Field Sample Collection

Task 1.—Mercury Isotope Sampling

Twenty-four environmental samples of new (2018–19) or archived (2008–09) particulate samples from the terrestrial BNC or surface sediment from Sinclair Inlet (OU B Marine and outer Sinclair Inlet) or representative bays in Puget Sound (table 2; fig. 5) were collected as described in Opatz and others (2023). The three representative bays are similar to Sinclair Inlet in size, depth, and geometry. Briefly, surface (0–2 centimeters [cm]) bed-sediment samples were collected by a boat-deployed box corer into plastic or glass containers. Particulate material was collected from floor troughs in Dry Docks 5 and 6 using a plastic spatula into plastic jars. Particulate material was collected from the bottom of monitoring wells (MW) 709, 412, and 722, which are in contaminated fill material, by lowering pre-cleaned polyethylene tubing to the bottom of the well and pumping the slurry into a jar. After settling, overlying water was decanted and the remaining particulate material was shipped for analysis. Samples were shipped frozen to the USGS Mercury Research Laboratory (MRL) in Middleton, Wisconsin. Samples were lypholized and homogenized prior to mercury concentration and mercury stable isotope analyses.

Table 2.    

Total mercury and methylmercury concentrations and mercury stable isotope ratios in 2018–19 and archived 2008–09 sediment samples from the Bremerton Naval Complex and Sinclair Inlet, Bremerton, Washington.

[Location: SI, OU B Marine or outer Sinclair Inlet; Bay, Representative Bay; BNC, Terrestrial Bremerton Naval Complex. Sample site short name: Location shown in figure 5. DD, dry dock; MW, monitoring well, PSEMP LT, Puget Sound Ecosystem Monitoring Program – Long-Term sampling site. Sample date: YYYY-MM-DD, year-month-day. Abbreviations and symbols: δ202 average, natural abundance mercury isotope '202-Hg', ratio relative to a reference standard (NIST 3133); Δ199 average, natural abundance mercury isotope '199-Hg', ratio data; na, not analyzed; ng/g, nanogram per gram; dw, dry weight. See Opatz and others, 2023 for method details and quality control information]

Table 2.    Total mercury and methylmercury concentrations and mercury stable isotope ratios in 2018–19 and archived 2008–09 sediment samples from the Bremerton Naval Complex and Sinclair Inlet, Bremerton, Washington.

Location	USGS station ID	Sample site short name	Matrix description	Sample date (year-month-day)	Total mercury (ng/g dw)	Methylmercury (ng/g)	Calculated percent methyl of total	δ202 average	Δ199 average
SI	473323122382110	BNC-52	Surface sediment	2008-08-15	776	2.3	0.3	−0.41	0.01
SI	473334122373810	PSNS-71	Surface sediment	2008-08-18	565	1.2	0.2	−0.45	0.02
SI	473328122375200	BNC-61	Surface sediment	2009-08-04	1,170	2	0.2	−0.42	0.07
SI	473305122391400	BNC-26	Surface sediment	2009-08-04	758	2.7	0.4	−0.48	−0.01
SI	473316122384800	BNC-42	Surface sediment	2009-08-04	883	2.5	0.3	−0.49	0.04
SI	473249122394300	PSEMP LT-34	Surface sediment	2018-04-16	780	2.1	0.3	−0.38	0.04
SI	473249122394300	PSEMP LT-34	Surface sediment	2019-04-15	810	2.38	0.3	−0.41	0.05
SI	473242122390300	PSEMP LT-40030	Surface sediment	2018-04-16	620	3.36	0.5	−0.47	0.05
SI	473242122390300	PSEMP LT-40030	Surface sediment	2019-04-15	950	5.32	0.6	−0.43	0.07
SI	473232122383110	Pt Orchard	Surface sediment	2009-08-03	505	0.41	0.1	−0.51	−0.02
SI	473208122403700	SI-6	Surface sediment	2009-08-06	669	3.2	0.5	−0.45	0.05
SI	473228122395800	SI-11	Surface sediment	2009-08-06	661	3.2	0.5	−0.47	−0.01
SI	473241122385900	SI-18	Surface sediment	2009-08-06	751	0.84	0.1	−0.47	0.06
SI	473306122375000	SI-28	Surface sediment	2009-08-06	272	2.4	0.9	−0.59	−0.03
Bay	470406122543110	Budd Inlet	Surface sediment	2008-08-13	168	2.7	1.6	−0.6	0.01
Bay	480123122312710	Holmes Harbor	Surface sediment	2008-08-11	30.1	0.65	2.2	na	na
Bay	474323122384910	Liberty Bay	Surface sediment	2008-08-12	176	4.7	2.7	−0.56	0.03
BNC	473336122381405	DD5 NE trough	Particulate material from dry dock trough	2018-11-27	220	1.76	0.8	−0.42	0.19
BNC	473336122381404	DD5 SW trough	Particulate material from dry dock trough	2018-11-27	150	0.43	0.3	−0.47	0.12
BNC	473316122383308	DD6 SE trough	Particulate material from dry dock trough	2018-11-20	200	0.73	0.4	−0.87	0.13
BNC	473316122383306	DD6 SW trough	Particulate material from dry dock trough	2018-11-20	16	0.32	2	na	na
BNC	473322122390701	MW722	Particulate material from well bottom	2019-03-18	44,000	54.9	0.1	−0.48	0.04
BNC	473323122382902	MW412	Particulate material from well bottom	2019-02-20	21,000	142	0.7	0.18	−0.04
BNC	473324122382202	MW709	Particulate material from well bottom	2019-02-27	18,000	6.7	0.04	−0.46	0.07

Task 2.—Dry Dock Sampling

Twenty-six water samples were collected in November 2018 from within the dry dock system to identify the source of mercury previously measured in pump well samples. The pump well water is a composite of groundwater, seawater, and, when a nuclear ship is in dock, cooling water (recirculated Sinclair Inlet seawater). Drainage relief tunnels along the east and west walls of the dry docks drain water seeping from the walls (table 3; fig. 6A) into the pump well and out to Sinclair Inlet through the drainage system (fig. 6B). When Sinclair Inlet seawater is circulated through a nuclear ship in dock as non-contact cooling water, the water, called auxiliary seawater, also is collected in the wet well and pumped back out to Sinclair Inlet (fig. 6B). In a separate system called the process water collection system (fig. 6B), floor troughs along the east and west walls of the dry docks drain water that accumulates in the dry docks themselves, including precipitation, process water during ship maintenance activities, and floor seep water. This process water is passed through sand traps before typically being discharged to the City of Bremerton, although on rare high flow or turbidity occasions (see fig. 6B) it is discharged directly to Sinclair Inlet.

Table 3.    

Total mercury concentrations in the dry dock systems, Bremerton Naval Complex and Sinclair Inlet, Washington, November 2018.

[Abbreviations: μS/cm, microsiemens per centimeter at 25 degrees Celsius; mg/L, milligram per liter; ng/L, nanogram per liter; mg/kg, milligram per kilogram dry weight. –, not analyzed; Dates are presented as YYYY-MM-DD, year, month, day]

Table 3.    Total mercury concentrations in the dry dock systems, Bremerton Naval Complex and Sinclair Inlet, Washington, November 2018.

USGS station ID	Sample name	Number of samples	Specific conductance (μS/cm)	Calculated percentage of groundwater	Total suspended solids (mg/L)	Filtered total mercury (ng/L)	Particulate total mercury (ng/L)	Whole total mercury (ng/L)	Total mercury of solids (mg/kg)
Dry Dock 6, sampled 2018-11-20
473316122383301	Pump well	6	37,700 1(37,000–38,000)	19	1.27 (0.99–1.58)	2.36 (1.30–5.31)	3.08 (1.96–5.69)	5.43 (3.26–11.0)	2.47 (1.45–4.66)
473316122383302	Cooling water intake	1	46,500	0	1.83	0.33	3.99	4.32	8.31
473316122383303	Auxillary seawater	1	44,000	5	1.71	0.66	1.47	2.13	3.60
473316122383305	Groundwater seep– northwest	1	9,850	79	12.85	8.73	5.60	14.3	19.9
473316122383306	Groundwater seep– southwest	1	30,600	34	0.3	3.25	0.529	3.78	4.31
473316122383307	Groundwater seep– northeast	1	11,300	76	11.76	35.1	16.4	51.5	67.9
473316122383309	Groundwater seep– mideast	1	13,800	70	0.17	2.71	0.434	3.14	3.58
473316122383308	Groundwater seep–southeast	1	15,500	67	0.15	1.87	0.427	2.30	2.72
473316122383310	Floor seep	1	–	–	0.14	2.15	0.420	2.57	2.99
473316122383304	Process water	1	27,900	40	9.01	3.55	4.49	8.04	12.5
Dry Docks 1–5, sampled 2018-11-27
473339122380002; 473339122380003	Pump well (Dry Dock 4)	2	31,200; 32,200	32	5.44; 3.82	3.06; 4.09	9.07; 9.82	12.9; 13.2	1.67; 2.57
473338122375801	Drainage water (Dry Docks 1–3)	1	28,500	39	1.45	0.86	30.5	31.4	61.9
473336122381403	Groundwater seep–northwest (Dry Dock 5)	1	16,300	65	0.17	5.69	1.72	7.4	9.1
473336122381408	Groundwater seep–midwest (Dry Dock 5)	1	32,200	31	0.27	51.5	2.16	53.7	55.8
473336122381404	Groundwater seep–southwest (Dry Dock 5)	1	37,400	20	4.72	1.02	2.87	3.9	6.8
473336122381405	Groundwater seep–northeast (Dry Dock 5)	1	19,900	57	0.19	4.45	5.49	9.9	15.4
473336122381407	Groundwater seep–mideast (Dry Dock 5)	1	42,700	8	0.48	3.56	1.32	4.9	6.2
473336122381406	Groundwater seep–southeast (Dry Dock 5)	1	39,100	16	0.6	3.04	2.14	5.2	7.3
473341122374701	Process water (Dry Dock 3)	1	15,000	68	3.46	1.00	2.29	3.29	5.58
473336122381402	Process water (Dry Dock 5)	1	21,100	55	2.6	1.65	3.51	5.16	8.67

¹

Average (range in parentheses).

Grab samples for specific conductance, TSS, and mercury (table 3) were collected into polyethylene terephthalate glycol (PETG) bottles from wall seeps of the drainage relief tunnels in Dry Docks 5 and 6, one floor seep in Dry Dock 6, the pump wells for Dry Docks 4 and 6, the cooling water intake and auxiliary seawater for Dry Dock 6, process water from Dry Docks 3, 5, and 6 (all shown in fig. 6) and drainage water from Dry Dock 1+2+3 (not shown in fig. 6). Samples were stored on ice until further processing at the USGS Washington Water Science Center (WAWSC) in Tacoma, Washington (see section, “Field Processing”).

Task 3.—Thermal Surveys

Thermal surveys were conducted along the western unwalled shoreline of the BNC from Charleston Beach to the edge of the seawall to identify cold-water locations that may indicate groundwater discharge locations for potential contaminant sampling in the task 6 spatial survey (fig. 7). The OU A part of the unwalled shoreline was surveyed with a georeferenced fiber-optic distributed temperature sensing (FO-DTS) cable from July 30 to August 7, 2020 (fig. 7A). This data collection window targeted daytime negative low tides coincident with high daytime air temperatures (up to 28 degrees Celsius [°C] on August 4, 2020) to maximize temperature differentials between cold discharging groundwater and warm receiving surface water. The daytime negative low tides ranged from approximately −3 to −5 feet (ft) NAVD 88. Elevations are presented in ft NAVD 88 in this report, where elevations in Mean Lower Low Water are elevations in NAVD 88 plus approximately 2.5 ft. Temperature was recorded every 15 minutes at 1.01 m spacing along the 1,500 m Sensornet Oryx DTS FO-DTS cable. Cable deployment configuration was an out-and-back U-shape with approximately 2 m distance between the parallel cable lengths. The cable was geolocated with a global positioning system every 45 m to maintain constant elevation during cable deployment. The elevations were approximately −2.5 to −7.5 ft NAVD 88 on the outbound cable and approximately −3.5 to −8.5 ft NAVD 88 on the return cable.

The FO-DTS could not be deployed in the western OUBT portion because of shipyard piers and spill-containment buoys. Instead, during a negative low tide on August 18, 2020, a georeferenced temperature survey with a calibrated multi-parameter sonde (600XLM, YSI, Inc.) was conducted on foot from the westernmost pier (Mooring G) to the edge of the seawall (fig. 7B). Approximately 150 survey points and nearbed water temperature measurements at 10-foot spacings were made at elevations ranging from approximately −2.5 to −6 ft NAVD 88. Additional method details, the original data files, and quality assurance and control procedures for both thermal surveys are available in Opatz and others (2023).

Task 6.—Spatial Survey of Trace Elements and Other Parameters along Unwalled Shorelines of Operable Unit A and Western Operable Unit B-Terrestrial, 2021

A spatial survey of nearshore groundwater, pore water, overlying surface water, seeps, and sediment along the unwalled shorelines of OU A and western OUBT was conducted in 2021 to better understand the potential transport of terrestrial contaminants to Sinclair Inlet via groundwater. The sampling windows targeted maximum groundwater discharge to the nearshore, presumed to be during the negative low tides, and occurred on March 30–April 1, April 26–29, May 24–27, June 8, and June 10, 2021. The groundwater locations included the nine wells in OU A and western OUBT that were instrumented for time-series data collection during the same time period (MW204, MW203, MW206, MW241, MW720, MW721, MW715, MW722, and MW718; table 4; fig. 4). The nearshore locations included 10 locations identified during the thermal surveys as cold-water discharge locations or adjacent non-cold-water locations selected for comparison purposes (locations 1, 2, 3, 5, 6, 7, 10, 13, 14, and 17; table 5; figs. 7A and 7B). A porewater sample was collected at each of the 10 nearshore sites and, in some cases, additional samples, including overlying surface water, seeps, additional pore water, and sediment, also were collected (table 5). One nearshore location and an adjacent groundwater well were concurrently sampled on each day over 11 days plus pre-sampling in-field method testing and 2 days of follow-up repeat sampling. In total, 49 water samples and 5 sediment samples were collected from the 19 locations (tables 4 and 5), including approximately 10 percent quality-control samples (see section, “Quality Assurance and Control”).

Table 5.    

Summary of nearshore sampling locations used in the spatial survey of trace elements and other parameters along unwalled shorelines, Sinclair Inlet, Washington, 2021.

[Locations of nearshore sampling sites are shown in figures 7A and 7B. Sediment samples were collected on May 25–26, 2021. Abbreviations: ID, identification number; USGS; U.S. Geological Survey; ft, feet; NAD 83, North American Datum of 1983; NAVD 88, North American Vertical Datum of 1988. Water sample collection date: YEAR-MONTH-DAY. Type of sample collected: PW, pore water; SW, surface water; FD, field duplicate; EB, pore water equipment blank; Sed, surface sediment; --, no sample collected; near, closer to shore than the primary sampling location; far, farther offshore than the primary sampling location; deep, pore water collected 3 feet below the sediment-water interface (as compared to 8 inches for the primary pore-water samples)]

Table 5.    Summary of nearshore sampling locations used in the spatial survey of trace elements and other parameters along unwalled shorelines, Sinclair Inlet, Washington, 2021.

Nearshore sampling site ID	USGS station ID	Short name	Latitude	Longitude	Sediment-water interface elevation	Water sample collection date	Type of sample collected
			(NAD 83)	(NAD 83)	(ft, NAVD 88)		PW	Seep	SW	Sed	Others
1	473320122390601	Sinclair Inlet nearshore 1 (24N/01E -2211 TW-T05)	47.55605	−122.6518	−5.38; −7.78; -12.05	2021-05-26	PW	Seep	SW	Sed	PW near; PW far
2	473321122390801	Sinclair Inlet nearshore 2	47.55582	−122.6525	−6.02	2021-05-24	PW	--	--	Sed	--
3	473317122390801	Sinclair Inlet nearshore 3 near MW715	47.55498	−122.6525	−7.08	2021-04-28	PW	Seep	--	--	PW FD
5	473316122391401	Sinclair Inlet nearshore 5 near MW721	47.5546	−122.654	−6.95	2021-04-27	PW	Seep	--	--	PW FD
6	473316122391801	Sinclair Inlet nearshore 6	47.55459	−122.655	−7.53	2021-04-26	PW	--	SW	-	SW FD; EB
7	473315122392301	Sinclair Inlet nearshore 7 near MW241	47.55437	−122.6565	−6.68; −7.58; -9.98	2021-04-29	PW	Seep	--	Sed	PW near; PW far; Sed near; Sed near FD
10	473313122392701	Sinclair Inlet nearshore 10 near MW206	47.55363	−122.6576	−10.55	2021-04-01	PW	--	SW	--	--
13	473309122393001	Sinclair Inlet nearshore 13 near MW203	47.55259	−122.6586	−7.17	2021-05-25	PW	--	--	--	--
14	473308122393401	Sinclair Inlet nearshore 14 near MW204	47.55227	−122.6596	−9.39	2021-03-31	PW	--	--	--	--
17	473304122393901	Sinclair Inlet nearshore 17 near Charleston Beach	47.55126	−122.661	−4.76	2021-03-30	PW	--	--	--	PW deep

At each groundwater well, two water-level measurements were taken upon arrival, then two sampling ports were deployed—one approximately 1 ft below the water surface (to sample a freshwater layer, if present) and the other approximately 2 ft above the bottom of the well (to sample the seawater layer, if present). The sampling ports were made of slotted Teflon attached to Teflon tubing that was attached to peristaltic pump tubing on the land surface. Well water was purged through both ports at a low-flow rate [approximately 100 milliliters per minute (mL/min) to purge 4 to 11 liter (L) total from each port]. While purging, field parameters of pH, oxidation-reduction potential, water temperature, dissolved oxygen, specific conductance, and turbidity were monitored using a calibrated multi-parameter sonde (In Situ AquaTroll 600, Fort Collins, Colorado) until readings stabilized (consecutive readings were within 3 percent of each other).

Samples for chemical analysis were collected by low-flow peristaltic pump (approximately 100 mL/min) to reduce the potential to induce flow from other locations within the sand pack surrounding the well screen. Unfiltered samples were collected per USGS field collection protocols (U.S. Geological Survey, 2002, 2006; Lewis and Brigham, 2004) for field sulfide and total iron (Chemetrics, Midland, Virginia), radon, trace elements (arsenic, cadmium, chromium, copper, lead, manganese, nickel, and zinc), including a separate PETG bottle for mercury, and TSS. Filtered samples were collected through a pre-conditioned 0.45-µm filter into bottles for trace elements (arsenic, cadmium, chromium, copper, lead, manganese, nickel, and zinc); cations (calcium, magnesium, sodium, iron, manganese, and potassium); anions (chloride, silica, sulfate, and fluoride); alkalinity, DOC, and stable isotopes of water. Stable isotopes of hydrogen and oxygen in water samples can be used to better understand groundwater/surface-water interactions, such as the relative proportion of a water sample comprised of groundwater versus surface water. Results are reported as the relative difference between isotopes—δ(²H/¹H) and δ(¹⁸O/¹⁶O) in parts per thousand (or per mil), where a negative value indicates that the sample is depleted in the heavy isotope relative to the international measurement standard (Révész and Coplen, 2008a, 2008b).

Concurrent with groundwater sampling, nearshore sampling occurred via a boat and floating pontoon. At each candidate groundwater discharge location (figs. 7A and 7B), hydraulic gradient was repeatedly measured in the 2 hours leading up to low tide using a manometer board to confirm groundwater discharge to the nearshore (that is, a positive hydraulic gradient) per USGS methods (Rosenberry and LaBaugh, 2008). Once a positive hydraulic gradient was confirmed, a custom direct-push stainless steel pore-water sampler (fig. 8) was installed approximately 8 in. below the sediment-water interface within a target bed-surface elevation range of 0 ±10 ft NAVD 88. The pore-water sampling depth of 8 in. below the sediment-water interface targeted existing forms of mercury (and other parameters) presumably sourced from the terrestrial environment rather than from overlying water. The target vertical elevation band encompassed the elevations of cold-water locations identified during the thermal surveys (generally from 0 to −10 ft NAVD 88) and historical sediment trace-element concentration exceedances in intertidal sediments near Charleston Beach (generally from −2.5 to +8.5 ft NAVD 88; CH2M, 2020). The exact location of each sampling location was determined by substrate suitability—a cobble- and boulder-free area of mostly silt and fine sediment was needed to push the pore-water sampler to the desired depth and attain a surface seal with a compressible foam seal to reduce potential inflow of surface water at the sediment-water interface. The horizontal coordinates and vertical elevation of each nearshore location were determined using a Trimble R8 GPS receiver with a Trimble survey controller handheld device. Data were collected using the real-time kinematic method. Real time corrections were obtained using a MiFi jetpack to access the Washington State Reference Network Puget Region SubNet per Rydlund and Densmore (2012).

Pore water was pumped through Teflon tubing inserted into the stainless-steel sampler at a low flow rate (less than 100 mL/min) to purge (less than 500 milliliters [mL]), measure field parameters, and collect water samples for chemical analysis. Unfiltered water samples were collected for radon and mercury. Filtered water samples were collected through a pre-conditioned 0.45-µm filter for trace elements, cation, anions, alkalinity, DOC, and stable isotopes of water. At three nearshore locations (nearshore locations 1, 6, and 10), samples of overlying surface water also were collected via peristaltic pump. Opportunistic samples of visible seeps were collected at four locations—at the edge of the seawall and under each of the three piers (nearshore locations 1, 3, 5, and 7). At two locations (nearshore locations 1 and 7), two additional porewater samples were collected along a transect perpendicular to the shoreline (targeting −10 ft and +10 ft NAVD 88). At one location (nearshore location 17), a deep pore water sample was collected from 3 ft below the sediment-water interface in addition to the 8-inch-deep sample. Five sediment samples were collected by a hand-push corer into perfluoroalkoxy (PFA) jars for total and methylmercury analysis, WhirlPak bags for analysis of grain size and organic carbon, and Whirl-Pak bags for further processing at the WAWSC laboratory before analysis for trace elements and ions. Due to the coarse nearshore substrate, sediment samples were only able to be collected at nearshore locations 1, 2, and 7.

Field Processing

During the nearshore spatial survey, field processing per USGS methods (U.S. Geological Survey, 2018) occurred in a mobile laboratory at the field site. Alkalinity was determined using the inflection-point method. Bottles for radon were shipped overnight to the USGS National Water Quality Laboratory (NWQL) for immediate analysis. Filtered and unfiltered water samples for trace elements analysis were acidified to pH <2 with nitric acid. Filtered samples for DOC were acidified to pH <2 with sulfuric acid. The sample bottles were stored in sealed bags on ice and transported to the WAWSC for any additional processing before shipping.

For dry dock sampling, tidal studies, and spatial survey, unfiltered water samples for mercury were processed at the WAWSC in a laminar-flow hood for mercury (Lewis and Brigham, 2004). Briefly, the samples were filtered through a filter held in a Savillex PFA filtering tower assembly into a 250- or 500-mL PFA or PETG bottle for FTHg or FMeHg analysis. The filtrate was 2 percent acidified with 6 molar hydrochloric acid within 6 hours of sample collection. The filters were transferred to PFA petri dishes and stored frozen for PTHg analysis. The process was repeated for FMeHg and PMeHg when applicable. Bottles were stored in the dark at room temperature and shipped to the USGS MRL for analysis.

The TSS was measured for dry dock sampling, tidal studies, and the spatial survey at the WAWSC. An aliquot was filtered gravimetrically through a 0.4-µm pore-size, 47-millimeter (mm) diameter, Nuclepore polycarbonate filter and determined to a precision of 0.001 milligram (mg), (Huffman and others, 2012).

For the nearshore survey, Whirl-pak bags of sediment for grain size and organic carbon analysis were shipped to the USGS Pacific Coastal Marine Science Center (Santa Cruz, CA). Sediment samples for analysis of trace elements and ions (arsenic, cadmium, calcium, chromium, copper, iron, lead, magnesium, manganese, nickel, potassium, sodium, and zinc) were dried at the WAWSC before shipping to AGAT Laboratories, Calgary, Canada. Samples of filtered water were shipped to the USGS Reston Stable Isotope Laboratory for analysis of stable isotopes of oxygen and hydrogen in the water molecule.

Analytical Methods

At the MRL, FTHg was determined using U.S. Environmental Protection Agency (EPA) method 1631, revision E (Environmental Protection Agency, 2002) that includes oxidation, purge and trap, desorption, and cold vapor atomic fluorescence spectrometry. FMeHg was measured using a modified version of EPA method 1630 (Environmental Protection Agency, 1998), which included distillation pre-treatment followed by ethylation, gas chromatography separation, and analysis by isotope dilution inductively coupled plasma mass spectrometry (ICP-MS; Jackson and others, 2009). The method detection limits for FTHg and FMeHg were approximately 0.05 and 0.04 ng/L, respectively.

Filters for PTHg analysis were prepared by oxidation with 5-percent bromine monochloride (weight per volume percentage) and heated for 5 days at 50°C (Olund and others, 2004) prior to analysis using cold vapor atomic fluorescence spectrometry (Environmental Protection Agency, 2002). Filters for PMeHg were prepared by distillation coupled to ethylation, gas chromatography separation, and ICP-MS detection (DeWild and others, 2004; Jackson and others, 2009). The method reporting limit was approximately 0.05 ng mercury per filter for PTHg and 0.04 ng methylmercury per filter for PMeHg.

STHg was determined by atomic adsorption following direct combustion (Environmental Protection Agency, 1998). SMeHg was determined by distillation, gas chromatography separation, and speciated isotope dilution mass spectrometry (Jackson and others, 2009). Sediment and particulate samples for mercury stable isotopes were digested in concentrated aqua regia (3:1 hydrochloric and nitric acids) for 8 hours at 90°C. After digestion, samples were filtered to remove any remaining particulate material and diluted to a final acid content of 10 percent (weight per volume).

Mercury stable isotope analysis was performed using a multicollector ICP-MS using operating protocols outlined elsewhere (Yin and others, 2016; Janssen and others, 2019). Mercury stable isotope values are reported in δ notation (for example, δ²⁰²Hg) for mass dependent fractionation and Δ notation (for example, Δ¹⁹⁹Hg or Δ²⁰⁰Hg) for mass independent fractionation following standard nomenclature (Blum and Bergquist, 2007).

At the NWQL, ²²²Radon was determined by Standard Method D5072-98 (American Society for Testing and Materials, 1998). Dissolved organic carbon was determined by Standard Method 5310B (American Public Health Association, 1992). Major ions were determined by ion chromatography or inductively coupled plasma chromatography (Fishman and Friedman, 1989). Trace elements were determined by ICP-MS (Garbarino and others, 2006).

Stable isotopes of water were determined at the USGS Reston Stable Isotope Laboratory in Reston, Virginia, using an isotope-ratio mass spectrometer (Révész and Coplen, 2008a, 2008b).

Sediment samples were analyzed for grain-size analysis and carbon contents at the USGS Pacific Coastal and Marine Science Center sediment laboratory and for elements at AGAT Laboratories as described in Opatz and others (2023).

Mercury Concentrations and Isotopes in Sediments

Mercury concentrations in archived sediments from Sinclair Inlet (“SI” in table 2 and figure 5, including samples from both OU B Marine and outer Sinclair Inlet) ranged from 272 to 1,170 ng/g dw (dry weight) (median = 758 ng/g dw). The mercury concentrations measured in the three Puget Sound representative bays (fig. 5) were from 30.1 to 176 ng/g dw (“Bay” sites in table 2), reflecting lower mercury concentrations than “SI” samples and samples collected on the terrestrial BNC (“BNC” sites in table 2). Total mercury concentrations in particulate material from Dry Docks 5 and 6 were variable, reaching values as low as 16 ng/g dw. Total mercury concentrations were highest in particulate material collected from the bottom of monitoring wells MW722, MW412, and MW709 (18,000–44,000 ng/g dw, table 2).

Methylmercury in sediments represented less than 1 percent of the total mercury pool present in OUB Marine and Sinclair Inlet sediments. Methylmercury concentrations reached a high of 5.32 ng/g in marine sediments at PSEMP LT-40030 and 142 ng/g in particulate material from well MW412. Methylmercury concentrations generally were high when total mercury concentrations were high in sediments and particulate material (table 2). Methylmercury concentrations in Puget Sound representative bays were lower than hotspots within the terrestrial BNC, but methylmercury made up a larger percentage of the total mercury ranging from 1.6 to 2.2 percent. All samples were candidates for mercury stable isotope analysis with the exception of two low mercury concentration samples from the southwestern corner of Dry Dock 6 and sediment from Holmes Harbor.

Mercury stable isotopes measured across all sediments and particulate material ranged from −0.87 to 0.18 ‰ for δ²⁰²Hg and −0.04 to 0.19 ‰ for Δ¹⁹⁹Hg (table 2; fig. 9). No discernible Δ²⁰⁰Hg was measured in samples (Opatz and others, 2023). Isotope values were similar between Sinclair Inlet samples (average δ²⁰²Hg = −0.48 ± 0.07 ‰ and average Δ¹⁹⁹Hg = 0.03 ± 0.04 ‰, 1 standard deviation), including OU B Marine and outer sampling locations, and samples collected at other Puget Sound Inlets (table 2; fig. 9). Archived sediments from the terrestrial BNC (average δ²⁰²Hg = −0.45 ± 0.03 ‰ and average Δ¹⁹⁹Hg = 0.03 ± 0.03 ‰) were similar to Sinclair Inlet and Puget Sound representative bays, but δ²⁰²Hg outliers were measured in particulate material collected from Dry Dock 6 and monitoring well MW412. Isotope values of δ²⁰²Hg from Dry Dock 6 were 0.42 ‰ lower than the average values for the Sinclair Inlet and Puget Sound Bays whereas material from monitoring well MW412 was 0.63 ‰ higher. These deviations from the uniform δ²⁰²Hg values observed at all other sites may indicate a different Hg source. The sample collected from Dry Dock 6 was from the southeastern floor drain (“DD6 SE trough”), which collects all particulates washed off the dry dock and may not be representative of the legacy contamination measured in the sediments. Particulates from well MW412 correspond to the highest mercury concentrations observed in these samples (table 2) and may relate to the original (unprocessed) source material or a different Hg source on-site that was not represented in the sediment samples (fig. 10).

Mercury stable isotope values (δ²⁰²Hg and Δ¹⁹⁹Hg) collected from this study overlap with literature measurements of contaminated sediments, including those collected from sites of metallic mercury catalyst usage, metal refining, mercury and gold mining, and paper mill activity (Eckley and others, 2020; fig. 9). The sediments or particulate material at BNC, Sinclair Inlet, or the Puget Sound representative bays do not overlap with most of literature values absent of direct point source inputs (termed Reference-No point source; fig. 9), typically regions outside of direct points sources fall below a δ²⁰²Hg of −1 and have a larger range of Δ¹⁹⁹Hg. This comparison indicates that mercury in BNC, Sinclair Inlet, and the larger Puget Sound is derived from industrial discharges of mercury. Despite the wide range of mercury concentrations measured in this study, the isotope values for δ²⁰²Hg were nearly constant, with the exception of Dry Dock 6 SE Trough and well MW412, suggesting that an industrial mercury profile is present in the region (fig. 10). These data demonstrate that mercury from industrial or urban discharges dominate the region and little mercury input is received from overland runoff (wetland or forested regions) or direct deposition (atmospheric precipitation). Lastly, low Δ¹⁹⁹Hg values (average across all samples 0.04 ±0.05, 1 standard deviation) indicate that little photochemical processing occurred within the samples, with the exception of particulates from Dry Dock 5 (0.19 and 0.12‰) and Dry Dock 6 (0.13 ‰) and support the conclusion that little post-processing occurred between suspended particulate material release and sedimentation.

Mercury in the Dry Dock System

WTHg concentrations in Dry Dock 6 system in the pump well ranged from 3.26 to 11.0 ng/L (table 3). The specific conductance (37,000–38,000 µS/cm) indicated a mix of mostly seawater with some groundwater (approximately 19 percent). Percentage of groundwater was calculated assuming a two end-member mixing model, with seawater specific conductance of 46,500 µS/cm (Cooling Water Intake value) and a groundwater specific conductance of 500 µS/cm (for example, MW206, see table 6). Mercury concentrations were low (WTHg <5 ng/L) in the seawater component (Cooling Water Intake and Auxillary Seawater), which is consistent with previous measurements of Sinclair Inlet seawater (1.09 ng/L, Strivens and others, 2018). Mercury concentrations also were low (WTHg <5 ng/L) in four of the six sampled seeps in Dry Dock 6. The two seeps with elevated mercury concentrations were in the northwest (WTHg = 14.3 ng/L) and northeast (WTHg = 51.5 ng/L) sections of the drainage relief tunnels. TSS also were elevated in these samples (about 12 mg/L) and specific conductance was about 10,000 µS/cm, indicating primarily (>75 percent) groundwater composition.

WTHg concentrations in the Dry Dock 1-5 system in Dry Dock 4 pump well were 12.9 and 13.2 ng/L, mostly bound to particulates (table 3). Particulate mercury was elevated (30.5 ng/L) in a sample of drainage water from Dry Docks 1–3 and additional sampling is needed to identify if there are sources of mercury-laden particulates in Dry Docks 1, 2, or 3. Mercury was elevated (WTHg = 53.7 ng/L) in one seep on the mid-west side of Dry Dock 5 mostly in the filtrate. Mercury concentrations were low (WTHg from 3.29 to 8.04 ng/L) in the three samples from the Process Water system.

Cold-Water Locations in the Nearshore Identified by Thermal Surveys

The two nearshore thermal surveys—FO-DTS survey and sonde walk—identified approximately 10–15 cold-water locations as candidate sampling locations for the 2021 nearshore spatial survey (figs. 7A and 7B). Adjacent non-cold-water locations also were identified for comparison (figs. 7A and 7B). FO-DTS data from August 1 through 4, 2020, were visualized using the USGS DTSGUI-v1.0.0 (https://code.usgs.gov/water/espd/hgb/DTSGUI/tree/v1.0.0) in figure 11. The x-axis represents distance along the cable; here the cable entered the water near the western-most pier (Mooring G) at 55 m and ran south along the OU A shoreline to Charleston Beach at 550 m. The y-axis represents time with temperature data collected every 15 minutes. This 4-day window captured extreme negative low tides (for example, approximately −5 ft NAVD 88 at 1145 on August 4, 2020) coincident with high daytime air temperatures (for example, up to 28 °C) on August 4, 2020, to maximize temperature differentials between cold discharging groundwater and warm receiving surface water. Vertical dark blue bands in figure 11 indicate potential groundwater discharge locations as persistent cold temperatures regardless of tide or time of day. This also is indicated on the bottom bar (temperature statistics) by areas with low maximum water temperatures and low standard deviations. For example, a zone of at least three cold-water locations was indicated along a protected, north-facing shoreline adjacent from well MW206 (fig. 7A—Location 10 and two additional unnamed candidate cold-water locations). Well MW206 (the red square on the parking lot inset in fig. 11) was the only nearshore well that contained fresh water/low specific conductance. There may be subsurface structures, as well as the protected north-facing shoreline orientation, limiting inland seawater movement in this area, resulting in groundwater discharge. A visible seep in the rip rap along this section was noted during cable deployment in July and retrieval in August.

Other candidate cold-water locations identified by the FO-DTS survey included two broad areas on either side of the outfall pipe (near the USGS marine logger) in the inlet just north of Charleston Beach, one zone along a steep rip-rap ledge just before the cable exited the water by the truck inspection gate, and a few other locations where cold water was more prominent in the deeper cable (approximately −8.5 ft NAVD 88) than the shallower cable (approximately −4.5 ft NAVD 88; fig. 7A). In total, eight candidate cold-water locations were identified along the OU A shoreline and two additional non-cold water locations were identified for comparison purposes (fig. 7A).

The sonde walk during low tide on August 18, 2020, identified cold-water locations along the western OUBT shoreline including under piers (fig. 7B, for example, Locations 3 and 5). Two visible seeps were noted—one at the edge of the seawall and one under Mooring E. In total, five candidate cold-water locations were identified along the western OUBT shoreline and two additional non-cold water locations were identified for comparison purposes (fig. 7B).

The shoreline along OU A was steeper with large riprap, other than the section near Charleston Beach, as compared to the western OUBT. However, the cable was able to be at similar vertical elevations along OU A (approximately −2.5 to −8.5 ft NAVD 88) as compared to the shoreline along OUBT (−1 to −6 ft NAVD 88).

Tidal Fluctuations of Water Level, Temperature, and Specific Conductance in Nearshore Wells

The 11 monitored nearshore wells were approximately 10 (MW720) to 150 ft (MW412) from the shoreline (table 4; fig. 4). The screened intervals were 9.5 to 20 ft long (table 4) and generally encompassed the range of Sinclair Inlet water-level elevations—the bottom of the screened interval ranged from −4 to −14 ft NAVD 88; the top ranged from approximately 0 to +14 ft NAVD 88; Sinclair Inlet water-level elevations ranged from approximately −6 to +12 ft NAVD 88 (table 4).

Site 1 Wells

The two monitored wells near Site 1 soils had varying characteristics despite being in close proximity (450 ft from each other and similar distances from the shoreline). Water levels in well MW709 fluctuated up to 5 ft and the mean level was centered on the mean water level in Puget Sound (fig. 12A). In contrast, water-level fluctuations in well MW412 were smaller (up to 3 ft) and the water level was approximately 3 ft lower than the water level in well MW709 (fig. 12A). One hypothesis is that well MW412 is within the capture zone of the Dry Dock 6 pump system, while well MW709 is outside of dry dock influence and may directly exchange with Sinclair Inlet along the unwalled shoreline. This hypothesis is consistent with subsurface particle tracking model results (see fig. 3B, modified from Jones and others [2016]).

The time-series water temperature data (fig. 12B) also indicate different connectivity with Sinclair Inlet between the two wells. The water temperature near the bottom of well MW709 had a tidal signal and ranged from 13 to 16 °C over the 8 months of monitoring. The water temperature near the bottom of well MW412 had no tidal signal and was colder than well MW709, ranging from 11 to 14.5 °C, especially in the summer months, indicative of groundwater influence. Further, there was a surface freshwater lens in well MW412 during the winter months, coincident with precipitation recharge events (fig. 12C) whereas the water in well MW709 was seawater year-round (collected over the same period but was outside of acceptable criteria for publication).

Western BNC Wells

The nine nearshore OU A and western OUBT monitoring wells also were tidally influenced, although to varying degrees not directly related to distance from shoreline. Maximum sub-daily water-level fluctuations ranged from 3.89 to 11.2 ft (table 6; fig. 13; app. 2). There appeared to be high hydraulic connectivity between Sinclair Inlet and the monitoring wells based on tidal oscillations in the wells that were typically less than 4 hours behind Sinclair Inlet tidal oscillations. Well MW722 appeared the most connected to Sinclair Inlet as indicated by a large water-level elevation range (table 6; fig. 13; app. 2), seawater composition (fig. 14; app. 3), a large seasonal range of temperatures that tracked with Sinclair Inlet (highest temperatures in late summer and lowest annual temperatures in late winter) and sub-daily temperature tidal oscillations (fig. 15). MW720 also appeared well connected to Sinclair Inlet with a large tidal signal in water-level elevation and temperature (table 6; figs. 13–15; app. 2–3).

Table 6.    

Ranges of water elevation, specific conductance, and water temperature in nine nearshore monitoring wells, western Bremerton Naval Complex, Washington, 2020–21.

[Location of monitoring wells are shown in figure 4. Abbreviations: ft, feet; NAVD 88, North American Vertical Datum of 1988; μS/cm, microsiemens per centimeter at 25°C; °C, degrees Celsius; YYYY-MM-DD, year, month, day]

Table 6.    Ranges of water elevation, specific conductance, and water temperature in nine nearshore monitoring wells, western Bremerton Naval Complex, Washington, 2020–21.

Site short name	Period of time- series record	Water level elevation (ft above NAVD 88)		Water-level range (ft)	Specific conductance (μS/cm)		Water temperature (°C)
		Minimum	Maximum		Minimum	Maximum	Minimum	Maximum
Marine logger	2020-06-05 to 2021-10-15	−5.98	12.21	18.19	–	–	7.1	24.6
Operable Unit A
MW204	2020-06-05 to 2021-10-15	3.05	9.85	6.80	88	37,500	3.1	17.9
MW203	2020-06-02 to 2021-11-03	2.63	110.65	18.02	13	34,400	0.5	20.9
MW206	2020-06-02 to 2021-11-03	3.95	110.56	16.61	10	620	0.9	20.1
MW241	2020-06-02 to 2021-11-03	4.44	9.66	5.22	33,500	41,500	10.4	17.2
Operable Unit B—Terrestrial
MW720	2020-06-02 to 2021-11-04	0.82	11.21	10.4	4,660	35,800	9.1	17.9
MW721	2020-06-02 to 2021-11-04	3.41	7.3	3.89	806	2,100	12.8	18.8
MW715	2020-06-02 to 2021-11-04	2.62	7.29	4.67	987	22,800	12.8	16.3
MW722	2020-06-02 to 2021-11-03	0.79	12.0	11.2	17,500	44,300	9.4	20.6
MW718	2020-06-02 to 2021-11-03	4.42	8.92	4.5	165	1,370	13.2	20.3

¹

Impacted by surface runoff directly into well.

In addition to MW722 and MW720, four other wells contained seawater, with specific conductance values ranging from 22,800 to 44,300 µS/cm—MW715, MW241, MW204, and MW203 (table 6; fig. 14; app. 3). Three wells—MW718, MW721, and MW206—were either brackish or freshwater (specific conductance <2,500 µS/cm) despite their proximity to the shoreline. MW718 is the farthest inland well and was selected, in part, to be representative of an upland freshwater well. MW206 is a similar distance from the shoreline as adjacent seawater wells (MW203 and MW204), so there may be a physical subsurface barrier at that location limiting seawater mixing in the nearshore. The water temperature time series supported this—the temperature in MW206 was not tidally driven and instead remained constant at around 13 °C other than precipitation-derived pulses owing to the nature of the well construction that allowed parking lot runoff into the well during the period of monitoring (fig. 15). MW203 and MW206 were impacted by surface runoff directly into the well owing to the vented cap during the data-collection period, evident in all three time-series parameters (fig. 15; app. 2–3).

Mooring infrastructure also may be buffering seawater exchange, which may explain the brackish instead of seawater conditions in MW721. Further, there was a more delayed seasonal temperature lag in MW241, MW721, and MW715, each of which are behind moorings, as compared to MW720 and MW722. However, MW241 was full seawater despite the long seasonal temperature lag and minimal sub-daily tidal variation. Another anomaly was MW206, which had a large tidal influence in the water-level elevation (about 6 ft range) but was freshwater (low specific conductance and a constant temperature year-round).

The time-series data for these nearshore wells indicate that the conditions in these wells are spatially heterogeneous and represent rapidly changing conditions in nearshore tidally influenced artificial fill rather than steady-state groundwater conditions. The groundwater/surface-water mixing zone along the unwalled shorelines of the BNC extends at least 100 ft inland, likely farther inland in some places (estimated up to 700 ft inland in some areas during the Remedial Investigation). Conditions are highly variable spatially and temporally owing to tidal exchange, seasonal precipitation, and surface and subsurface infrastructure such as moorings and buried material that may limit seawater connectivity or may provide preferential flow paths from the BNC to Sinclair Inlet.

Mercury Concentrations in Site 1 Monitoring Wells over Spring and Neap Tidal Cycles, 2018–19

The water in the monitoring wells in Site 1 soils (MW709 and MW412) was seawater during the 2018–19 tidal studies, indicating that the groundwater/surface-water mixing zone extends inland at least 150 ft along this unwalled shoreline. This is consistent with the time-series results (see fig. 12 and section, “Tidal Fluctuations of Water Level, Temperature, and Specific Conductance in Nearshore Wells”). The near-surface and near-bottom water in MW709 was seawater throughout the 25-hour spring and neap tidal studies (fig. 16A–16B; table 7). There was fresh groundwater in the near-surface samples of MW412, especially during spring and neap ebb tides when the Sinclair Inlet water level declined below the water level in the well (figs. 16C–16D). MW709 is located on the Site 1 outcropping while MW412 is located approximately 450 ft west of MW709 and may be more buffered from direct seawater exchange because of its proximity to the Z Street pier. MW412 also appears to be within the pumping zone of Dry Dock 6 (fig. 3), which may explain the consistently lower water level and smaller water-level fluctuation in MW412 as compared to MW709 (figs. 12, 16A–16D).

Concentrations of WTHg in MW709 generally were less than 50 ng/L during the tidal studies except for spikes during ebb tides up to 965 ng/L (figs. 17A–17B; table 7). The spikes at the beginning and end of each study were mostly due to an increase in PTHg concentrations (for example, PTHg = 948 ng/L or 98 percent of the WTHg concentration in the first near-bottom sample during the spring tidal study). The high concentrations in the samples near the beginning of each study may have been impacted by particulates stirred up from the bottom of the well during purging and sampler installation. However, elevated PTHg concentrations (250 and 72.6 ng/L) also were measured in the near-bottom final spring and neap samples, respectively, when the sampling port had not moved during the 25 hours of sampling. Further, elevated WTHg concentrations up to 212 ng/L were measured during ebb tide in the middle of the spring study, primarily as FTHg (174 ng/L). Other than this spike for a few hours during the spring ebb tide, concentrations of FTHg in MW709 were less than 40 ng/L and typically 15–25 ng/L throughout spring and neap tidal studies.

It is hypothesized that during large negative ebb tides, inland seawater containing elevated mercury extracted from contaminated soils drains past MW709 to Sinclair Inlet. This tidally driven change in subsurface hydraulic gradient also may stir up contaminated particulates in the bottom of the well. This resulted in elevated mercury in the near surface samples primarily as FTHg and elevated mercury in the near bottom of the well primarily as PTHg. The high FTHg spike in the near surface samples occurred during a 15 ft tidal differential, when Sinclair Inlet water levels were approximately 10 ft lower than the water level in the well during maximum ebb (fig. 17A). This high-mercury draining phenomenon did not appear to occur during the neap tidal study, which included an 8-ft tidal differential when the Sinclair Inlet water levels were approximately 4 ft lower than the water level in the well during ebb (fig. 17B; FTHg <40 ng/L). The maximum FTHg concentration in MW709 during the July 2018 ebb (174 ng/L) was less than one-half of the maximum measured during the ebb of the November 2011 tidal study (496 ng/L; Paulson and others, 2013).

Mercury concentrations and partition between filtered and particulate phases in MW412 differed from that in MW709 during the tidal studies. WTHg concentrations in MW412 were very high (figs. 17C–D; table 8). WTHg concentrations in the near surface samples ranged from 238 to 1,040 ng/L and 653 to 1,910 ng/L during the spring and neap tidal studies, respectively. These high concentrations were mostly in the filtered form (percent PTHg of WTHg <15 percent). WTHg concentrations were lower in the near-bottom samples than in the near-surface samples, ranging from 21 to 351 and 13 to 96 ng/L during the spring and neap tidal studies, respectively, with larger contributions from the particulate fraction (percent PTHg of WTHg ranged from 18 to 89 percent). The surface concentrations were high throughout both tidal studies, regardless of tidal stage. Soil boring analysis during the Remedial Investigation identified high-mercury soils in the upper layers at MW412 (up to 75.4 mg/kg at a depth interval of 5–7 “ft,” the units presumed to be “ft below ground surface”; U.S. Navy, 2002). Mercury concentrations were low in deep soil layers (less than 0.5 mg/kg at depth intervals from 10 to 27 ft). This high mercury band appears to be approximately 8–10 ft vertically higher than the water level in the well during 2018–19 but may be indicative of high mercury shallow soils near MW412 that are the source of the high-mercury near-surface water in MW412.

Trace Elements and Other Parameters Along Unwalled Shorelines, 2021

During negative low tides in spring 2021, 58 environmental and quality-control samples were collected from groundwater wells, seeps, nearshore pore water, surface water, and sediment for a large suite of field parameters, stable isotopes of water, major ions, and trace elements to better understand groundwater/surface-water interactions along unwalled shorelines of the BNC, including the transport of terrestrial contaminants to Sinclair Inlet via direct discharge. Nearshore locations were selected for pore water sampling only if a positive hydraulic gradient was measured to indicate groundwater discharge to surface water in the nearshore. Nearshore hydraulic gradients were tidally driven and often only positive in the 1–2 hours prior to and including low tide. Hydraulic gradients along the western shoreline adjacent to OU A (fig. 7A, location numbers 10–17) were minimally positive, typically with 10 mm or less of positive hydraulic gradient. Hydraulic gradients along the northern shoreline adjacent to western OUBT (fig. 7B, location numbers 1–7) were more strongly positive, including under the piers, ranging from +10 to +165 mm. The largest hydraulic gradient was measured at nearshore location 2 (+165 mm). The high hydraulic gradients under the piers and in nearshore location 2 suggest high permeability and hydraulic conductivity in these nearshore areas allowing marine water to infiltrate into the terrestrial soils during high tides and then subsequently drain as a mixture of marine water and groundwater from the terrestrial soils to Sinclair Inlet during low tides.

At the targeted nearshore elevation (final sampling elevations ranged from −4 to −12 ft NAVD 88), the seabed substrate typically was composed of sands underlying varying amounts of gravel, cobble, and riprap material. At locations with a steep shoreline and cobble and riprap substrate (common along the OU A shoreline), the pore water sampler was installed in a sandy spot between cobbles to ensure a flat seal on the sediment-water interface. Pore water sampler installation was easier in locations with a gradual slope and sandy substrate, including at Charleston Beach (Location 17) and under the piers (Locations 3, 5, and 7). Location 2 contained silts and clays, and Location 1 had unconsolidated gravel over silts and clays. The custom pore water sampler (fig. 8) was successfully installed and used for pore water sampling at 10 nearshore locations of the BNC shoreline.

Characterization of Groundwater and Surface-Water End Members

Field and laboratory parameters indicated a range of conditions in the nearshore samples (tables 9–10). Two wells—MW206 and MW718—were used as groundwater end members characterized by low (more negative) values of stable isotopes of water, low specific conductance (<1,000 μS/cm), low DOC, low concentrations of major ions, including calcium, magnesium, potassium, sodium, chloride, fluoride, manganese, and sulfate, and high silica concentrations (complete results available at U.S. Geological Survey, 2022). The stable isotope ratios of these two wells are consistent with regional Puget Sound groundwater stable isotope ratios (fig. 18; table 4.1). The three surface-water samples were used as seawater end members and were characterized by high (least negative) values of stable isotopes of water, high specific conductance, high concentrations of major ions, and low silica concentrations. The other water samples—pore water, seeps, and the remaining groundwater well samples—were a mix of these two end members. Stable isotopes of water were used to define a two end-member mixing model (fig. 18) and calculate the percentage of groundwater in each sample (fig. 19). Specific conductance, silica, and sulfate correlated well to the stable isotope model and also could have been used. The percentage of groundwater was calculated as:

$$Percentage of groundwater=\frac{({\delta }^{18}O sample)-({\delta }^{18}O surface water)}{({\delta }^{18}O groundwater)-({\delta }^{18}O surface water)}*100$$

,

(4)

per Phillips and Gregg (2001), using δ¹⁸O values (tables 9–10), where

δ¹⁸O sample

is the ¹⁸O ratio in the sample, per mil;

δ¹⁸O surface water

is the ¹⁸O ratio in the surface-water end member, the average of the three surface-water samples, per mil; and

δ¹⁸O groundwater

is the ¹⁸O ratio in the groundwater end member, the average of the four fresh groundwater samples, per mil.

One of the two groundwater end member wells—MW718—is the farthest inland well (approximately 130 ft from the shoreline) and had been selected for the study because it was previously known to be upland of the saltwater mixing zone and could potentially serve as an inland reference well. It is unclear why the other groundwater end member well—MW206—does not have salt water because it is one of the closest wells to the shoreline (about 30 ft, tables 2 and 6). The other OU A wells—MW204, MW203, and MW241—were primarily seawater (17.6–37.5 percent groundwater) as was MW722 (23.3 and 25.7 percent groundwater). In contrast, MW720, MW721, and MW715 were majority groundwater (52.5–78.5 percent groundwater).

Many of the pore water and seep samples resembled seawater, with less than 2 percent groundwater contribution (table 10; fig. 19). The samples with the greatest groundwater contribution (>10 percent) included Seep 3, Seep 5, PW17 deep, and PW 7 near. Samples with 2–10 percent groundwater included Seep 1, PW17, PW7 far, and PW1 near. All other samples had less than 2 percent groundwater contribution. The areas with highest groundwater discharge, therefore, appear to be under the piers, at the edge of the seawall, and at Charleston Beach. These are areas where cold water or seeps were found during the thermal studies, positive hydraulic gradients, or high percent groundwater contribution based on stable isotopes of water data. It is hypothesized that pier and seawall infrastructure or the fill material in which they are located may provide preferential flow paths between the terrestrial subsurface and Sinclair Inlet.

Concentrations of Trace Elements in Water and Sediment Samples

Concentrations of trace elements in many water samples were low or non-detectable (tables 11–12). Arsenic was the most frequently detected trace element in the spatial survey (fig. 20). Arsenic was detected in unfiltered groundwater samples at concentrations greater than the Criteria/Remediation Goal of 5 μg/L in MW203, MW241, and MW722. Concentrations were highest in MW203 (approximately 50 μg/L), primarily in the dissolved or colloidal fraction (<0.45 um). Arsenic also was elevated in three nearshore Location 7 filtered seep and pore water samples (23.7–37.2 μg/L; table 11). The two Location 7 sediment samples exceeded Washington State Marine Sediment Cleanup Objectives and Screening Levels (Washington State Department of Ecology, 2013) for arsenic (521–553 µg/g; table 13). Although sediment samples were not collected from under the other two piers, arsenic was detected in unfiltered seep samples from both, including one elevated concentration (Location 5 unfiltered seep = 27.2 µg/L). The particulates in these unfiltered seep samples may represent resuspended fine surface sediment. The deep pore water sample at Charleston Beach (Location 17) was slightly elevated for arsenic (6.4 µg/L). Arsenic concentrations were low (<5 µg/L) in many other groundwater and nearshore water samples. Arsenic concentrations were low at the other two sediment samples (Locations 1 and 2) although “Sed 1 near” (28.9 μg/g) was greater than the Puget Sound Natural Background level (11 μg/g). The results suggest that there is an arsenic hotspot near MW241 and nearshore Location 7 (fig. 20) where arsenic concentrations are elevated in groundwater, seep, pore water, and sediment samples. Offshore of MW203, which had the highest groundwater concentrations of arsenic, the nearest porewater sample (PW13) was less than detection for arsenic (seep and sediment samples were not collected).

Table 13.    

Field and laboratory parameters, and trace elements from sediment samples from the nearshore spatial survey, Bremerton Naval Complex, Washington, 2021.

[Abbreviations: ng/g, nanogram per gram; μg/g, microgram per gram; μm, micrometer; <, less than; %, percent; –, not applicable. Inorganic results are presented as dry weight concentrations. Exceedances in bold. Natural background and cleanup criterion: Washington State Department of Ecology, 2013; complete results available in Opatz and others (2023)]

Table 13.    Field and laboratory parameters, and trace elements from sediment samples from the nearshore spatial survey, Bremerton Naval Complex, Washington, 2021.

Natural background, cleanup criterion, or sample short name	Percent mud (<63 μm)	Percent total organic carbon	Total mercury (ng/g)	Methyl- mercury (ng/g)	Percent methyl of total mercury	Arsenic (μg/g)	Cadmium (μg/g)	Chromium (μg/g)	Copper (μg/g)	Iron (%)	Nickel (μg/g)	Lead (μg/g)	Zinc (μg/g)
Puget Sound Natural Background	–	–	200	–	–	11	0.8	62	45	–	50	21	93
Washington State Marine Sediment Cleanup Objective	–	–	410	–	–	57	5.1	260	390	–	–	–	410
Washington State Marine Sediment Cleanup Screening Level	–	–	590	–	–	93	6.7	270	390	–	–	–	960
Sed7 near	4.36	0.29	230	0.28	0.12	524	0.95	148	939	9.10	98.4	539	4,820
Sed7 near, field duplicate	3.94	0.93	280	0.17	0.06	521	0.88	158	783	11.20	101	614	5,610
Sed7	2.21	0.17	330	0.17	0.05	553	1.03	180	1,120	7.56	140	592	3,090
Sed2	27.12	1.47	840	0.55	0.07	3.1	1.22	27.7	71.3	2.99	42.1	412	252
Sed1 near	6.18	1.18	1,300	0.99	0.08	28.9	0.44	59.3	206	5.94	44.9	137	650

Cadmium was detected in a single nearshore sample (unfiltered Seep 5 = 1.03 µg/L) and was not detected or detected at low levels (<2.5 µg/L) in the well samples (table 11). Sediment cadmium concentrations were low (<1.5 µg/g; table 13).

Copper exceeded the Remediation Goal or Compliance Criteria in MW204, MW720, and MW722 (17.3–60.0 µg/L; table 11) and was primarily in the dissolved or colloidal form (<0.45 µm) in groundwater. Copper was not detected in any nearshore filtered water samples but was detected in nearshore sediment and particulate samples (fig. 21). Copper was measurable in the two unfiltered seep samples (Seep 5 = 106 µg/L; Seep 3 = 11.2 µg/L). Because copper was not detected in the corresponding filtered pore water sample, the copper in these unfiltered pore water samples is hypothesized to be particulate-bound to nearshore sediment particulates in the unfiltered sample rather than from a terrestrial source. Sediments at nearshore Location 7 exceeded Washington State Marine Sediment Cleanup Objectives and Screening Levels (Washington State Department of Ecology, 2013) for copper (783–1,120 µg/g; table 13). Copper concentrations were greater than Puget Sound Natural Background in sediments at Locations 1 and 2. The results indicate the copper is present in the dissolved or colloidal form in terrestrial samples (nearshore monitoring wells) but in the particulate form in nearshore samples (unfiltered seeps and sediment near Moorings G and F). It is unknown if this is due to different sources of copper in the terrestrial versus nearshore marine environments, or because of partition and sorption affinity to particulates during transport from the terrestrial to the marine environment.

Lead concentrations were not detected or low (<2 µg/L) in all groundwater, seep, pore water, and surface-water samples (table 11) with the exception of the two unfiltered seep samples (Seep 5 = 72.1 μg/L; Seep 3 = 8.68 µg/L). Lead was detected in the five sediment samples ranging from 137 µg/g at Location 1 near to 614 µg/g in the field duplicate of location Sed7 near, all of which exceed Puget Sound Natural Background of 21 μg/g (table 13).

Nickel was detected slightly greater than the Remediation Goal or Compliance Criteria in MW204 and MW241 (surface sample only, table 11). Nickel was detected in one filtered nearshore sample (PW17 deep = 8.1 µg/L). The highest concentration was in an unfiltered seep sample (Seep 5 = 53.1 µg/L). Nickel was detected in the five sediment samples ranging from 42.1 µg/g at Sed2 to 140 µg/g at Sed7 (table 13). The Sed7 samples exceeded Puget Sound Natural Background of 50 μg/g.

Zinc was detected at concentrations greater than the Remediation Goal or Compliance Criteria in MW204, MW241 (surface sample only), MW720, and MW722 (table 11). Zinc was detected in four nearshore samples, including the highest measured concentrations (filtered water from PW6 = 934 µg/L and unfiltered water from Seep 5 = 609 µg/L). Blank results indicated no contamination of zinc or other trace elements due to sampling equipment. Zinc was detected in the five sediment samples at concentrations greater than Puget Sound Natural Background, including exceedances at nearshore Location 7 of Washington State Marine Sediment Cleanup Objectives and Screening Levels (3,090–5,610 µg/g; table 13; Washington State Department of Ecology, 2013). The results suggest a zinc hotspot near Mooring G with elevated concentrations in groundwater, seeps, pore water, and sediment (fig. 22).

Whole total mercury (WTHg) exceeded the OUBT Compliance Criteria of 25 ng/L in MW720 (54.8–55.2 ng/L), MW718 (93.1 ng/L in the bottom sample), and MW722 (1,020–1,030 ng/L). Total mercury in the other wells ranged from 0.411 to 12.3 ng/L (table 12). In contrast to the trace elements above, in which the “total” or whole-water chemical concentration was determined by the analysis of an unfiltered water sample, whole-water total mercury (WTHg) was calculated as the sum of FTHg and PTHg, where FTHg is all chemical forms of mercury in the filtrate passing through a 0.45-µm nominal pore size filter and PTHg is defined as all chemical forms of mercury captured on the filter. Owing to the low analytical detection levels, FTHg was detected in all nearshore water samples (except PW7 near), ranging from 0.21 ng/L in a surface-water sample to 153 ng/L in Seep 1. Both nearshore samples that exceeded the 25 ng/L groundwater criteria were at Location 1 (Seep 1 = 153 ng/L FTHg; SW1 = 67.6 WTHg; table 12). Total mercury was detected in the five sediment samples, including exceedances at Locations 1 and 2 (840–1,300 ng/g; table 13). The current study’s mercury results provide additional data that there is a mercury hotspot adjacent to Site 2 soils in the nearshore area by Location 1 at the edge of the seawall (fig. 23) where there were elevated concentrations in well water, seep, porewater, surface water, and sediment samples. Two additional samples in the vicinity of nearshore Location 1 had elevated mercury—Seep 3 (13 ng/L) and sediment at Location 2 (840 ng/g).

Methylmercury was not detected along OU A (wells or nearshore samples) and was frequently detected in OUBT (table 12). Methylmercury concentrations in OUBT wells ranged from <0.04 to 0.58 ng/L (filtered) and <0.008 to 0.217 ng/L (particulate), with the highest concentration in the bottom of MW715 (WMeHg = 0.797 ng/L), which had strongly reducing conditions (−188 millivolt oxidation-reduction potential) and the highest well DOC concentrations (3.30 and 5.30 mg/L), iron concentrations (1170 and 2110 mg/L), and manganese concentrations (541 and 584 mg/L). Methylmercury was detected in all four seep samples (filtered = 0.06–0.22 ng/L) and in pore water at nearshore Locations 1, 2, and 5 (filtered = 0.04–0.42 ng/L). Methylmercury also was detected in the surface-water sample at Location 1 (WMeHg = 0.371 ng/L). Filtered methylmercury contributed between 0.01 (MW722 bottom) and 37 percent (PW5) of the filtered total mercury concentration. Previous investigations have found that biogeochemical characteristics determine the presence of and levels of methylmercury, including DOC concentrations, oxidation-reduction conditions, presence of terminal electron acceptors such as iron and sulfate, and the capability of the microbial community to methylate by containing methylation genes (Paulson and others, 2018; Bravo and Cosio, 2020). Methylmercury was detected in the five nearshore sediment samples ranging from 0.17 ng/g in Sed7 to 0.99 ng/g in Sed1 near, representing less than 0.2 percent of the total mercury sediment concentration (table 13).

Stormwater

The data from Conn and others (2018) indicate that mercury in PSNS015 is associated with tidal flushing, not with fresh stormwater. The single high mercury concentration in a freshwater sample from 2009 was not confirmed in the follow-up 2011–12 sampling. The additional data indicate that the FTHg estimate of 119 grams per year (g/yr) by Paulson and others (2013) from the PSNS015 outfall is a maximum load estimate; updated estimates are approximately 1–77 g/yr (table 15; app. 5). Adding the paired PTHg data, WTHg load estimates ranged from 0.5 to 2.2 g/yr for the fresh stormwater component and approximately 1–230 g/yr from tidal flushing. There is still high uncertainty in these loading estimates owing to a lack of tidal flow data for PSNS015, and, also, as with any non-continuous sampling, uncertainty around whether these measured concentrations captured the full range of mercury concentrations in the storm-drain system that may be present under different tidal and seasonal conditions.

The remaining BNC stormwater outfalls were estimated in Paulson and others (2013) to have negligible mercury loads based on four samples. A 2010–13 Navy survey collected 2–11 samples from each of 13 of the BNC stormwater systems for a suite of metals including mercury (Brandenberger and others, 2018). Total recoverable mercury maximum concentrations exceeded 25 ng/L in 4 of the 13 systems, including PSNS015 (346 ng/L) and maximum concentrations from 26.5 to 80.6 ng/L in the other 3 systems. These maximum concentrations are higher than the data in Paulson and others (2013). Mercury monitoring, including particulate mercury and bi-directional flow data for all 24 of the stormwater outfalls, is needed. That, in combination with periodic sampling from seaward vaults during ebb tides, could reduce uncertainty in these loading estimates.

Dry Docks

The measured FTHg concentrations in the pump wells for Dry Docks 1–5 and 6 were similar to those reported in Paulson and others (2013; table 15; app. 5). The average annual discharge estimate for Dry Docks 1–5 provided by the U.S. Navy from January 2019 through December 2021 (Duy Pham, U.S. Navy, written commun., June 10, 2022) was the same as the value used in Paulson and others (2013)—13,800 cubic meter per day (m³/d, or 3.21 million gallons per day (Mgal/d), range=1.31–7.25 Mgal/d). Therefore, the estimated FTHg loads are similar: 3.2–26 g/yr (Paulson and others, 2013) versus 16–21 g/yr (this study). PTHg and WTHg load estimates were not calculated in Paulson and others (2013) and are estimated in this study to range from 46 to 50 and 65 to 67 g/yr, respectively (table 15; app. 5). This suggests that most mercury discharged to Sinclair Inlet from Dry Docks 1–5 is in particulate form although there is high uncertainty because this is based on only two particulate samples. There were two samples with elevated mercury from the Dry Docks 1–5 system—drainage water from Dry Docks 1–3 with elevated particulate mercury (PTHg = 31.4 ng/L) and a wall seep sample along the midwest wall with elevated filtered total mercury (FTHg = 51.5 ng/L). Mercury was low (WTHg <10 ng/L) in the remaining samples contributing to the Dry Docks 1–5 pump well Additional sampling for filtered and particulate total mercury in various locations, including repeat sampling of the drainage water from Dry Docks 1–3 and new and different wall seeps contributing to Dry Docks 1–5, could refine these values.

Recent annual average discharge estimates provided by the Navy for Dry Dock 6 were differentiated between two conditions—with cooling water (recirculated Sinclair Inlet seawater) and without cooling water. The “without cooling water” discharge estimates for both periods (January 2017 through November 2018 and January 2019 through December 2021) were similar to the value used by Paulson and others (2013)—19,000, 18,100, and 17,300 m³/d, respectively (table 14), or approximately 4.5 Mgal/d (2019–21 range = 3.6–5.9 Mgal/d). From 2017 and 2021, Dry Dock 6 was operating under the “without cooling water” condition approximately 52 percent of the time (Duy Pham, U.S. Navy, written commun., July 1, 2022). The other 48 percent of the time it was operating under “with cooling water” condition, which approximately doubles the discharge (2019–21 average = 34,600 m³/d or 9.1 Mgal/d, range = 6.6–15.2 Mgal/d). All dry dock pump well samples were collected when cooling water was contributing, therefore load estimates were calculated with the “with cooling water” discharge value. Updated FTHg loads were higher than those in Paulson and others (2013), 16–67 g/yr in this study versus 8.2–14 g/yr in Paulson and others (2013), owing to a slightly higher range of measured mercury concentrations and the higher discharge value. PTHg and WTHg load estimates were not calculated in Paulson and others (2013) and are estimated in this study to range from 25 to 72 g/yr and 41–140 g/yr, respectively (table 15; app. 5). The data indicate that filtered and particulate forms of total mercury are contributing to the WTHg load discharged from Dry Dock 6 to Sinclair Inlet. There were two wall seeps from Dry Dock 6 with elevated mercury in both filtered and particulate forms—a northeast seep (WTHg = 51.5 ng/L) and a northwest seep (WTHg = 14.3 ng/L). Mercury was low (WTHg <10 ng/L) in the remaining samples contributing to the Dry Dock 6 pump well.

Approximately 10–30 percent of the total mercury from Dry Dock 6 is recirculated Sinclair Inlet seawater, estimated at WTHg = 13–26 g/yr, based on measured concentrations from the Cooling Water Intake and Auxillary Seawater (table 3) and the difference between the discharge with and without cooling water (16,400 m³/d). Therefore, the Dry Dock 6 whole total mercury load estimates from the BNC are approximately 28–112 g/yr (table 15; app. 5).

The two seep samples with elevated WTHg greater than 50 ng/L were from tunnels adjacent to the area between Dry Docks 6 and 5. The elevated mercury concentrations in the northeast seep of Dry Dock 6 and midwest seep of Dry Dock 5 may have originated from contaminated soils/fill material in the area between Dry Docks 6 and 5 and were transported via groundwater to the dry dock pump systems. A sample of drainage water from Dry Docks 1–3 had elevated particulate mercury (30.5 ng/L) and additional sampling is needed to identify if there are sources of mercury-laden particulates in Dry Docks 1, 2, or 3 that are contributing to the Dry Dock 1–5 mercury load to Sinclair Inlet. No other mercury pathways were identified from the dry dock component sampling. Process water load estimates were low, ranging from 0.2 to 0.7 g/yr for FTHg and 0.7 to 2 g/yr for WTHg (app. 5).

These updated estimates suggest that the Dry Dock systems (WTHg about 94–180 g/yr) may be a larger pathway of mercury from the BNC to Sinclair Inlet than previously thought (Paulson and others, 2012; 2013) and may be on the same order of magnitude as the PSNS015 storm-drain system (WTHg about 1–270 g/yr). Similar to the storm-drain system, there is uncertainty with the dry dock loading estimates because the mercury concentrations and discharge rates will vary depending on how many and the types of ship in dock and special servicing and repair activities.

Direct Discharge Along Shorelines

A key assumption during remedial investigations was that chemical concentrations in groundwater decrease to ambient seawater concentrations by the time groundwater reaches the discharge point to Sinclair Inlet surface water. In the current study, elevated levels of some trace elements, such as arsenic, zinc, and mercury, were measured in nearshore monitoring wells, shoreline seeps, and marine porewater above Remediation Goals or Compliance Criteria for groundwater. These samples were collected from within the dynamic groundwater-marine water nearshore mixing zone (fig. 24) where fresh groundwater discharge (FGD) interacts with both a shallow intertidal saltwater cell (ISC) and a deep saltwater wedge (DSW) and the shape of the mixing zone varies under different tidal and precipitation conditions. The BNC nearshore monitoring wells rarely represented fresh groundwater, and, more typically, represented saline water within the groundwater-marine water nearshore mixing zone. Marine water preferentially extracts some chemicals such as mercury from subsurface materials to salt water through complexation to chloride ions (Grassi and Netti, 2000). On the BNC, concentrations of arsenic, copper, zinc, and mercury exceeded criteria, and were 2–50 times higher in a saltwater well (MW722) as compared to a nearby freshwater well (MW718), both of which are near Site 2 soils. When exceedances of criteria occurred in monitoring wells, it was within saltwater wells. Concentrations of trace elements in the fresh and brackish wells (MW206, MW718, and MW721) were low, with the exception of PTHg in the bottom of MW718 (80.7 ng/L) and are the best indicators of nearshore groundwater concentrations from the study. Concentrations of trace elements in the Sinclair Inlet seawater samples also were low.

Therefore, for some chemicals, such as mercury, the major transport mechanism in the nearshore likely is ISC rather than FGD (fig. 24). During each high tide, the ISC infiltrates nearshore contaminated subsurface materials, extracts contaminants, and carries them back to Sinclair Inlet on the subsequent low tide. Concurrently, fresh groundwater discharge generally is moving from areas of higher hydraulic head (uplands) to areas of lower hydraulic head (Sinclair Inlet), per steady-state model simulations (for example, Jones and others, 2016). As the fresh groundwater passes through contaminated subsurface materials, it also extracts contaminants, albeit less efficiently than seawater for some chemicals, and the contaminants are further diluted with clean seawater during transport to Sinclair Inlet. For example, at the mercury hotspot near the edge of the seawall, WTHg concentrations decreased from greater than 1,000 ng/L in well MW722 to 153 ng/L in Seep 1 to 1–11 ng/L in nearshore pore water.

In shorelines where direct discharge is occurring, it is difficult to estimate chemical loads because the discharge estimates need to account for both the FGD and discharge associated with the ISC (fig. 24). A steady-state groundwater model was updated for the BNC (Jones and others, 2016) and, while it does not account for the ISC or other tidal dynamics, provides the best estimates of steady-state groundwater discharge. Total direct discharge (FGD + ISC) can be estimated with manual field measurements of seepage over tidal cycles, coupled with geochemical tracers of groundwater and seawater. For example, Rosenberry and others (2013) measured highly variable seepage rates in a Puget Sound intertidal zone ranging from −217 cm/d during flood tides to +300 cm/d during ebb tides. The authors concluded that “volumes of water and associated nutrients and chemicals that flow across the sediment-water interface each day are much larger than previously thought [Simonds and others, 2008], resulting in a substantially larger mixing volume in the pore waters beneath the broad intertidal zone that covers more than 4.6 km² of bed in Lynch Cove.”

Manual field seepage measurements have not been made in the nearshore of the BNC but may by as highly variable over tidal cycles as those measured in nearby Lynch Cove given the coarse substrate along the BNC shoreline. To estimate trace element fluxes to Sinclair Inlet, an estimate of the ISC volume discharged to Sinclair Inlet during ebb tides is needed (that is, volume per ebb × 706 ebb per year). It is not appropriate to use a net discharge value (that is, subtract the flood tide volume) because when multiplied by a chemical concentration the load would falsely indicate that chemicals are being loaded back into the nearshore during each flood tide, when instead they are diluted in Sinclair Inlet and each flood tide contains relatively clean seawater. The groundwater discharge estimates from Jones and others (2024), therefore, represent a minimum possible discharge. In the absence of measured discharge data, an initial estimate of the range of possible volumes discharged during ebbs along the unwalled shoreline from Charleston Beach to the edge of the seawall ranges from essentially no discharge where seepage rates are close to 0 cm/d to more than 11,000 m³/d, or more than 30 times the estimate of steady-state groundwater discharge for that shoreline (336 m³/d; Jones and others, 2024). This range is based on approximately 1,000 m of shoreline length, approximately 15 m of intertidal height and linearly variable seepage rates from 0 cm/d during slack tide to 300 cm/d during max ebb tide. A hypothetical exercise using variable rates from 0 to a maximum of 35 cm/d during max ebb (within the average measured range in nearby Hood Canal; Rosenberry and others, 2013) results in discharges, and therefore, chemical loads, nearly four times higher than those determined using a net value (table 16).

Estimated mercury loads (table 15) and other trace element loads (table 17) were estimated using the Jones and others (2024) discharge as the minimum and four times the Jones and others (2024) discharge as an estimated maximum. There is very large uncertainty with these numbers—they should be considered rough estimates for demonstration purposes only. Actual site-specific fluxes could be determined with manual field measurements such as those used in Simonds and others (2008) and Rosenberry and others (2013) coupled with a quantitative determination of intertidal area along the unwalled BNC shoreline. Uncertainty would still exist, as it is difficult to adequately characterize the temporal variability in the flux rates.

Using these highly uncertain estimates based on the described assumptions, the volume of direct discharge along unwalled shorelines (<0.5 Mgal/d) is minor compared to stormwater (about 1.3 Mgal/d), Dry Docks 1–5 (about 3.5 Mgal/d), and Dry Dock 6 without cooling water (about 4.8 Mgal/d). Based on the concentration results from the select nearshore sampling locations in this study, multiplied by the estimated discharges, mercury loads (table 15, app. 5) from direct discharges (WTHg about 12–90 g/yr) are estimated to be third largest after stormwater (WTHg about 1–270 g/yr) or the dry dock systems (WTHg about 94–180 g/yr). Loads of other trace elements via direct discharge generally are an order of magnitude higher than mercury loads because concentrations are an order of magnitude higher. For example, estimated loads along OU A and western OUBT are 860–3,460 g/yr of filtered arsenic; 1,230–5,000 g/yr of filtered copper; 760–3,100 g/yr of filtered nickel; and 16,200–67,000 g/yr of filtered zinc (table 17).

Data Gaps

Some nearshore locations warrant additional study to better quantify the spatial and temporal variability of discharge and contaminant concentrations. These are:

These areas have elevated concentrations of one or more trace elements and evidence of high discharge (presence of seeps and large measured hydraulic gradients discharging to Sinclair Inlet). Current concentrations of trace elements other than mercury are unknown near Site 1 soils, such as in MW709, MW412, or the adjacent nearshore environment. The extent of the contaminated sediment (and fill material) in the nearshore is unknown (distance inland, depth of contaminated sediment, spatial heterogeneity). Recirculating seawater through contaminated sediments, resuspension of nearshore bed sediment, and physical sloughing of shoreline material all may contribute to elevated trace-elements concentrations in these nearshore areas. The annual long-term monitoring sampling results indicate that all nearshore wells, even those behind seawalls, such as 392R, 380, and 410R, contained seawater (EA Engineering, Science, and Technology, Inc., PBC, 2022). Therefore, these groundwater/surface-water interaction discussions and implications may be relevant not just for unwalled shorelines, but also for walled shorelines along the BNC. Additional nearshore porewater and sediment sampling coupled with geophysical tools to better quantify site-specific tidal discharge and the extent of the ISC would help identify any other nearshore areas with high discharge and elevated trace-elements concentrations.

Other data gaps identified in this study include:

The data gaps that have the most impact on the Conceptual Site Model at this time likely are (1) seepage rates/discharge volumes for the intertidal saltwater cell, especially adjacent to Sites 1 and 2 soils, and (2) comprehensive stormwater monitoring of concentrations and bi-directional flows.