Research Reports

Permanent URI for this collectionhttps://www.lib.ncsu.edu/resolver/1840.20/6

Browse

Now showing 1 - 20 of 569

Potential Impacts of Legacy Lithium Mining on Water Quality in North Carolina
(NC WRRI, 2025-08-31) Vengosh, Avner; Williams, Gordon D.Z.
As the search for new lithium deposits continues and new mines become permitted in North Carolina, understanding the potential water quality impacts of legacy lithium mining is important to enable sustainable lithium extraction. This project aims to evaluate the possible impact of the legacy of lithium mining on the quality of surface water and groundwater in the vicinity of legacy lithium mining and the Carolina Tin-Spodumene Belt (TSB) region of North Carolina. As part of this project, we conducted extensive sampling and analyses of the major and trace elements (50 elements) as well as strontium isotopes (87Sr/86Sr ratios) of 99 stream water and 93 groundwater samples in 6 sampling campaigns in a legacy lithium-pegmatite mining district in North Carolina. Analysis of the quality of surface water downstream from outfalls of two legacy lithium mines (Hallman Beam and Kings Mountain) and an active lithium processing site (Livent) shows marked increases in the overall salinity (reflected by Electrical Conductivity; EC as well as high calcium and sulfate), lithium (>1000 μg/L), rubidium (>25 μg/L), and cesium (>10 μg/L) immediately downstream from the outfalls in comparison to the background levels in local surface water. Further downstream, these concentrations decrease to the background levels. No other potential contaminants were observed and impacted streams were characterized by neutral pH. Elevated levels of lithium and rubidium in the impacted stream waters were associated with higher 87Sr/86Sr (radiogenic), Rb/Sr, and low K/Rb ratios reflecting the natural composition of the lithium pegmatite rocks and from processing facilities, distinct Ca/SO4 (molar) ratios (~1) reflecting gypsum dissolution, likely from gypsum generated during lithium processing. Overall, the concentrations of lithium in streams associated with legacy lithium mining and processing were higher than background streams within the Carolina Tin-Spodumene Belt (TSB) region, which were higher than lithium levels in common streams in the Piedmont region of North Carolina. We observed a high correlation (R2=0.97, p<0.01) between EC and lithium in the stream waters, suggesting a novel and rapid methodology to monitor lithium occurrence in surface waters near lithium mining and processing sites through using real-time EC measurements as a proxy for lithium occurrence in surface waters. Analysis of groundwater show systematically higher lithium, rubidium, and cesium in groundwater underlying legacy lithium mining (Kings Mountain) relative to the regional groundwater from TSB region, which in turn were higher relative to common groundwater in the Piedmont region of North Carolina. However no other contaminants such as arsenic were detected in the groundwater underlying the legacy lithium mine. In contrast, in the northern section of Gaston County, we have identified a cluster of wells with high levels of contaminants, including arsenic (up to 346 μg/L), lithium (890 μg/L), and cesium (46 μg/L). A previous study in the area has also identified a cluster of high arsenic levels and suggested a naturally occurring phenomena of groundwater contamination from geogenic sources. We observed direct correlations between 87Sr/86Sr and lithium as well as Rb/Sr and low K/Rb ratios in the apparent geogenic contaminated groundwater. These correlations suggest that intensification of water-rock interactions with lithium pegmatite rocks that has caused mobilization of lithium into the groundwater. On the other hand, arsenic contamination in groundwater in North Carolina is typically derived from interaction with other rock types such as metamorphic volcanics and mica schist. We posit the that the mix of aquifer rocks in that area that includes mica schist and lithium pegmatite would generate a unique groundwater chemistry in which As is elevated from interaction with the mica schist rocks, while the lithium, 87Sr/86Sr, Rb/Sr, and K/Rb signatures are affected by reactions with the lithium pegmatite rocks. While the apparent geogenic groundwater contains high arsenic levels, the regional TSB groundwater has typically low arsenic concentrations. A parallel study of the quality of groundwater in the vicinity of legacy lithium mining of Kings Mountain show systematically elevated levels of lithium, rubidium, and cesium, and yet, negligible levels of arsenic or other EPA regulated contaminants. Overall, while lithium in groundwater in the TSB region can be derived from both naturally occurring (geogenic) source and the legacy of lithium mining and processing, arsenic appears to exclusively occur in geogenically-sourced groundwater. Therefore, we posit that the ratio of arsenic to lithium in groundwater can be used to detect the possible origin of groundwater and links to legacy lithium mining and processing. The distinction between naturally occurring contamination and water-quality impact from legacy lithium mining is the key component of this research. Our geospatial analysis reveals that the groundwater quality is not dependent on the proximity to the legacy lithium mining (Hallman Beam and Kings Mountain) and the active lithium processing site (Livent) in the TSB region. Therefore, wells located near these lithium mining and processing sites were not affected and the quality of the TSB groundwater is primarily controlled by naturally occurring processing and reactions with aquifer rocks, some containing lithium pegmatites with distinctive geochemical and isotopic fingerprints.
Probabilistic Streambank Erosion Curves for Advancing Assessment And Prediction in the Context of Climate Change
(NC WRRI, 2025-10-31) Castro-Bolinaga, Celso; Hopkins, Kristina; Cabas, Ashly; Sanchez, Karol; Holberg, Sam; Acharya, Shishir
Streambank erosion is a ubiquitous morphodynamic process in riverine environments. However, when pronounced, high rates of streambank erosion are detrimental to the stability and function of streams. Over time, such high rates can lead to excessive channel-bed erosion and deposition, increased loading of fine-grained material and nutrients, and poor water quality and habitat conditions. The overarching goal of this research project was to advance assessment and prediction of streambank erosion in the context of climatic fluctuations in streams’ water-sediment regimes. The project was organized into two phases at answering the following research questions: (1) how streambank erodibility parameters vary in response to rapidly changing hydro-environmental conditions? and (2) how do different geomorphic characteristics and hydrological fluctuations control the probability of exceeding annual erosion rates? To answer the posed research questions, the following specific objectives were pursue: (1) to characterize the temporal variability of streambank erodibility parameters (i.e., τc – critical shear stress, K – erodibility coefficient, and α- numerical exponent greater than zero) under changing physical, hydrological, and hydro-environmental conditions; and (2) to develop a probabilistic streambank erosion curve for quantifying the probability of exceedance of streambank erosion rates under variable conditions. The technical approach integrated two components. The first component consisted of field measurements and laboratory analyses across seven streambanks located along reaches of Crabtree Creek and Walnut Creek in Raleigh, North Carolina. The second component coupled high-resolution hydro-environmental monitoring and process-based probabilistic modeling at a selected streambank in Crabtree Creek. Monthly Jet Erosion Tests (JET) normalized by instantaneous measurements of hydro-environmental conditions were performed to characterize the temporal variability of streambank erodibility parameters. At the selected streambank in Crabtree Creek, data from a soil sensor that recorded high-frequency (5 min.) soil moisture content and temperature for almost a year was combined with streamflow data from nearby USGS gages to construct a site-specific probabilistic streambank erosion curve. Results from this project revealed that: (1) streambank erodibility parameters are dependent on rapidly changing hydro-environmental conditions, showing variability with changing moisture content and soil temperature at the same streambank; (2) the degree of variability of streambank erodibility parameters with changing moisture content and soil temperature is dependent on the streambank composition, with the percent of fine material playing an important role in modulating the magnitude and the direction of such degree of variability; (3) the shape of probabilistic streambank erosion curves provides a direct indication of a streambank susceptibility to experience erosion over given time horizon as dictated by concurrent changes in erodibility parameters and hydro-environmental conditions; and (4) the overall shape of the probabilistic streambank erosion curve is controlled by the probability distributions of hydro-environmental conditions and flow rates, highlighting the importance of the frequency and magnitude of erosion-generating hydrological events.
Microplastic Sources, Abundance, and Transport Dynamics in Water and Sediments of a Representative Appalachian Headwater Basin
(NC WRRI, 2025-10-31) Miller, Jerry; Love, Jason; Gray, Austin; Barrett, Nathaniel; Youker, Robert
Microplastics (plastic particles < 5 mm in size) represent a contaminant of emerging concern on a global scale. The primary objectives of this 2-year study were to (1) further our understanding of the short-term spatiotemporal variations in the concentrations of MPs in both water and sediments of Richland Creek and similar headwater basins in the southern Appalachians, (2) investigate the abundance of MPs in caddisfly (Hydropsychidae) larvae and cases as well as the acute toxicity of MPs to caddisfly larva using laboratory exposure experiments, and (3) assess the source(s) of MPs to the streams, including atmospheric inputs. MP concentrations and characteristics were determined, in part, by analyzing 458 water samples collected primarily from 8 sites within the Richland Creek Basin and an additional 156 samples from two sites within the Cullasaja Watershed over a range of flow conditions including baseflows, stormflows, and the Hurricane Helene event. Mean MP concentrations in water were unexpectedly high for predominantly forested, lightly developed headwater basins, falling within the upper quartile of globally reported values (ranging from 0 to 70.21 MPs/L). Most particles were fibers (>90 %), black and blue in color, and predominantly composed of PET (15.9 %), styrenes (15.0%), PP (12.2%), and PA (9.4%). At a site, MP concentrations were highly variably, exceeding an order of magnitude, and were observed to be higher, lower, or exhibit no difference between base- and stormflows, depending on location and upstream basin characteristics. During an event, MP concentrations increased, decreased, or showed no pattern with increasing discharge depending on the site and event. Variations in concentration are attributed to temporal changes in rainfall-runoff characteristics and MP source and source inputs which included runoff from urban/suburban areas and the decomposition of larger plastic debris, among others. Similarities in the morphotype, color, and composition of MPs in water and atmospheric deposition samples suggest that atmospheric deposition is an important, if not the predominant, source of MPs in these headwater basins. Atmospheric deposition at three sites located across the region ranged from 0 to 670 MPs/m2/day. Atmospheric deposition during Hurricane Helene was 8.1 times, 3.4 times, and 22.7 times higher than pre-Helene rates at our three monitoring sites. Mean MP concentrations in channel bed sediments ranged between 500 and 1,000 MPs/kg at most sites along Richland Creek prior to Helene, indicating that they act as an important sink and secondary source of MPs that influence downstream particle transport and fate. Both caddisfly larva and casings contained MPs; mean concentrations ranged between 0 and 2.4 MPs/caddisfly and 0 to 3.79 MPs/case, respectively. Their occurrence in benthic biota suggests that MPs may be transferred to higher tropic organisms. Acute toxicity tests conducted on caddisflies indicate that microfibers were consistently the most toxic particles, followed by fragments and tire wear particles. Denser MPs posed less of a threat than more buoyant particles. Our data indicated that that the full range of variations in MP concentrations need to be determined to assess potential risks to biota, and the analysis of atmospheric deposition should be included in future studies of riverine ecosystems.
Improved Quantification of Stormwater Dilution in Wastewater Treatment Plant Influent Using Organic Mater Fluorescence
(NC WRRI, 2025-11-20) Osburn, Christopher L.; Zepeda, Sara Gabriela Cornejo
Improved Quantification of Stormwater Dilution in Wastewater Treatment Plant Influent using Organic Mater Fluorescence. Wastewater treatment plants (WWTPs) are designed to process residential and industrial sewer flow as influent; however, they often receive addition water during and after rainfall through inflow and infiltration (I&I). These unplanned inputs dilute the influent, reduce treatment efficiency, and increase operational costs. The objective of this study was to use organic mater fluorescence in a source apportionment model (FluorMod) to detect and quantify I&I contributions to WWTP influent. A reference library of 142 excitation-emission matrices (EEMs) was developed, representing six potential sources: wastewater, rainwater, groundwater, river water, soil leachate, and street runoff. Samples from ten municipal WWTPs across North Carolina were analyzed, including facilities spanning both the Coastal Plain and Piedmont regions. Each sources’ contributions to each unknown sample were estimated using FluorMod and, coupled with descriptive statistics, Pearson correlation, and time-series analysis, we gained insight into the variability and dynamics of I&I. Two Coastal Plain WWTPs showed median influent dilution by groundwater exceeding 83%, while others showed variable contributions from both rainwater and groundwater, with an overall median dilution of 16% and 9% respectively. Several Piedmont WWTPs showed median dilution values near 0%, only exhibiting episodic dilution events following substantial five-day antecedent precipitation (AP5). In several systems, the number of contributing I&I sources increased with higher AP5, consistent with rainfall activating additional inflow and infiltration pathways. Higher and chronic groundwater dilution occurred in the Coastal Plain relative to the Piedmont. However, the observed variability in I&I dynamics suggests that other factors including, but not limited to, sewer infrastructure conditions, hydraulic connectivity, soil saturation, geologic setting, and the spatial distribution of rainfall, likely play significant roles in modulating dilution behavior. Results for each plant were disseminated in two outreach efforts. First, reports that summarized the project scope, methods, and time series results were shared with each plant shared. Second, offers for an online discussion of results and deeper explanation of the FluorMod approach were made, culminating in three Zoom meeting held prior to the end of the project. Project results demonstrate that fluorescence-based source apportionment is a practical, cost-effective tool for identifying and quantifying I&I contributions from multiple sources. Future applications extend to exfiltration of sewage and other potential pollution sources that carry a fluorescence signature. The project outcomes represent a substantial innovative method to corroborate results from other I&I methods and investigate I&I issues in sewersheds at scale. FluorMod’s adaptability and accessibility make it a promising diagnostic for WWTP operators and may complement conventional field techniques to support more targeted infrastructure assessments, long-term monitoring, and investment planning to address I&I challenges.
Flood Typing and Mixed Population Study
(NC WRRI, 2025-03-31) Awasthi, Chandramauli; Archfield, Stacey A.
Flood frequency analysis is critical for managing flood risk, designing hydraulic structures, and developing floodplain management strategies. Accurately estimating the flood peak of varying associated risks can significantly reduce potential hazards to water infrastructure and human life. Despite long-standing recognition of the diverse nature of floods caused by different hydrological and meteorological processes, traditional flood frequency analysis methods often assume a homogeneous population of flood events. This study investigates five flood frequency approaches suitable for mixed flood distribution populations and evaluates their applicability to rivers experiencing floods from multiple hydrological processes. We aim to address three key research questions: (a) when application of a mixed frequency model is needed, (b) how an expert-based flood classification compares with a purely data-driven flood classification technique, and (c) which mixed flood frequency model can best estimate flood quantiles for varying return periods. First, we systematically conduct synthetic experiments to explore the application of the considered models, followed by their application to selected natural basins in the USA. The implications of this study extend to various stakeholders, including engineers, planners, and policymakers. By adopting a mixed population flood frequency analysis, decision-makers can gain a deeper understanding of flood risk, leading to more informed and effective flood mitigation strategies. Future research directions include refining the classification of flood-generating mechanisms and exploring the application of mixed population models in different climatic and geographical contexts.
Novel Mass Spectrometry Approaches for the Identification of Pesticides and Per- and Polyfluoroalkyl Substances in North Carolina Drinking Water Sources
(NC WRRI, 2024-01-18) Knappe, Detlef; Baker, Erin
Water quality can be adversely impacted by a wide range of organic contaminants, including pesticides and per- and polyfluoroalkyl substances (PFAS). To support the assessment of water quality by liquid chromatography, ion mobility spectrometry, and tandem mass spectrometry, a database of collision cross sections was developed for pesticides. The database can be used in both targeted and nontargeted analyses of aqueous samples. Within the database, a clear separation occurs between pesticides and highly halogenated exposomic compounds, such as per- and polyfluoroalkyl substances (PFAS), polybrominated diphenyl ethers (PBDEs), and polychlorinated biphenyls (PCBs), which will be useful in nontargeted workflows as unknown features can be tentatively assigned to specific exposomic classes without a true identification. The utility of these methods for the assessment of water quality and human exposure was explored. Using the database, atrazine was identified in 3 out of 13 surface water samples (11-40 ng/L). To assess the potential impact of land-applied biosolids on aqueous PFAS concentrations, 13 surface water sites and 9 private wells were sampled near fields that had received biosolids. Eight of the surface water sites contained detectable PFAS at concentrations up to 1,880 ng/L, and eight of the private wells contained detectable PFAS at concentrations up to 194 ng/L. The primary subclasses identified in the samples were perfluoroalkyl sulfonic acids and perfluoroalkyl carboxylic acids.
Developing a Watershed Prioritization Index to Guide Restoration Initiatives in the Upper Little Tennessee River Basin
(NC WRRI, 2024-03-28) Gibbs, W. Keith; Styers, Diane M.; Martin, Thomas H.
Our research objectives were to: 1) analyze geospatial data across the Upper Little Tennessee River Basin (ULTRB) to 2) rank subwatersheds for risks to water quality at multiple spatial scales and 3) link land cover change over time to water quality and aquatic biota. We extracted land cover and landscape feature data at multiple spatial scales to categorize attributes and rank subwatersheds from least to most susceptible to impairment. Water quality and aquatic habitat were sampled in a subset of subwatersheds to verify and validate geospatially derived prioritization rankings. Historical fish and macroinvertebrate assemblage data were compiled and georeferenced to identify gaps in knowledge and target supplemental sampling efforts. Although the ULTRB is predominantly forested (>87%), ~8.5% is developed and ~2.5% is agricultural. Some subwatersheds exceeded 78% developed land use and others exceeded 34% agricultural land use within a 30m riparian zone. Altered land uses (i.e., agriculture and development) were negatively correlated to steep slopes and concentrated in areas of gentler slopes found in riparian zones. Soil erodibility was not strongly correlated to land use or slope steepness. Agricultural land uses had the strongest negative correlation to larger substrate sizes and stream reaches with more agriculture had higher percentages of finer substrates (i.e., sand and silt). Development was also positively correlated to the percentage of silt within stream reaches. Visual habitat assessments were negatively correlated to percentage of agriculture and development. Macroinvertebrate indices of biotic integrity scores were also negatively correlated with agriculture and development. However, fish indices were only negatively correlated to development. Macroinvertebrate indices were most correlated to physical habitat and positively correlated to bottom velocity and percent cobble. Macroinvertebrates were negatively correlated to percent silt and weakly correlated to turbidity. Fish indices were negatively correlated to bottom velocity and positively correlated to riffle embeddedness and chain roughness. Macroinvertebrate indices were positively correlated to visual habitat assessments, whereas fish indices did not have a strong correlation to visual habitat assessments. In general, sites with the highest percentage of agriculture had the lowest habitat quality and, consequently, the lowest biotic integrity. Sites within more developed watersheds generally had suboptimal habitat quality but had better ecological integrity compared to agricultural subwatersheds. Sites with intact riparian zones and forested headwaters had the highest habitat quality and ecological integrity. These results confirm the importance of healthy riparian zones for maintaining water quality and ecological integrity, especially in areas that are highly susceptible to erosion and intense rainfall. We recommend continued monitoring of water quality, biota, and land use changes throughout the region to further elucidate these relationships. We also recommend targeting watersheds with the greatest impairment or risk of impairment to provide the most ecological uplift during restoration efforts. Effective restoration and watershed management will require active participation from individual landowners. With regional partners, we will help to establish watershed coalitions to educate landowners, promote hydrologically connected communities, and enhance stewardship of our local watersheds. Additionally, the transferability of techniques we used could be tested in other regions of the state with inherently different landscape attributes.
Fecal Contamination Source Tracking and Forecasting to Support Recreational and Cultural Development in the Black River Watershed
(NC WRRI, 2024-04-30) Harris, Angela; Emanuel, Ryan; Obenour, Daniel
The Coharie Tribe expressed interest in learning more about specific CAFO and human impacts on water quality along the Great Coharie River, with special attention to spatial and temporal variability of these impacts. The Tribe’s interest also aligns with opportunities to answer fundamental research questions about water quality in North Carolina’s Coastal Plain and areas similarly impacted by industrialized agriculture. We proposed research questions and objectives intended to help meet these needs while also advancing fundamental science concerning biological water quality. What are the dominant sources of fecal contamination at 3 sites in the watershed? How do these sources vary across base flow and storm flow conditions, and warm and cold seasons? Can forecasting tools be developed to capture elevated levels of fecal contamination in the watershed to support Tribal decision-making related to the use of river? To address these questions, low frequency and high frequency (determined by base and storm flow) sampling events were conducted at three sites in the Great Coharie River watershed from March 2022 to March 2023. Each sample was evaluated for total coliforms, Escherichia coli, molecular microbial source tracking (MST) targets, Total suspended solids, Total Kjeldahl Nitrogen, the sum of Nitrate and Nitrite, Ammonia, Total Phosphorus, Orthophosphate, Temperature, pH, dissolved oxygen, specific conductivity, and turbidity. An automated DNA extraction kit was optimized for use with surface water filters and nucleic acid was extracted from each sample. Following DNA extraction, samples were analyzed using polymerase chain reaction microbial source tracking targets including LA35, Pig2bac and hCYTB484 which is sensitive and specific to swine and poultry combined, swine, and human fecal contamination, respectively. Results from the MST assays indicate that the watershed likely suffers from fecal contamination from the three target hosts. Escherichia coli data collected throughout the study indicate that the three sites have contamination levels that would deem the Great Coharie River as impaired per the United States Environmental Protection Agency’s Recreational Water Quality Criteria. An explanatory model using multiple linear regression was developed with the available data. The preliminary model revealed that 2-day total rainfall is an important covariate to help explain E. coli concentrations in the watershed. In addition to the explanatory model, a random forest predictive model was constructed to predict whether the water was at an E. coli concentration above or below the EPA recommended threshold (i.e., 126 MPN/100 mL). When performing covariate importance analysis, it was determined that 2-day total rainfall was the most important covariate of any examined in this study when predicting E. coli. Lastly, a generalized estimating equation was developed, and this model also found 2-day total rainfall to be an important covariate. It is recommended that caution be used during primary contact recreation, which includes activities that involve increased likelihood of water ingestion or immersion like swimming, at each of the three sites due to elevated levels of E. coli- especially following an intense rainfall event. We also recommend that a land use study be undertaken to determine potential management strategies to minimize fecal contamination at the Marsh Swamp site, which experienced higher levels of E. coli contamination than other sites. Additionally, further studies into the human contributions to the watershed could be performed to help identify if further interventions related to human waste management need to be made.
Remove Per- And Polyfluoroalkyl Substance From Water By Combined Ion Exchange and UV/Sulfite Treatment
(NC WRRI, 2024-03-15) Sun, Mei
Per- and polyfluoroalkyl substances (PFAS) are contaminants of concern in North Carolina. Ion exchange (IX) is effective in separating PFAS from contaminated drinking water, but incapable of chemically breaking down the contaminants. Thus, waste streams from IX, including spent resins and regenerant, contain high levels of PFAS and are challenging to be disposed of safely. This project addresses this problem by strategically combining IX and resin regeneration via chemical PFAS degradation. We hypothesize that in this treatment train, PFAS can be separated from contaminated water and subsequently mineralized for safe and feasible waste management. With the goal of sustainable and cost-effective management of IX wastes for PFAS removal, this project includes tasks of determining the most suitable chemical treatment for IX wastes, optimizing the treatment conditions, and assessing the feasibility of the treatment train for water treatment practice. In this project, we proposed and evaluated several variations of a treatment train for PFAS removal and defluorination for water treatment. The most effective option for the treatment train design we identified in this project combines IX, NaCl/methanol regeneration, distillation, and UV/sulfite defluorination. Using this treatment train, PFAS at environmentally relevant levels (ng/L) can be captured by IX to near 100% by either PFAS-specific or generic resins. The spent resins can be regenerated using 60% methanol/5% NaCl solutions and reused in the following treatment cycles, with much better regeneration performance from the generic resins than PFAS-specific resins. The waste regenerant can be distilled to separate methanol for reuse, with simultaneous PFAS enrichment and volume reduction of the waste stream. Finally, the concentrated waste (still bottoms) can be treated with UV/sulfite to transform PFAS into nontoxic fluoride. This process is effective for most PFAS with almost 100% removal and >75% defluorination, although a few short-chain sulfonates and GenX are more challenging to treat. The generic resins (e.g., IRA910 tested in this project) are recommended for the treatment train over PFAS-selective resins and can be reused for treatment over multiple cycles. All the detailed conditions for each treatment step were optimized. Overall, PFAS with different chain lengths and functional groups can be effectively removed and eventually defluorinated in different water matrices using the proposed treatment train.
Internet of Water: Research and Development Toward a Linked Data System and Foundational Knowledge Network for the Internet of Water: Final Report
(NC WRRI, 2023-10-31) Doyle, Martin; Onda, Kyle
Internet of Water: Research and Development Toward a Linked Data System and Foundational Knowledge Network for the Internet of Water is a project designed to improve the discoverability of water-related datasets. The main objective of this project was to establish a foundational framework for a contributor-based system that facilitates regular harvesting and cross-referencing of metadata through two key components: (1) facilitating the Internet of Water community in publishing detailed, machine-readable, and cross-referenced metadata (linked data); and (2) developing a centralized crawler/harvester to catalog all the linked data into a single knowledge graph, serving as a part of the index for eventual Internet of Water search utilities. The project built on previous efforts to continue and enhance the development of Geoconnex, a linked data system and foundational knowledge network for water data. Geoconnex can be conceptualized as an operationalization of the Open Geospatial Consortium Environmental Linked Features Interoperability Experiments for the United States in the domains of water science and management. Our methods encompassed data-publisher-oriented research and engagement, development of Geoconnex architecture, and implementation of the Geoconnex system. The results of this project include a performant infrastructure leveraging semantic technology and open, modern API standards that allow data providers to independently publish metadata on the web in a manner that results in their data becoming linked to other providers’ data where spatially, hydrologically, and topically relevant. If adopted by data providers at a wide scale, this infrastructure would enable users to easily find water information across a variety of topics and geographies via user-friendly search applications. Through this project, we have concluded that further work is needed to encourage participation in the system and use of the infrastructure. We recommend that future work include the development of a user-friendly interface, technical advancements to improve the usability of the system and enable scaling, a promotional campaign to build the userbase, and the development of a governance structure. Potential future work could also establish best practices to use the same technologies to enable the automatic translation of observation and model data across data systems, fostering improved interoperability of data in addition to the improved discoverability already enabled by the currently implemented infrastructure.
Enabling rural water utilities to create, transfer information, and use digital service area boundaries: Year 2 final report
(NC WRRI, 2023-10-31) Ward, Ashley; Onda, Kyle
Enabling Rural Water Utilities to Create, Transfer Information, and Use Digital Service Area Boundaries is a project that aims to increase equity by lowering the digital divide for small utilities by providing free, open-source tools that support regional and statewide planning. The project’s objectives are to (1) enable water and wastewater utilities to create and maintain digital service area boundaries through the development of an open-source, free technology called BoundarySync (formerly ABOUT-US); and (2) create an online tool that spatially links digital service area boundaries with parcel data to better equip small rural utilities to meet new Environmental Protection Agency (EPA) regulatory reporting requirements through the development of the Lead Service Line Inventory Tool (LSLIT). The creation of BoundarySync, in combination with LSLIT, will assist small rural utilities in meeting EPA’s new reporting requirements and provide more complete information on where lead service lines are likely to exist in their service areas. This information may inform the allocation of federal funding to support lead service line replacement and could help utilities pursue external funding for infrastructure improvements and enhance decision-making. In addition, BoundarySync will provide the state with up-to-date data to support regional and statewide planning. In Year 1, we developed BoundarySync, a user-friendly tool for the creation and maintenance of digital service area boundaries and assessed how rural utilities could use digital service area boundaries in combination with other data to solve a pressing challenge. Through these engagements, we learned that new EPA rules requiring water utilities to develop lead service line inventories placed a significant burden on rural water utilities which often do not have access to GIS software. In Year 2, we developed the LSLIT to give small rural utilities an easier way to comply with the new rules. Our methods included (1) close collaboration with the North Carolina Rural Water Association (NCRWA) to identify and reach a representative group of users to provide feedback on tool planning and development, (2) engagement with the North Carolina Department of Environmental Quality (NC DEQ) on the future hosting and maintenance of BoundarySync, (3) leveraging additional funding for software development support to ensure that BoundarySync had a modern and intuitive design, (4) working with a Ph.D. student and a postdoc to integrate digital service area boundaries created in BoundarySync with parcel data through LSLIT, and (5) participating in NCRWA conferences and workshops and developing instructional and promotional videos to encourage adoption and use of BoundarySync and LSLIT. Our results include (1) BoundarySync; (2) engagement with a representative group of water utilities, NCRWA leadership, and NC DEQ in a user-based design process to refine BoundarySync’s usability; (3) the Lead Service Line Inventory Tool (LSLIT); (4) engagement with rural water utilities in the planning, development, and continued improvement of LSLIT; (5) a GitHub repository for BoundarySync and LSLIT enabling other states or entities adapt these tools to their needs; (6) two instructional videos (Getting Started in BoundarySyc and Using the Admin Role in BoundarySync) and a promotional video about BoundarySync; and (7) a promotional video that details the challenges rural water utilities face in modernizing their data infrastructure and highlights NCRWA’s work. Through our engagements with North Carolina rural water utilities and partners in other states, we have discovered that there is a great deal of need for these two products not only in North Carolina but in rural areas across the US. As a result, we recommend that in addition to continuing to host, maintain, and promote these two tools in North Carolina we develop a national version of LSLIT and promote the use of LSLIT and BoundarySync throughout the US.
Performance Evaluation of Novel Resins in Flow-Through Columns for PFAS Removal from Drinking Water and Treated Wastewater
(NC WRRI, 2023-11-30) Coronell, Orlando; Leibfarth, Frank; Manning, Irene; Gorzalski, Alexander; Chew, Guan
Per- and polyfluoroalkyl substances (PFAS) are known bioaccumulative and persistent chemicals which pollute natural waters globally. However, there exists a lack of granular sorbents optimized to efficiently remove PFAS at environmentally relevant concentrations. To address this need, we previously developed a platform approach for fabrication of regenerable granular sorbents (i.e., ionic fluorogel (IF) resins) that selectively remove PFAS over background competing contaminants. In this project, we modified 1st generation IFs to realize structurally tunable and chemically stable resins with systematic variations of network architecture and cation density. Chemical stability was demonstrated through accelerated degradation studies. Investigation of the library of chemically stable IFs enabled insights into structure–property relationships important to high affinity PFAS binding. Results show that the chemically stable IF resin outperformed a leading commercially available ion-exchange resin in batch and column tests. Using as a guiding principle in sorbent development the cost and availability at large scale of base materials, we developed a 2nd generation IF resin based on trifunctional crosslinkers with a more translationally relevant profile in terms of chemical inputs and cost. Batch results from challenge tests with various NC waters showed that the 2nd generation IF resin had a PFAS-removal performance level intermediate between those of two leading commercial ion exchange resins. The 2nd generation IF resin had a substantially better performance in drinking water sources than in wastewater matrices. Results also showed that there is the potential to improve PFAS removal performance via optimization of the resin synthesis procedure. Additionally, because of the high modularity and copolymerization amenability of the platform chemistry used for IF synthesis, we hypothesize that future tuning of the fluorinated backbone can be used to optimize PFAS partitioning, and resin swelling and flexibility and enable superior PFAS removal across water matrices.
A citizen science internship program to quantify racial and economic disparities in lead levels in drinking water across North Carolina
(NC WRRI, 2023-09) Cooper, Caren; Berglund, Emily; Johnson, Valerie
“A Citizen Science Internship Program to Quantify Racial and Economic Disparities in Lead Levels in Drinking Water Across North Carolina” (Grant #21-17-W) has four objectives: 1) characterizing lead plumbing in North Carolina, 2) mentoring underrepresented undergraduate researchers, 3) increasing public awareness and self-efficacy related to drinking water, with a focus on marginalized communities, and 4) refining a statistical model to estimate water lead levels using citizen science data. To characterize lead plumbing in North Carolina, we engaged participants in the Crowd the Tap citizen science program. Since Crowd the Tap was started until August 8, 2023 1796 North Carolina residents have screened their home through Crowd the Tap, identifying 1989 in-home plumbing materials and 1671 service line or well casing materials, of which 32 in-home pipes and 20 service lines or well casings were leaded. Of these households, 365 had their water tested in the lab and 208 conducted at-home lead tests. Crowd the Tap identified 226 households with some detectable lead levels, seven of which were greater than 10 ppb. There were nine positive at-home lead tests. We have recruited 43 undergraduate and post-baccalaureate research interns. We have recruited 12 Shaw University students, one student from North Carolina Agricultural and Technical University, one student from Montreat College, 28 students from North Carolina State University, and one post-baccalaureate intern. Altogether, 16 of the 43 interns have underrepresented racial or ethnic identities, and six others are people of color. Through these internships, students learned about public engagement through faith communities, community activist groups, community-based health organizations, and in their own communities. To increase public awareness and self-efficacy, we engaged people through various partnerships, which ensured that we engaged people of diverse backgrounds. We recruited 828 university students, 480 households connected to university interns, 296 households from high schools, 26 from faith communities, 21 from a corporate volunteer program, 31 from the Southeastern Wake Adult Day Center (a community-based health organization that supports primarily low-income Black residents of Wake County), and 109 households came to the project independently. Altogether, we recruited 969 White households, 238 Black or African American ones, 109 Asian, 71 Hispanic or Latino, three American Indian or Alaska Native, one Native Hawaiian or Other Pacific Islander, and 231 households with two or more races or ethnicities present. We are using the citizen science data collected from Crowd the Tap to test a Bayesian Belief Network (BBN) model to predict lead levels in household drinking water. Preliminary results indicate that water quality test strips have high degrees of error and that the BBN has low predictive performance for predicting lead levels in drinking water. Our analysis is ongoing, and we are continuing to explore BBN’s predictive ability for a range of samples. Thus, Crowd the Tap has made significant progress towards each of these objectives. These results can provide guidance to utility companies seeking to prioritize areas for lead infrastructure removal that prioritize equity. Furthermore, our program has supported the professional development of undergraduate researchers from Historically Black Colleges and Universities and/or have underrepresented identities.
Data-Driven Analytics Tools to Support Prioritized Management of Stormwater Infrastructure
(NC WRRI, 2023-09-19) Barclay, Nicole; Smith, Michael; Qi, Jingyi
Maintenance of aging underground infrastructure compounded with urbanization encroachment and economic pressures have caused a number of challenges for municipal agencies that are tasked with managing and operating these infrastructure systems. However, the budget and time constraints that municipalities often face present challenges for managing culverts. Reliable prediction of infrastructure conditions can help these municipal agencies manage this burden for culverts by providing decision support regarding optimal renewal, replacement, and maintenance, which offers reduced overall costs and improved system reliability. As opposed to traditional physics-based models, data-driven models (e.g., machine learning) offer benefits in processing large datasets where some data is missing or there is “noisy” data. Considering these compounding issues, our study aimed to develop predictive models for culvert condition assessment to better inform repair and replacement schedules. The overall goal is to complement and bolster existing toolsets for prioritization & asset management. The project objectives were to 1) estimate current conditions based on measured data and infrastructure characterization, 2) design models to predict conditions via data analytics, 3) train and validate the models by evaluating each model and verifying results, and 4) identify at-risk pipelines and culverts within the study area. This report describes the data-driven modeling approach via machine learning for predicting culvert conditions to identify at-risk culverts using Charlotte-Mecklenburg Stormwater Services existing data inventory. Several supervised machine learning models were considered and their performance evaluated, respectively. Simulation test results and discussion are included to showcase the key contributions and findings of this work. The results show that one model, the random forest classifier (RFC) gives the prediction performance. The main contributions highlighted in this report include: 1) the performance of four different machine learning models which are compared for the target application to predict culvert conditions, 2) the application of the model for stormwater infrastructure asset management, 3) the viewpoint that the machine learning approach offers economic savings for utilities, 4) the machine learning approach provides potential for improved reliability of stormwater infrastructure.
Empowering underrepresented citizens of Charlotte-Mecklenburg with a voice and role in municipal storm water program goals
(NC WRRI, 2022-08-31) Patterson, Bryan; Guyer, Regina; Raburn, Kari
In our research study “Our Voice Our Water” focused on community engagement and stormwater knowledge within three communities in Charlotte, NC. A goal of our study sought to address the linkages between stormwater and trash, litter, and/or illicit discharges. The neighborhoods selected were historically underserved communities of Historic Washington Heights, Lincoln Heights, and Northwood Estates. Primary research objectives included: 1) Applied research techniques by Johnson C. Smith University faculty, Rising Solutions, PLLC staff, and an AmeriCorps Member were implemented to access the behaviors and knowledge of stormwater issues with community citizens. Community members were encouraged to “Be Heard” by sharing their knowledge and perspectives. Research techniques included focus groups, surveys, and stream indexing. 2) The 3-1-1 Call Center Data was evaluated to gain a historic understanding of the citizens’ concerns and follow up actions by Charlotte Mecklenburg. 3) Engagement with neighborhood leaders and citizens to build community trust and understanding through education and outreach opportunities as well as community meetings. The “3-1-1, Having a Voice Workshop”, was a model to provide opportunities for citizens to have a voice by expressing concerns which, in turn, helps to build relationships, two-way communication, and enhance trust. The “Neighborhood Empowerment Model” was a powerful result of the research, it enabled us to create a new systematic approach to engage with stakeholders and communities. These purposeful interactions with neighborhood leaders and city and county municipalities-built community trust and understanding through education and outreach opportunities as well as increased community stream and residential cleanups.
Intensification of Anaerobic Digestion: Co-Digestion of Grease Interceptor Waste (GIW) with Thermally Hydrolyzed (TH) Biosolids
(NC Water Resources Research Institute, 2022-07-12) de los Reyes, Francis
Intensifying methane production during anaerobic digestion (AD) is a key factor in making anaerobic digestion economically feasible. We explore two key ideas in biomethane production: using thermally hydrolyzed (TH) biosolids, and co-digestion with grease interceptor waste (GIW). We used modified biochemical methane potential (BMP) tests in several Phases in triplicate: Phase 1 investigated the optimal COD-based substrate to inoculum (S/I) ratio for mono-digestion of TH biosolids and GIW, Phase 2 investigated co-digestion, and Phase 3 investigated codigestion under conditions modified to be more like those of full-scale operations. Application of a TH pretreatment to Phase 1 biosolids, Phase 3 mixed and thickened (M&T) biosolids, and Phase 3 co-substrate (M&T biosolids and GIW) increased the fraction of soluble to total COD by 51.7%, 25.2%, and 25.5%. Modified Gompertz models identified maximum methane potential (MMP) and maximum daily methane production rate (MDMPR). The most productive Phase 1 GIW mono-digestion had a MMP three times that of Phase 1 TH biosolids but had a lag period that was almost seven times longer. In Phase 2 biosolids monodigestions, TH increased MMP and MDMPR in biosolids but no significant effect was observed in GIW. Inclusion of TH GIW and GIW without TH in Phase 2 co-digestions both increased MMP and MDMPR, compared to mono-digestions of biosolids, but there was no significant difference in MMP or MDMPR between these co-digestions. These are consistent with very similar microbial communities at the end of the incubations (analyzed using 16S gene sequencing). Modified Gompertz Model fitting of Phase 3 results also showed no significant difference between TH and raw co-digestions of M&T biosolids and GIW in terms of MMP or MDMPR. Again, these were consistent with the microbial community analysis results, that showed very similar communities at the end of the runs. One reason for the difference in TH effects between Phase 3 and other Phases was the inclusion of primary solids in M&T biosolids. The benefits of TH appeared to be less dramatic in substrates that contained primary solids than in those exclusively composed of waste activated sludge. TH was shown to be beneficial in mono-digestion of waste activated sludge biosolids, more than doubling MMP and increasing MDMPR by more almost seven times. However, TH does not intensify anaerobic digestion when applied to co-substrates of biosolids and GIW. This study found that TH is suitable for application to biosolids at the current state of practice. Incorporation of TH GIW showed no consistent benefit over addition of untreated GIW. The application of TH to GIW does not always lead to increased methane yield or improved digestion kinetics: many factors, including the type of biosolids (WAS vs. primary sludge; thickened vs. unthickened) affect the methane yield.
Biological Condition in NC Urban Streams Phase III: Development of a Web Application
(NC Water Resources Research Institute, 2022-05-13) Paul, Michael; Allen, Diane
Using the model of biological potential along urban1 land cover gradients in North Carolina constructed using EPT and biotic index metrics, a web application (ArcGIS StoryMap) was constructed to visualize the results for every urban (>1% impervious cover) watershed in North Carolina. The web application allows any user to identify the potential EPT richness and NCBI score for any watershed constrained by urbanization. This constrained biological potential is provided as both a scaled value (comparable across the state) and ecoregion specific actual predicted score. Watershed managers should be able to use this information to help guide urban stream prioritization, planning, restoration and monitoring and the concept would support construction of a tiered aquatic life use framework. NCDEQ agreed to host the web application as part of the ArcGIS StoryMap service and the plan is to migrate the web application to the state for this purpose.
Tertiary Anammox for Sustainable Nitrogen Removal
(NC Water Resources Research Institute, 2022-04-28) Aziz, Tarek; de los Reyes, Francis; Young, Anthony
Concentrated discharges of nitrogen into our waterways have led to serious environmental impacts, such as eutrophication and algae blooms. Due to these discharges, stringent total nitrogen (TN) discharge limits have been placed on Water Reclamation Facilities (WRFs). Mainstream deammonification offers a novel approach for improving nitrogen removal at these facilities by harnessing anaerobic ammonia oxidizing bacteria (Anammox). However, the success of this process has been limited. Previous research at NC State explored the conversion of tertiary filters to mainstream deammonification filters which were shown to be capable of an average total inorganic nitrogen (TIN) removal rate of 91%, with effluent TIN reaching below 2 mg/L-N. However, this research suggested that nitrate loading concentrations were the limiting factor in meeting 2 mg/L-N TIN effluent limits. Incorporating partial denitrification (PdN), the conversion of nitrate to nitrite for subsequent use in Anammox, offers a promising solution. The goal of this research was to explore the TIN removal capability and feasibility of a PdN-Anammox (PdNA) filter under typical filter loading conditions, in comparison to a conventional denitrification (FdN) filter. Different carbon loading strategies in pilot scale filters confirmed that maintaining a nitrate residual of >1.5 mg-N/L allowed for the highest PdN conversion efficiencies and in turn increased Anammox activity. Furthermore, the carbon requirement (C/N ratio) of 5.1 g COD/g TIN in pilot scale FdN was higher than the 2.08 g COD/g TIN achieved in the PdNA filters, demonstrating the cost benefits associated with mainstream deammonification and the application of PdNA. In addition to greater than 50% reduction in supplemental carbon, other benefits include nearly 38% reduction in oxygen demand and reduction in excess sludge in comparison to conventional BNR processes. Through this pilot study, PdNA was demonstrated to provide TIN removal efficiencies of greater than 80%. With further research, stable TN removal at WRFs at typical filter loading rates can be achieved, and with that, the possibility for substantial operational expenditure (OPEX) savings.
Fostering Engagement in the Cottage Grove Neighborhood Through Community Involvement in Riparian Cleanup to Enhance Stormwater Quality
(NC Water Resources Research Institute, 2021-12-12) Sills, Stephen
This community engaged project with residents of the Cottage Grove neighborhood in East Greensboro addressed environmental justice concerns with unnamed tributaries of South Buffalo Creek that pass through a former landfill containing incinerated waste. Many communities of color have, over the last century, experienced the destruction of historical landmarks, erosion of social ties through urban renewal projects, and challenges to health by environmental hazards. This project involved the residents and children of the Cottage Grove community through in collecting oral histories of long-time residents, conducting surveys, volunteering during trash/litter clean-up days (also including university student and environmental groups), and finally in citizen-science stormwater data gathering. Twice monthly we meet with residents and community leaders. Youth environmental education included virtual summer camps, environmental justice literacy lessons, and an Earth Day challenge week. We held quarterly community workdays with over 242 volunteers the neighborhood, local universities, other environmental groups, and AmeriCorps. More than 250 bags of litter, a dozen shopping carts, bicycles, strollers, car seats, fire extinguishers, and other miscellaneous debris have been removed from the creek. We collected oral histories, surveys, and historical documentation instrumental in bringing $356,000 in additional funding to environmental justice activities in the neighborhood including a three-year Robert Wood Johnson Foundation Award.
Effects of Extreme Flooding on Water Quality in Areas of Dense Food Animal Production
(NC WRRI, 2022-02-07) Stewart, Jill; Serre, Marc; Kothegal, Nikhil; Christenson-Diver, Elizabeth
Storm events are expected to increase with climate change, with the potential to adversely impact environmental quality and public health. This study assessed impacts of Hurricane Florence on water quality in areas of dense food animal production. Twelve surface water sites in rural, eastern North Carolina were sampled before and after Hurricane Florence to assess storm effects and duration of impacts. Concentrations of fecal indicator bacteria Escherichia coli were not significantly different in our first sampling event, conducted 10 days after Hurricane Florence. Instead, the longitudinal data show that first runoff events, defined as rainfall following a dry period, are associated with increased concentrations of fecal indicator bacteria in these watersheds. Additionally, watersheds with larger commercial hog operations (CHOs) and CHOs closer to sampling sites, as well as sites with larger number of households closer to sampling sites had increased E. coli concentrations. Unlike E. coli concentrations, we found that mean ranks of microbial source tracking markers associated with swine wastes (pig-2-bac) and human wastes (HF183) were higher immediately after Florence (Mann Whitney U p=0.009, p=0.0003 respectively). The swine MST marker was markedly elevated at several sites in watersheds with CHOs, while the human MST marker suggested diffuse human contamination across study sites. Antimicrobial resistance among E. coli isolated from study waters was also higher in watersheds with CHOs. Resistant E. coli was detected in 46% (n=147) of samples collected downstream of CHOs compared to 22% (n=94) of samples collected in watersheds without CHOs, resulting in a relative risk of 1.47 (95% CI: 1.21, 1.78). Bacteria with multiple antibiotic resistance was isolated in 15 of the CHO-associated samples and 1 background sample. Occurrence of AMR does not appear to be driven by precipitation, suggesting other dynamics such as spray events or antibiotic use practices may better explain contributions to resistant bacteria to surface waters. These results help clarify the effects of extreme flooding on microbial contamination of surface waters and can inform strategies for waste management, antibiotic use, and one health surveillance. Ultimately, this work contributes information to build more resilient agricultural systems and communities better prepared to weather the storms that frequent the North Carolina coast.

Browse

Recent Submissions