Reports
Characterization of chloride sources in groundwater in the Piedmont, Blue Ridge, Ridge And Valley, and Appalachian Plateau Physiographic Provinces of Maryland
2026, Firdous, R., Ulizio, T., and Van Ness T.
Open File Report 26-02-01
Key Results
Groundwater and wastewater samples were collected from 28 sites across central and western Maryland to evaluate chloride sources and associated geochemical indicators. Samples were collected from septic effluent, wells at municipal landfills, agricultural wells, road salt–impacted wells, natural formation brine-affected wells, domestic wells, and a saltwater treated swimming pool. Chemical analyses included major ions, minor halides, nutrients, trace metals, and wastewater tracers (artificial sweeteners and caffeine). The purpose of this study was to: (1) characterize chloride concentrations across distinct sources, and (2) evaluate geochemical ratios and tracers that support source identification. Key results from this investigation are:
- Piper diagram analysis showed two dominant water types: Ca–Mg–HCO₃ waters, representative of natural and agricultural influences; and Na–Cl waters, characteristic of anthropogenic inputs (septic effluent, landfill leachate, road salt). Septic and formation brine samples plotted toward the Na+K field, while agricultural and landfill monitoring well water clustered in the Ca–Mg field.
- Chloride concentrations varied widely across source categories. Septic effluent samples exhibited the broadest concentration range from 37 to 4,000 mg/L (median 215 mg/L). Municipal landfill wells contained 20–110 mg/L chloride (median 65 mg/L), whereas agricultural well water ranged from 2.5 to 8.5 mg/L (median 3 mg/L), with elevated values observed near livestock facilities. Road salt–impacted wells showed the greatest variability among groundwater sources, with concentrations ranging from 26 to 750 mg/L (median 455 mg/L), and the highest value measured at a salt storage facility. Formation brine-affected wells contained 86–170 mg/L chloride, whereas the saltwater treated swimming pool sample exhibited 1,700 mg/L. Chloride concentration ranged from 11 mg/L to 910 mg/L in the two domestic well waters sampled.
- Sample pH values ranged from 5.8 to 8.7. Domestic well samples were acidic (5.9–6.5), whereas septic effluent and formation brine affected samples were neutral to alkaline (7.1–8.7). Specific conductance was strongly correlated with chloride (R² = 0.986). The highest conductance (12,000 µS/cm) and TDS (6,600 mg/L) occurred in a septic effluent sample, consistent with concentrated wastewater affected by water-softener brine.
- Septic effluent exhibited generally low dissolved oxygen (DO) concentrations (median < 5 mg/L) and elevated dissolved organic carbon (DOC) (up to 65 mg/L). Landfill monitoring well samples exhibited very low DO (<1 mg/L) and low DOC (≤4.2 mg/L). Domestic wells were generally oxic with low DOC, formation brine affected well water was oxygen-depleted (<1 mg/L). In contrast road salt-impacted wells showed variable DO conditions, ranging from oxygen-depleted (<1 mg/L) to well-oxygenated (>7 mg/L), with DOC not detected. Agricultural well waters were well-oxygenated with low DOC.
- Box-plot analysis of ion-to-chloride and nutrient-to-chloride ratios reveals distinct but overlapping chloride source signatures based on the combined MGS dataset and literature-derived source ranges (Panno and others, 2005). Na:Cl ratios in road salt–impacted samples cluster near and below halite mass ratios with a narrow interquartile range (MGS median: 0.289), whereas septic effluent exhibits frequent sodium enrichment (Na:Cl > 0.60). Calcium and magnesium display greater variability, with elevated ratios most pronounced in agricultural wells (Ca:Cl > 3.0), while road salt– and brine-affected samples remain consistently low. Landfill-affected groundwater exhibits an intermediate Na:Cl median of approximately 0.38, distinguishable from septic systems by lower sodium enrichment and from road salt by increased divalent cation variability. K:Cl and especially TN:Cl ratios provide additional discrimination, as they are elevated in septic effluent and animal-waste–affected samples, but remain low in mineral salt–dominated sources.
- When interpreted together, geochemical ratios provide practical fingerprints for chloride source identification. Road salt–impacted groundwater is characterized by Na:Cl ratios near and below that of halite and consistently low Ca:Cl, Mg:Cl, K:Cl, and TN:Cl. Septic effluent exhibits elevated Na:Cl, K:Cl, and TN:Cl and is uniquely associated with wastewater tracers. Agricultural groundwater is distinguished by elevated Ca:Cl and Mg:Cl ratios relative to sodium, often at low chloride concentrations. Landfill-affected groundwater shows intermediate Na:Cl ratios with elevated bromide (lower Cl:Br) and greater divalent-cation variability. Formation brine-affected groundwater is identified by low Cl:Br ratios and strontium enrichment. However, care must be taken in using ratios with cations for source identification, which can be affected by adsorption and desorption on subsurface materials during transport.
- Chloride:bromide ratios for septic effluent (246–3,649) and road salt–impacted wells (1,143–2,778) overlapped in the zone characteristic of rock salt (halite) inputs. Landfill wells exhibited moderate ratios (177–1,806), while formation brine samples plotted at lower ratios (142–156) due to bromide enrichment relative to chloride. Agricultural wells showed ratios of 264–708 that mainly reflect uncontaminated fresh groundwater (with respect to chloride). Domestic wells were variable, with one well plotting among the highest ratios.
- Chloride:bromide versus chloride concentration mixing curves provide the best differentiation of chloride sources in Maryland waters. Points for septic effluent and road salt affected waters trended towards the high-ratio halite end members for the chloride sources, landfill wells plotted along mixing zones toward the moderate ratio end members, and formation brine-affected wells formed a zone towards distinct low-ratio end members. The close agreement between observed field data and locally derived mixing trajectories demonstrates the value of Maryland-specific endmembers for chloride source interpretation. All of the agricultural well waters had low chloride concentration (<10 mg/L) such that no significant trend towards a high chloride source could be conclusively determined. If animal waste had produced substantially elevated chloride, a trend towards slightly higher Cl:Br might be expected based on literature data.
- Septic effluent samples clustered at high total nitrogen and high chloride with elevated Cl:Br, while agricultural wells showed high nitrogen and low chloride. Road salt impacted well water samples exhibited high chloride and Cl:Br and negligible nitrogen. Formation brine affected well waters also had low nitrogen and had low Cl:Br. This dual-tracer approach provided effective separation of sources where Cl:Br alone overlapped.
- Strontium showed the strongest correlation with chloride (ρ = 0.71, p < 0.0001) with enrichment in formation brine-affected wells. Barium showed a moderate overall correlation with chloride (ρ = 0.54, p = 0.0030) while a strong relationship was observed within the limited dataset of road salt–impacted wells (R² = 0.991), although this observation should be interpreted with caution. One well near a salt storage facility exceeded the USEPA MCL of 2.0 mg/L. Lithium showed weak correlation with chloride and limited tracer utility. These minor cations are also affected by cation exchange so care must be taken in interpretation. Other metals (Fe, Mn, As, Pb, Cu) were primarily controlled by redox, chemical precipitation, and adsorption/desorption conditions rather than salinity.
- Artificial sweeteners (acesulfame and sucralose) were detected at elevated concentrations in septic effluent and at lower concentrations in domestic and landfill well waters. Trace levels of acesulfame were detected in two road salt–impacted wells, indicating localized septic influence or mixed-source conditions. Caffeine was not detected in any samples.
- Many groundwater samples exhibited geochemical signatures consistent with contributions from more than one chloride source. The integrated use of Cl:Br ratios versus chloride mixing plots, other ion-to-chloride ratios, nitrogen, and wastewater tracers provided a regional-scale framework for constraining chloride sources, while recognizing that complete separation is not always possible in areas with overlapping land uses and complex hydrogeologic and geochemical conditions.
