Kirk Bonney

Amidst increasing need for efficiency gains and protection against vulnerabilities, the expression of infrastructure as digital twins is increasingly important to assess disruptions and planning. Water distribution networks (WDN) are one such infrastructure facing an increasing need for modernization and resilience. However, the dearth of high-resolution data on WDNs prevents robust simulations. This study explores the challenges and opportunities of synthetic water distribution models as a critical bridge to navigate the complexities of urban water networks. A synthetic infrastructure model is applied to San Juan and Mayagüez, Puerto Rico, to generate synthetic WDNs composed of reservoirs, pumps, links, and nodes, using the road network as a proxy for pipe location. The synthetic networks are then compared with true data from the Puerto Rico Aqueduct and Sewer Authority to assess their ability to reproduce key structural and hydraulic characteristics. Results show that the synthetic networks effectively capture large-scale network connectivity and spatial patterns, with moderate to strong correlations in pipe placement, nodal allocation, and pressure profiles. Challenges are apparent in synthesizing across urban and rural morphologies, with the synthetic models overestimating nodal density in rural areas while underestimating it in urban cores. Pipe diameter distributions also diverge, with synthetic models favoring smaller pipes, impacting redundancy and criticality assessments. Opportunities to improve synthetic models include integrating population-density based allocation strategies to influence looped versus branched structure, improved handling of reservoir connections, and refining hydraulic parameters to enhance realism. These findings demonstrate the potential of synthetic water distribution models as scalable tools for resilience analysis in data-scarce regions while highlighting strategies to improve the representation of sociotechnical complexities and hydraulic dynamics.

As concerns about water scarcity grow due to increasing aridity and extreme weather, there is a need for computational test beds to help explore water management challenges and elucidate interdependences within coupled human-natural systems. Prior approaches for water management analyses have ranged from process-based simulations to data-driven modeling. However, development of data-driven models for water availability can be challenging due to poor data availability and software limitations. To overcome these challenges, this work aims to develop a parsimonious, deep learning, long short-term memory (LSTM)-based surrogate of a process-based water management model, the Water Rights Analysis Package (WRAP). Synthetic streamflows were generated to create an ensemble of scenarios across a range of hydrologic conditions. These data were inputted into WRAP to obtain associated water allocation data for LSTM training, testing, and validation. We applied this computational framework to a case study of the Colorado River Basin in Texas. The trained LSTM emulates the water allocation process with low overall error and demonstrates minimal performance impact when out-of-distribution drought scenarios are used. Exploratory analysis of error patterns shows that the LSTM effectively captures the spatio-temporal water allocation patterns well and has no significant bias for underlying system attributes. These findings suggest that LSTMs have the potential to serve as surrogates for water management models. The computational framework developed could be used to explore water management scenarios for other basins to understand potential impacts from increasingly extreme hydrologic conditions.

Water distribution network (WDN) models are an essential tool used by water utilities for hydraulic analysis. Unfortunately, missing data and insufficient resources often make creating and maintaining these models unfeasible. Existing methods to address missing pipe properties, like sequential imputation for missing values and reconstruction using graph metrics, are designed to accommodate random patterns of missing information and require a significant percentage of the system's attributes to be known. However, these data completeness assumptions do not always align with real‐world scenarios where large sections of the WDN model have missing data. To address this challenge, this study proposes a data‐driven approach for estimating pipe diameter when considering different spatial patterns and degrees of data completeness (i.e., 0%–90%). Using data from 16 WDNs in Kentucky, this study compares the use of machine learning (ML) using topological and geospatial features against an existing deterministic approach. Results demonstrate that WDN models with pipe diameters predicted by the proposed ML method had comparable hydraulic performance to the ground truth models. Moreover, results showed that ML method performance varies between WDNs of differing topological classification. Insights from this study help advance the ability to leverage partial data to create and maintain WDN models amid uncertainty and inadequate resources.

The increasing potential of infrastructure disruptions calls for high-quality infrastructure models to be used in resilience analysis and decision making. Unfortunately, many utilities and communities do not have access to accurate and detailed models due to a lack of data and resources. Furthermore, security restrictions on sharing infrastructure models present roadblocks to research, analysis, and decision making. Recent advances in the development of synthetic water distribution models provide a potential solution to this problem. There is an opportunity to improve these methods by leveraging incomplete pipe datasets to aid synthetic network generation. To address this gap, we developed a methodology for synthetic network generation that incorporates partial pipe data using a modification of the minimum cost flow algorithm for network generation and pipe sizing. This methodology demonstrates how partial pipe data can be leveraged to improve site-specific synthetic network generation. For the study area of Mayagüez, Puerto Rico, a synthetic model generated using 50% of real pipe data matches the pressure of the validation system with an average error of 23.5 m of head, which improves upon the average error of 31.6 m of head produced by a synthetic model generated using no data of the real pipes. Additionally, synthetic networks are shown to replicate the pressure response under a disruption scenario of the validation network, suggesting potential use in resilience analysis.

The purpose of pvOps is to support empirical evaluations of data collected in the field related to the operations and maintenance (O&M) of photovoltaic (PV) power plants. pvOps presently contains modules that address the diversity of field data, including text-based maintenance logs, current-voltage (IV) curves, and timeseries of production information. The package functions leverage machine learning, visualization, and other techniques to enable cleaning, processing, and fusion of these datasets. These capabilities are intended to facilitate easier evaluation of field patterns and extraction of relevant insights to support reliability-related decision-making for PV sites. The open-source code, examples, and instructions for installing the package through PyPI can be accessed through the GitHub repository.