Assessing the Environmental and Economic Footprint of Leakages in Water Distribution Networks ^†

Abstract

All urban and agricultural water distribution networks (WDNs), irrespective of their physical and operational characteristics, encounter substantial leakages which result in significant water losses, environmental degradation through increased carbon emissions, and noteworthy economic burdens. The current work aims to quantify both the environmental impact, estimated in terms of CO₂ emissions, and the economic implications associated with leakages and evaluate the effect of the most widely used leakage reduction strategies. The current approach is applied to the water distribution network of the city of Patras in Western Greece.

1. Introduction

Both urban and agricultural water distribution systems are inherently susceptible to substantial leakages [1,2,3,4,5,6,7], resulting in significant inefficiencies and environmental concerns. Leakages pose significant challenges, as they intensify the already concerning water scarcity issues while increasing the associated operational expenditures and contributing to environmental degradation [8,9,10,11,12,13].

To mitigate the aforementioned challenges, various leakage reduction methods and techniques have been developed throughout the years, encompassing pressure regulation, pipeline rehabilitation, network surveillance, in situ pipe restoration, and comprehensive water balance assessments. Their effectiveness and feasibility considerations inherently depend on specific structural and operational water network characteristics.

The most widely implemented technique is network partitioning into smaller, hydraulically isolated areas (district metered areas; DMAs) while also calibrating the operational pressure (pressure management areas; PMAs) for a resilient reduction of water losses [14,15,16,17,18,19]. Network partitioning represents a cost-effective and economically viable leakage mitigation strategy, owing to its relatively small initial investment, compared to comprehensive network rehabilitation, since it capitalizes on the initial network infrastructure, rendering it an efficient and pragmatic solution for enhanced operational performance and prolonged network lifespan [14,15].

The current study aims to quantify the environmental and economic impacts of WDN leakages, specifically focusing on CO₂ emissions and water production costs, while also testing the effectiveness of a recently introduced statistical clustering methodology [14,20] by using the water supply system of the city of Patras (in Western Greece) as a case study.

2. Data and Area of Application

To estimate the total CO₂ emissions per kWh of production, we utilize the energy consumption and CO₂ emission data acquired from the Greek Public Power Corporation [21,22] and the Independent Power Transmission Operator [23], respectively, during the 6-month high water consumption period from May 2023 to October 2023 [7]. In order to estimate the water production costs, we use the high-resolution (i.e., 1-min) energy consumption and flow data during the same period (i.e., May 2023–October 2023), as well as the associated energy billing data, from three pumping stations, namely: (a) Karnavalika, (b) Glafkos 1, and (c) Glafkos 2 (see cyan pins in Figure 1) from the water distribution network (WDN) of the city of Patras.

The WDN consists of approximately 1000 km of pipeline grid and covers the entire metropolitan area of the city of Patras in Western Greece, serving more than 200,000 consumers, according to reports by the associated public authorities. Recent studies [24,25,26,27,28] report that although the network is divided into 86 district metered areas (DMAs, see Figure 1), it exhibits significant leakage rates (i.e., more than 40% of the total input volume), leading to an excessive loss of water and increased energy consumption [25,27].

3. Methods

3.1. Monthly Energy Consumption Estimation

The available dataset for the three pumping stations (see Figure 1) includes the following: (a) pumped flow and (b) power consumption. Upon analysis, certain values were identified as outliers, likely resulting from either random measurement errors or electromagnetic interference occurring during data transmission from the meter to the central database. Since the presence of such unrealistic values significantly impacts the estimation of the monthly energy consumption, these values were removed from the time series of the recorded measurements and were subsequently replaced by the mean of the preceding and succeeding observations.

3.2. CO₂ Emmisions per Kilowatt-Hour (kWh) of Electricity Production

The carbon footprint per kilowatt-hour (kWh) of production (expressed in grams of CO₂ per kWh, see [29,30,31]) over the aforementioned six-month period (i.e., May 2023–October 2023, see Section 2) was estimated using data acquired from the Greek Public Power Corporation and the Independent Power Transmission Operator, as follows:

$$\mathrm{C}\mathrm{a}\mathrm{r}\mathrm{b}\mathrm{o}\mathrm{n} \mathrm{F}\mathrm{o}\mathrm{o}\mathrm{t}\mathrm{p}\mathrm{r}\mathrm{i}\mathrm{n}\mathrm{t} \left({\displaystyle \frac{{\mathrm{g}\mathrm{C}\mathrm{O}}_2}{\mathrm{k}\mathrm{W}\mathrm{h}}}\right) = {\displaystyle \frac{\mathrm{T}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l} {\mathrm{C}\mathrm{O}}_2 \mathrm{E}\mathrm{m}\mathrm{i}\mathrm{s}\mathrm{s}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{s} ({\mathrm{g}\mathrm{C}\mathrm{O}}_2)}{\mathrm{T}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l} \mathrm{E}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{g}\mathrm{y} \mathrm{P}\mathrm{r}\mathrm{o}\mathrm{d}\mathrm{u}\mathrm{c}\mathrm{e}\mathrm{d} (\mathrm{k}\mathrm{W}\mathrm{h})}}$$

(1)

where the Total CO₂ Emissions represent the cumulative emissions from all power sources (i.e., coal, oil, hydropower, wind, solar, biomass), calculated in grams of CO₂, while the Total Energy Produced is the sum of the electricity generated over this period in kilowatt-hours from all power sources [32,33,34,35,36,37].

3.3. Hierarchical Clustering of WDNs

To minimize leakages and the associated environmental and economic costs, we apply the Serafeim et al., 2022 [14], approach for a resilient reduction of leakages, combining statistical clustering with hydraulic modeling. The current methodology aims to minimize water leakages while ensuring adequate hydraulic resilience within the network.

This is achieved through a hierarchical clustering approach, enhanced with topological proximity constraints to preserve the network’s structural integrity. Key strengths of this method include the following: (1) Use of the original pipeline grid as the connectivity matrix, which prevents unrealistic clustering outcomes; (2) a statistically rigorous and objective foundation relying solely on statistical metrics, thus avoiding user biases; and (3) efficiency and ease of implementation, requiring minimal computational resources and processing power. A brief description of the implemented procedure is presented below.

• Water Balance and Leakage Estimation

Leakage rates are estimated using the probabilistic Minimum Night Flow (MNF) method by Serafeim et al. (2021) [25] to accurately quantify water losses and establish water balance equilibrium within the network. The principal advantage of the current approach is that it allows for accurate water loss quantification, encompassing both active bursts and leakages, based exclusively on flow data from the network’s inlet, without requiring supplementary equipment dedicated to leakage localization.

2.

Hydraulic Model and Demand Distribution

The hydraulic model is designed with high nodal density to capture topographic variability and connectivity [38]. Water demand at each node combines demand- and pressure-driven components, with leakages redistributed based on nodal pressures.

3.

Initial Hydraulic Resilience Assessment

Pressure-driven leakages and nodal pressures are accounted for using Todini’s modified index [14], providing an initial resilience assessment.

4.

Clustering and Solution Selection

A hierarchical clustering approach with topological constraints (see [39]) optimally sizes and allocates pressure management areas (PMAs). The solution balances real loss reduction with hydraulic resilience and can be customized for network-specific needs.

4. Results

Figure 2a illustrates the monthly carbon footprint per kWh of production (in gCO₂/kWh) during the 6-month period from May 2023 to October 2023 estimated using Equation (1). The carbon footprint (gCO₂/kWh) varies significantly over the 6-month period due to factors primarily tied to the energy generation methods and consumption patterns across different seasons and weather conditions [40].

Figure 2b,c illustrate the pumped volume of water and the associated energy consumption, respectively, using the method described in Section 3.1. While the aggregated volume of pumped water reaches its seasonal maximum in August, each of the three pumps demonstrates a distinct operational pattern dictated by the hydrogeological characteristics of its respective aquifer.

Table 1. Comparative analysis of the initial condition (i.e., prior to clustering) and the final state (i.e., post-clustering) of the 3 pumping stations regarding the leakage rates and the associated environmental and financial cost (in tCO₂ and EUR, respectively).

Pumping Station	Prior to Clustering			Post-Clustering
	Lost Water (m3)	tCO2	EUR	Lost Water (m3)	tCO2	EUR
Karnavalika	140,723	26.24	63,325	70,362	13.12	31,662
Glafkos 1	55,550	8.21	24,998	38,885	5.74	17,498
Glafkos 2	102,940	14.98	46,322	61,764	8.99	27,793

provides a comparative analysis of the initial condition (i.e., prior to clustering) and the final state (i.e., post-clustering) of the three pumping stations regarding the leakage rates and the associated environmental and financial cost (in tCO₂ and EUR, respectively). More specifically, the leakage rates and the associated CO₂ emissions and economic costs are reduced up to 40%, highlighting the importance of targeted pressure management towards achieving substantial efficiency improvements.

5. Conclusions

Mitigating water leakages in WDNs is crucial for achieving environmental sustainability and economic efficiency. By reducing leakages through network partitioning and pressure management, water utilities can significantly reduce both the carbon emissions and the operational costs, contributing to global sustainability goals, as demonstrated by a case study in the city of Patras.

The Serafeim et al., 2022 [14] hierarchical clustering approach leads to a significant reduction in water loss (more than 40%), leading to enhanced environmental and economic performance by reducing energy wastage.

These findings tend to support clustering as a scalable and economically viable approach to achieve the objectives of optimizing water management practices and achieving sustainability goals. Future studies may investigate refinements in clustering techniques, especially those that incorporate the dynamics of single aquifers and seasonal variability, which would result in improved environmental and economic outputs.