Estimating Facility‐Level Monthly Water Consumption of Commercial, Industrial, Municipal, and Thermoelectric Users in Virginia

Abstract

Statement: Monthly data on facility-scale water consumption suggest that industrial and commercial consumption may be twice as high as previously estimated with significant seasonal variation.ABSTRACT: Understanding water consumption is an important component of water management. However, water consumption data are limited and consumption coefficients do not account for variability through time and across users. This study combines federally maintained discharge data with state-maintained withdrawal data at monthly time steps to estimate facility-level and spatially aggregated water consumption in Virginia between 2010 and 2016. We evaluate (1) the feasibility of using discharge and withdrawal datasets to estimate sub-annual water consumption, (2) how these consumption estimates vary depending on the level of spatial aggregation, and ( 3) what patterns of seasonality exist in consumption estimates. We find that a combined process of text matching and geospatial analysis is effective in matching facilities and yielding monthly time-series of water consumption. Our results suggest that median consumption in industrial (17%) and commercial (19%) facilities may be higher than median consumption coefficients in the literature (10%). Consumption estimates also demonstrated more variability across facilities and seasons than aggregate coefficients in the literature suggest. Combining this approach with institutional knowledge can assist in quantifying issues such as inter-basin transfers and infiltration that impact consumption estimates, ultimately allowing for more accurate accounts of water use and availability.

Prediction of multi-sectoral longitudinal water withdrawals using hierarchical machine learning models

“…This result mirrors previous discussions of pooled and unpooled regression models, where unpooled models fit tend to be overfit and less generalizable when a small number of observations are available (Gelman & Hill, 2007). High variance in water use observations could also potentially be indicative of previously observed data quality issues in self-reported water use (Chini & Stillwell, 2017;McCarthy et al, 2022;Zhang & Balay, 2014). The ability to reduce the impact of these errors in predictive models is another potential benefit to the ensemble modeling approach.…”

“…An analysis of correspondence between cluster assignment and sector indicated that clusters generally did not corresponded to a single sector. This suggests that there are certain patterns of water use behavior that cannot be explained simply by sectoral classifications, consistent with previous research (Attaallah, 2018;McCarthy et al, 2022). Thus, models were also fit at the large (k = 3) and small (k = 8) cluster levels, where data from all facilities within a single cluster were combined into a single model.…”

“…One limitation with sectoral grouping is that facilities in a single sector might actually exhibit very different pattern of water use (Attaallah, 2018;McCarthy et al, 2022). Thus, the small and large cluster grouping levels were determined based on the results of a hierarchical cluster analysis (Everitt et al, 2011) To determine the optimal number of clusters, facilities were divided into k ∈ {1, 15} hierarchical clusters based on Euclidian distance.…”

Identifying Opportunities for Nonpotable Water Reuse Based on Potential Supplies and Demands in the United States

“…These data are ultimately reported on a HUC-12 scale. Other studies have recognized the challenge of finding and using facility-specific water withdrawal and discharge data. , Additionally, estimates show that only 75% of industrial water withdrawal in the United States is self-supplied, so the demand values in this work are conservative and do not consider industrial water demands that are currently met by municipal sources . This analysis could also be enriched with on-site industrial reuse data.…”

“…Self-supplied industrial demand is larger than power and irrigation demand in 70% of the watersheds where both supply and demand exist (Figure S10). These findings highlight the need for not only greater exploration of municipal wastewater reuse for industrial facilities but also greater site-specific data transparency in industrial water use. , A subset of industrial water use is the application of reclaimed water for cooling power-generating facilities. One strong motivator of water reuse in power generation is that inconsistencies in water supply and ambient water temperature can have severe negative effects on the reliability of electricity supplied to the grid, which makes a consistent stream of cooling water increasingly valuable. , An important policy consideration is that the introduction of an additional water source for power plants may have unknown effects in lengthening the life span of inefficient power plants that are nearing retirement.…”