Category: Journal Publications

Abstract

Installing pumps as turbines (PaTs) in water distribution networks can recover otherwise wasted energy, as well as reduce leakage caused by high water pressure. However, a barrier to their implementation is the lack of information on their performance in turbine mode. Previous studies have proposed models to predict PaT characteristics based on pump best efficiency points (BEPs), using regressions with one or two dependent variables, or more complex artificial neural networks (ANNs). While ANNs were found to improve the accuracy of predictions, these models are known to be unstable with small datasets. Other types of regressions with multiple variables have not been explored. Furthermore, because only small datasets are available to train these models, multivariate regression methods could yield better results. The present study develops multivariate regression models to predict BEPs and characteristic curves of PaTs. A database of 145 BEPs and 196 characteristic curve PaT experimental records was compiled from previous literature. Twenty-four types of multi-variate regressions, as well as ANN were compared, with dimensioned and dimensionless versions of the datasets. The multivariate regression models consistently outperformed previous models, including ANN. The R² of the head and efficiency curves were 0.997 and 0.909, respectively. Results also showed that XGB regressors and a dimensionless dataset yielded the best-fit models overall. The high accuracy of the models, combined with their lower computational cost compared to ANN, make them a robust solution for selecting PaTs in practice.

Brisbois and Dziedzic, 2023

Water https://doi.org/10.3390/w15183290

Abstract

Leaks cause substantial economic and water losses for water distribution networks (WDNs). Thus, it is imperative to detect and localize leaks. This paper proposes an efficient set of methods for detecting and localizing leaks, with three main steps: model calibration, leak detection, and leak localization. First, demands, pipe diameter, and pipe roughness in the hydraulic model are calibrated with a genetic algorithm (GA). The X-bar method and cumulative sum control chart are then implemented to detect pipe bursts and incipient leaks, respectively. Lastly, a reduced search space is defined and searched with a GA to locate leaks. The proposed strategy was tested on the hypothetical WDN of L-Town from the Battle of the Leakage Detection and Isolation Methods. Results show that 12 of 19 leaks were detected, and 8 of the 12 leakages were accurately located. Compared with other approaches, the proposed approach is more efficient and equally effective.

Cai et al. 2022

Canadian Journal of Civil Engineering https://doi.org/10.1139/cjce-2023-0176

Abstract

The present study applies a business model framework to a water utility using a case study. A survey was developed and conducted with residential water customers in the City of Guelph, ON, Canada, with the objective of assessing their expectations in order to improve the business. Preferences differed by user type, underlining the benefit of distinguishing between customer segments (or user types) and customizing messaging. Users were not aware of all services covered by their water bill, and although expecting high quality services they did not support rate increases. Stakeholder collaborations could be used to improve awareness, offer decentralized solutions, and resolve issues collectively.

Dziedzic and Karney, 2016

Utilities Policy https://doi.org/10.1016/j.jup.2016.06.002

Abstract

A gradient-based network assessment is proposed as a way to understand and improve the cost and hydraulic performance of distribution networks. It seeks to reduce the computation time of optimizing water distribution networks with varying demands by using a sequence of shorter time cycles to approximate a fuller range of costs. To allow system evolution, diameters are adjusted after each cycle according to anticipated capital, energy, and damage costs. Accounting for risks of low- and high-pressure hydraulic failure not only adds needed redundancy to the network, but also helps guide the gradient search. The method was applied to the well-known network of “Anytown.” Results indicate that the technique can be effectively applied to different scenarios and generate robust solutions. Furthermore, its lower computational intensity should facilitate its application as part of a broader optimization process and thus better enable the assessment of more storage, pumping, and control alternatives.

Dziedzic and Karney, 2016

Journal of Water Resources Planning and Management
https://doi.org/10.1061/(ASCE)WR.1943-5452.0000573

Abstract

A survey was developed and conducted with residential water users in the City of Guelph, ON, Canada, with the objective of assessing their awareness, preferences, concerns, motivations, and priorities. The overall goal of this data is to improve the water system on different fronts: infrastructure, conservation programs, communication with users, and long-term strategies. Results highlight the local concerns with water scarcity, currently addressed by conservation programs, as well as water quality, aging infrastructure, and costs. Correlations between user type and answers were seldom found, showing that different residential customer segments share concerns and motivations. Even so, feedback must be sought from all customer segments, residential as well as industrial, commercial, and institutional, through different channels. The findings will allow the utility to identify preferred solutions to current issues and openness to change, as well as gaps in user and utility knowledge.

Dziedzic and Karney, 2015

IWA Water Practice and Technology https://doi.org/10.2166/wpt.2015.094

Abstract

Previous studies have naturally related water distribution network performance to the ability to deliver sufficient pressure and flow. The present paper emphasizes that performance also depends on the efficiency of delivering these requirements. Accordingly, an efficiency-based performance index is proposed. It is the geometric average of four performance metrics: reliability, vulnerability, resilience, and connectivity. These are themselves based on the energy efficiency, hydraulic capacity, and structural ability of the system to deliver water under a range of conditions. The metrics are applied to two example networks and variations of these, enabling the assessment of their relevance, their sensitivity to system changes, and permitting a comparison to existing metrics. Variations represent different redundancy increasing strategies, recognized for improving performance. The proposed performance index generally follows a similar trend as the previous indices, increasing with network pressure. Nevertheless, it varies differently and penalizes networks with unnecessarily high pressures. Because the index is based on energy and demand efficiency metrics, it automatically complies with the energy and mass balances of the network. Moreover, the new metric is easily interpreted and can be applied to various systems, whether complex or involving multiple scenarios.

Dziedzic and Karney, 2015

Journal of Water Resources Planning and Management
https://doi.org/10.1061/(ASCE)WR.1943-5452.0000564

Abstract

Descriptive energy metrics, calculated for each component, represent how the hydraulic state of a network evolves and how energy flows vary temporally and spatially. More specifically, these metrics describe how the energy supplied is partitioned between the energy that is dissipated, lost, and delivered throughout the system. The metrics are meant to support planning, from local (e.g., pump or pipe renewal) to system-wide (e.g., leakage or pressure management) decisions. Whereas aggregate results are indicators of system capacity, efficiency, greenhouse gas emissions, and costs, the comparison of component metrics allows for the identification of specific pipes, tanks, or pumps for which changes would be most beneficial. Furthermore, analysis of the temporal variation of energy flows facilitates the assessment of operation under multiple scenarios. The metrics are applied to a case study of the Toronto water distribution system and show, based on two scenarios provided by Toronto Water, that on average, less than 27% of the energy supplied is actually delivered to users. This system inefficiency has important economic and environmental repercussions. Nevertheless, changes to operations, such as improved pump maintenance or scheduling, have significant potential to lower costs and exploit lower greenhouse gas emission factors.

Dziedzic and Karney, 2015

Journal of Water Resources Planning and Management https://doi.org/10.1061/(ASCE)WR.1943-5452.0000555

Abstract

Water utilities have large amounts of data at their disposal, which are seldom being used to their full potential. Integrating water
billing records with land-use and demographic data organizes information and makes inherent correlations easier to understand, facilitating communication to stakeholders. This data was integrated for three Ontario (Canada) municipalities, Barrie, Guelph, and London. A summary tool was created, with proposed metrics and charts, that facilitates comparisons between cities, definition of benchmarks, and identification of targets for conservation. More than 60% of consumption in these cities is residential, and mostly lies below the Ontario average of 267 L/cap · day. Water user clusters were created through self-organizing maps, K-means, and hierarchical clustering, and selected according to their pseudo-F and Rand statistics. Users within the same or similar property codes were found to cluster together. The application of data-mining methods provides actionable information for utilities seeking to reduce demands and increase system sustainability.

Dziedzic et al. 2014

Journal of Water Resources Planning and Management
https://doi.org/10.1061/(ASCE)WR.1943-5452.0000462