lines generally spaced between 75 and 250 metres –, interpreted with local geological knowledge and possibly integrated with ground measurements, allows for a detailed reconstruction of the subsoil at different depths. This approach provides valuable information to identify protected areas and groundwater recharge areas that can be supplied using water from intense precipitation (which would otherwise flow to the sea). Aquifer recharge is also a tool to counteract saline intrusion in coastal areas, allowing groundwater to be used for drinking purposes without any treatment.
Real-time monitoring of the distribution network
Efficient water management also relies on responsible consumption and minimizing losses in distribution networks. These networks are a complex meshes of interconnected pipes through valves, bends, sections with different diameters and materials. There are multiple pumping systems for water input into the pipes, various types of withdrawal, consumption variations due to tourism, climate variation, or industrial needs. Their management is therefore subject to numerous variables that are not always predictable.
Water leaks can manifest in different ways: if, during a walk, we notice water surfacing, we face a visible leak, which is easily identifiable; apparent leak can instead be attributed to technical problems with measurement devices, for example. Most leakages, however, remain unseen, as water does not surface but disperses into the underlying soil without giving easily identifiable signs of its presence.
This problem can be solved through the optimization of network management, that is, through the creation of water districts and their real-time monitoring. A water district essentially consists of the sectioning and confinement of a limited portion of the water network (around 10-15 km), feeding it from a single point. By monitoring the flow at the entry point, it is possible to have almost immediate evidence of consumption evolution in the area, thus bringing to light any leakages as they occur.
By studying the network and territory’s conformation and delimiting areas with adequate and homogeneous piezometric levels, feed and monitoring points are integrated with regulation and stabilization devices of pressure, as well as water quality control. Stabilizing and lowering the network pressure also implies less stress on the pipes, which are no longer subject to excessive pressure fluctuations and related operational inefficiencies.
Artificial Intelligence for predictive maintenance
Cutting-edge technologies like smart meters and Artificial Intelligence algorithms further enhance water network management, allowing for predictive maintenance and efficient resource allocation. By taking into account incoming flows, smart usage, and new sensor data, we can obtain perfect mass balances and immediate evidence and precise predictions regarding breaks in the water network, facilitating the research and resolution of the leakage.
In addition to that, it is also possible to perform a pre-localization of leaks. We have, for example, launched a project – named “Aquarius” – in the city of Brescia, which involves placing 520 noise loggers on the network, covering approximately 160 km. These sensors “listen” every night to the noise of possible water leakages from the pipes, analysing frequencies, intensity, and other parameters, and overlapping the results between nearby sensors. On average, after a week of listening, the data processing system can automatically close the report or place a possible water leakage on the map. The mapping report identifies, with high precision, the most likely point of water leakage and reduces the manual discovery – which remains necessary – and therefore the resolution time.
As said, the ability of AI algorithms to reproduce the behaviour of non-linear phenomena, combined with the increasing availability of data, represents a great opportunity for water management. Currently, at A2A, we are developing a predictive maintenance system that uses the history of breaks and the characteristics of the infrastructure to prioritize the pipe replacement plan. Simpler regression and correlation algorithms are instead used experimentally for monitoring seasonal flow variations in the districts.
In conclusion, investments in research and technological development are essential for adapting to a future where water resources are increasingly precious. By combining technology, awareness, and targeted investments, we can mitigate the impact of climate change on our lives and ensure sustainable water management for future generations.