Attachment_2__PHASE_II_ANACOSTIA_SOP_V5_0001.pdf

July 27, 2020

U.S. Geological Survey Maryland Water Science Center Project Work Plan, Quality Assurance Project Plan, and

Standard Operating Procedures For

Anacostia River Tributary Sediment Study- Phase II

Timothy Wilson

Submitted to

Dev Murali, RPM District Department of Energy & Environment

1200 First Street NE, Washington DC 20002

Version 5: 7/27/20

Phase II Anacostia River Tributary Sampling QAPP/SOP 07/05/23

Contents

1.0 Introduction

1.1 Overview of Sampling Program

1.2 Sampling Schedule and Inventory

1.3 Storm-Sampling Criteria

1.4 Description of Samples and Data to be Collected

1.4.1 Discrete Samples for Sediment Characterization

1.4.2 Large-Volume Composite Samples

1.4.3 Low-Flow Samples

1.4.4 Metal Samples

1.4.5 Bed-Sediment Samples

1.5 Large-Volume Sampling in the Lower Anacostia River

1.6.1 Storm Sampling

1.6.2 Low-Flow Sampling

1.6.3 Sampling for Calibration of Turbidity Sensor

1.7 In-Situ Water-Quality Data

1.8 Target Analyte List

1.9 General Field Operations for Sampling

1.10 Hydrodynamic Measurements in the Lower Anacostia River

1.11 Project Organization and Communication

1.12 Project Responsibilities

1.13 Safety

1.14 Contact Information

Chapter 2. Overview of Sampling Methods and Quality Control Procedures

2.1 Description of Sampling Operations

2.2 Sampling Equipment

2.3 Equipment Cleaning

2.4 Station Set-up

2.5 Preparation of Automatic Sampling Equipment

2.6 Collection of Large Volume Samples

2.7 Collection of Low-Flow Samples

2.8 Water-Quality Monitoring

2.9 Sample and Data Retrieval

2.10 Sample Records

2.11 Large Volume Sample Processing

2.12 Preparation of Metal Samples

2.13 Sample Shipping

2.14 Data Reporting

2.15 Operation of Hydrodynamic Equipment

2.15.1 Equipment

2.15.2 Operating Parameters

2.15.3 Data Handling

2.16 Quality Assurance Procedures

2.16.1 Sample Collection

2.16.2 Blanks and other Quality Assurance Samples

2.17 Corrective Actions

2.18 References Cited

Chapter 3. Standard Operating Procedures

List of Tables

Table 1. Phase II sampling locations Table 1.1. Sampling site groups Table 1.2 Number and type of samples to be collected at each station Table 1.3. Transit and pumping rates for flow velocities.

Table 1.4. Target Analytes.

Table 2.1. Summary of Quality Control and Assurance Samples used in this study.

List of Figures

Figure 1. Location of tributaries and sampling points in the Anacostia River Watershed, Washington D.C.

Figure 2. Location of the USGS gage station at Buzzard Point, Washington D.C.

Figure 3. Schematic of equipment used to collect large-volume composite samples from the Lower Anacostia River.

Figure 4A. Map view of areas for study as identified in hydrodynamic model results Figure 4B. Longitudinal view of predicted sediment accumulation in the Lower Anacostia River Figure 5. Schematic of ADCP equipment platforms.

Chapter 3. Attachments SOP A. General setup of sampling station and equipment

A.1. General Layout and Equipment SOP B. Procedures to be used for cleaning equipment

B.1.0 Inlet tubing lines, ISCO pump and distributor arm tubing, peristaltic pump tubing, plastic connectors B.2.0 Filter plates, canisters, and associated transfer equipment B.3.0 POC filtering equipment B.4.0 Poly-Bottles used in Automatic samplers B.5.0 Large-volume sample container B.6.0 GFF Filters

SOP C. Procedures to calibrate Automatic samplers and Data Sondes C.1.0 Calibration of auto-samplers

C.2.0 Calibration of Water-Quality Data Sondes C.2.1 Turbidity C.2.2 Specific Conductance

C.3.0 Example Program for calibrating an autosampler SOP D. Procedures Used for Sample Collection

D.1.0 Site Setup D.2.0 Automatic Samplers D.2.1 Setup D. 2.2 Large-Volume Samples.

D.2.3 Discrete Samples for SS and POC.

D.2.4 Low-Flow Samples D.3.0 Sample and Data Retrieval

SOP E. Chain of Custody Forms

SOP F. Procedures for Processing Large-volume Samples F.1.0 General F.1.1 Filtering F.1.2 Labeling

SOP G. Procedure for preparing Suspended Sediment and Particulate Organic Carbon Samples G.1.0 General.

G.2.0 Discrete Samples for Particulate Organic Carbon (POC) analysis.

G.2.1 Sample Processing G.2.1.1. Equipment G.2.1.2 Filtration of POC samples

G.2.1.3 Sample Labeling.

G.2.1.4 Shipping for POC analysis G.3.0 Samples for Suspended Sediment Analysis.

G.3.1 Shipping

SOP H. Procedures for Preparing Field Equipment Blanks SOP I. Preparation of Laboratory Analytic Blanks and QA Samples

Anacostia River Tributary Sediment Study Phase II Workplan

1. Introduction

As part of the Anacostia River Remedial Investigation conducted by the Washington, D.C. Department of Energy & Environment (DOEE), the U.S. Geological Survey (USGS) has been tasked with sampling five major gaged tributaries ( the NE and NW Branch of the Anacostia, Lower Beaverdam Creek (LBDC), Watts Branch, Hickey Run), one minor non-gaged tributary (Nash Run), and the tidal Lower Anacostia River (at Buzzards Point). In addition, hydrodynamic data will be collected in the tidal Lower Anacostia River. This work represents the second phase of the Anacostia Tributary study and is built on the 2016-17 USGS study. Data generated in Phase II can be used by the DOEE to update the Anacostia River contaminant data base, will serve to demonstrate the effects on ongoing remediation efforts in the tributary watersheds, and can serve as a check on the results of the Lower River contaminant-loading/hydrodynamic model.

The 5 gaged tributaries being resampled was chosen because they represent the largest tributary loadings measured in the Phase I study. Nash Run is included because it provided the largest estimated loads of the non-gaged Anacostia tributaries within in Washington D.C. area. Additional focus is being placed on Lower Beaverdam Creek because of the short data record (monitoring of all types was initiated in 2016) and because Phase I data showed to be largest tributary source of contaminated sediment to the Lower Anacostia. The Lower Anacostia River sampling was included because, with the addition of a new USGS discharge and water quality (WQ) monitoring station at Buzzard Point, chemical data will provide insight into the mass of contaminated sediment moving from the Anacostia into the Potomac River system.

In this project large-volume (LV) samples will be collected during storms and low-flow conditions.

Suspended sediment collected will be analyzed for polychlorinated biphenyl (PCB) congeners, polychlorinated aromatic hydrocarbons (PAHs), organochlorine pesticides (OCPs) and tracem metals (total and dissolved fractions); samples for analysis of suspended sediment (SSC) and particulate organic carbon (POC) concentrations will be collected concurrently with the LV samples. The chemical data will be combined with continuous measurement of discharge and water-quality parameters at the permanent USGS gage stations. Trailer deployed equipment will be used to sample at Nash Run. The collection methods and analytic work used in the Phase II are consistent with the methods used in the 2016-17 study, allowing the results to be combined and compared with data from the Phase I study. Several refinements to the methods used in Phase I:

• Sediment-loaded filters will be freeze-dried before analysis, allowing a more accurate measure of sediment mass sent for analysis, and therefore, more accurate determination of contaminant concentrations. Removing water from the sediment also reduces interferences in the OCP analyses, providing lower detection levels for OCP compounds.

• Spiked sediment samples will be submitted for analysis, allowing a better estimate of recoveries of contaminants from sediment.

• A larger number of SSC and POC samples will be collected, and several sediment-loaded filters will be prepared for direct POC measurement. These steps will refine the SSC POC values and allow more accurate carbon-normalized concentrations calculated. .

• A procedure will be implemented to split up to LV samples collected at LBDC for PCB/PAH and possibly pesticide analysis. Phase I sampling demonstrated that sufficient sediment is obtained during storms to allow the loaded filters prepared from LV samples to be split, while still obtaining sufficiently low reporting limits for PCB/PAHs. Depending upon the analytic refinements made to pesticide analysis, split samples may be obtained for these compounds as well.

• Discharge in Nash Run will be measured during several storm events, thereby helping to refine discharge estimates for all small ungaged tributaries along the Anacostia River.

• The Lower Anacostia River at Buzzards Point will be sampled to help quantify the sediment and contaminant loading to the Potomac River, and ultimately to Chesapeake Bay. The sampling will measure the cross-channel variability in flow and sediment concentrations under varying discharges. This data will allow calculating SSC and sediment loads using the continuous turbidity currently being measured at Buzzards Point.

• Water velocity and other hydrodynamic data will be collected in the Lower Anacostia using acoustic Doppler velocity meters deployed on the riverbed. Velocity profiles near the riverbed allow shear-stresses to be calculated as a function of river conditions. These data will help in understanding the conditions under which bed sediment in the Lower Anacostia is eroded; and will help in refining the hydrodynamic model of the river.

This document presents the work plan and standard operating procedures (SOPs) to be used in the Phase II sampling; it may be updated as methods are refined during the study.

1.1 Overview of Sampling Program

Sampling will be conducted on five major gaged tributaries and one minor non-gaged tributary of the Anacostia River during storm and low-flow conditions. NEB, NWB, Watts Branch, Hickey Run, and Nash Run will be sampled during 1 storm event, LBDC and the Buzzard Point (BP) in the Lower Anacostia will be sampled during 3 storms. Figure 1 shows the tributary watersheds and the sampling locations; details are listed in Table 1. All tributaries will be sampled at least once under low flow conditions. Storm samples will be collected using automatic samplers to collect aliquots of river water. 2 large-volume (~60 L) composite samples of river water will be collected during each storm 1 sample is dedicated for PCB/PAH analysis and 1 for OCP analysis. Multiple discrete samples will be also be collected for SSC and POC analysis. LV samples will be returned to the MD Water Science Center where the suspended sediment will be removed by filtration. During low-flow sampling, sediment will be collected using stream-side filtration. The sediment-laden filters will be freeze-dried, then sent to a contract laboratory for PCBs, PAH and OCP analysis;

separate samples will be collected for analysis of total metals. River stage, discharge, turbidity, and other WQ parameters are currently being continuously measured at the sampling sites by the USGS.

Sampling on the Lower Anacostia will be conducted by watercraft traversing the river during rising, cresting and falling portion of the hydrograph. The boat will stop at a minimum of 10 equal-width intervals across the river. At each stop, equal volume aliquots will be collected for the LV composite sample by slowly lowering inlet lines attached to a weighted sampler slowly lowered to the riverbed, while river water is pumped into a collection vessel. The weighted sampler will be configured to collect a sample for SSC and POC analysis.

Buzzard Point

Fig. 1 Location of tributaries and sampling points in the Anacostia River Watershed, Washington D.C.

Table 1. Phase II sampling locations

Site name Site identifier

USGS Site

ID

Latitude Longitude Description

Major tributaries, gaged sites1 NE Branch Anacostia River, Riverdale MD NEB 01649500 38 o 57’ 36.9” 76 o 55’ 33.5” Presently gaged and monitored

NW Branch Anacostia River, Hyattsville, MD

NWB 01651000 38 o 57’ 08.4” 76 o 57’ 57.78” Presently gaged and monitored

Hickey Run HR 01651770 38 55’ 0.5” 76 58’ 09.2” Presently gaged and

Watts Branch WB 01651800 38 54’ 04.0” 76 56’ 31.9” Presently gaged and

Lower Beaverdam Creek LBDC 01651730 38 o 54’ 58.23” 76 o 55’ 55.26” Presently gaged and undergoing calibration

Lower Anacostia at

Buzzards Point LABP 01651827 38 o 51’ 54.6” 77 o 00’ 37.2” On right bank, 0.5 mi upstream from mouth.

Minor tributaries, un-gaged sites

Nash Run NR TBD 38o 54’ 36.4 “ 76 o 56’ 30.24” Anacostia Ave. NE, app.

300ft N of Polk St NE.

1.2 Sampling Schedule and Inventory

Sampling will begin in late June 2020 and will be conducted by grouping tributaries. Table 1.1 lists the order the tributaries will be sampled. As mentioned previously, NEB, NWB, Watts Branch, Hickey Run and Nash Run will be sampled during 1 storm event and once during low-flow; LBDC and LABP will be sampled during 3 storms and during 1 low-flow. The sites are grouped because of limited to equipment and manpower availability. Storms samples will be collected in late spring and early summer; one sample each will be collected at LBDC and LABP in late summer. Low-flow sampling will be conducted throughout summer when flows are normally the lowest and will be completed over approximately 3 weeks.

Table 1.2 lists the number and types of samples and blanks to be collected at each site. Each of the different type of samples are described in detail later in the workplan and are summarized in the table notes.

In addition to the LV samples for contaminant analysis, SSC (up to 12 samples) and POC (up to 12 samples) will be collected at each site during each sampling trip; the grain size distribution in two of the SSC samples will be determined. Bed-sediment samples will be collected at LBDC after each sampling event, and will be analyzed for PCB,PAH, OCP, Organic carbon and grain size. As described in Chapter 2, filter blanks (unused filters, handled as samples) and field equipment blanks (DI water passed through equipment and filters in the field) will be collected throughout the program.

Table 1.1. Sampling site groups

Sampling Group

Sites Type Season

1 NW, NEB, LBDC Storm Spring/early summer 2 WB, HR, NR Storm Spring or early summer 3 LBDC and LABP Storm Early summer 4 LBDC and LABP Storm Summer 5 LABP Storm Late summer 6 NWB, NEB, LBDC, WB, HR, LABP Low-flow Mid-summer

NW: Northwest Branch Anacostia NEB: Northeast Branch Anacostia WB: Watts Branch HR: Hickey Run LBDC: Lower Beaverdam Creek LABP: Lower Anacostia at Buzzard Point

Table 1.2. Number and type of samples to be collected at each station

Stream ID Storm samplesa

Low-flow samplesb

Bed sediment samples

Metal samplesg

Suspended sedimentc

Particulate organic carbonc

Grain size analysisc

NEB NEB 1 1 0 3 24 24 2

NWB NWB 1 1 0 3 24 24 2

Watts Branch WB 1 1 0 3 24 24 2 Hickey Run HR 1 1 0 3 24 24 2 Nash Run NR 1 -- 0 3 12 12 1 Lower Beaverdam Creek

LBDC 3 1 5 9 48 48 8

Lower Anacostia at Buzzard Point

LABP 3 1 0 4 66f 66f 8

Filter blanksd FB 1 1 0 0 0 0 0 Field Equipment blanksd

FEB 3 1 0 4 0 0 0

Field Duplicate or splite

LBDC 2 1 1 0 0 0 0

SRMh NIST 1944 3

Total Samples 20 9 6 32 222 222 25

The various sample groups listed in Table 1.2 are detailed later in this SOP. The table notations summarize the various groups:

a. Storm samples collected for chemical analysis include: 1 LV (60-70L) sample for PCB/PAH, 1 LV sample for pesticides, 1-4L sample for total metals. As described in note c, 12 discrete samples are also collected for SSC, and 12 discrete 1L samples processed for POC.

b. Low-flow samples include: 1 set of GFF filters for PCB/PAH and 1 set for pesticides; typically set has 6 filters, plus 2-1L samples for metal analysis (1 whole water, 1 filtered). As described in note c, up to 12 samples are collected for SSC and 12 samples collected for POC; these samples are collected whenever a new filter installed.

c. During the collection of each storm and low-flow chemical sample, numerous discrete samples will be collected: 12 discrete 1L samples for SSC analysis; 2 of which will also have the grain-size distribution measured. Additionally, up to 12 discrete samples will be collected and processed for POC analysis.

d. Filter blanks (FB) and field equipment blanks (FEB) developed in the project include:

1 FB will be archived for each group of sites sampled for storms.

1 FB will be submitted with group 1 samples. The remaining FBs will be archived.

1 FB will be submitted with all low-flow samples that are collected.

1 FEB will be produced for storm samples collected at group 1 sites (NW, NEB, LBDC, spring) 1 FEB will be produced for storm samples collected at group 2 sites (WB, HR, NR, spring) 1 FEB will be produced for storm samples collected at group 3 sites (LBDC and LABP summer and fall) 1 FEB will be produced for all low-flow samples collected at group 4 sites (collected in late spring).

e. Field duplicate or sample splits. If enough sediment was obtained in storm samples collected at

LDBC, then two storm samples will be split and treated as duplicates. A minimum of 10 grams of sediment will be submitted in each split. One duplicate will be produced at LBDC during the low-flow event.

f. Additional samples to be collected at LABP include 36 samples for SSC and POC analysis.

These samples will be collected for the purpose of calibrating the turbidity probe currently in operation at this station. 12 pairs of samples (1 sample from each pair will be for SSC and 1 for POC) will be collected during the collection of the LV storm cross channel samples. These discrete samples will be collected using an autosampler with an inlet line installed near the turbidity sensor.

g. Metal samples represent 3 sub-samples: 1 unfiltered sample for total metals, 1 filtered sample for dissolved metals, and if enough suspended sediment is obtained in the LV storm samples, suspended sediment may be submitted for analysis of extractable metals. These 3 sub-samples (whole-water, filtered, and suspended sediment) are counted as 1 metal sample in the table. In addition, bed-sediment will be collected at LBDC after each storm event and will be analyzed for metals (and organics).

h. Standard Reference Material (NIST Standard 1944, NY Harbor sediment), will be extracted and analyzed three times for PCB/PAH and OPs. Analysis will be staggered throughout the project

1.3 Storm-Sampling Criteria

In order to produce samples that can be compared within and among tributaries, it is necessary to set certain hydrologic and sample criteria defining an “acceptable” storm event. For both storm and low-flow samples, other criteria are used to define an “acceptable” sample; specially the percentage of storm hydrograph represented by the LV sample (the coverage), and the mass of sediment obtained. The sample mass helps set the lower detection level for the analytic work, and obtaining a suitable mass reduces the number of chemical concentrations reported as “non-detect”. However, the sediment collected needs to represent the variability in sediment-bound contaminant concentrations that can occur during a storm event, so is important that the sediment is obtained over as much of the storm hydrograph as possible. Together, meeting both criteria (mass obtained and coverage) increase the confidence that the reported concentrations represent the “average” concentrations during a storm event. Flow-weighted sampling schemes help assure that reported concentrations are representative but obtaining an accurate flow-weighted LV sample is difficult as the total volume of discharge over the course of any storm event (and thus the sampling interval and number of aliquots obtained) is not known prior to sampling.

The following criteria will be used in this work:

1) Storm sampling will only be conducted if no precipitation has occurred within the previous seven

(7) days of the sampling date, and the pre-storm stream discharge is steady or gradually falling.

This requirement allows residual water in upstream storm-drains to move downstream and allows fine-grained sediment to settle in depositional areas.

2) Criterion for an “acceptable storm” is based on characteristics of each storm hydrograph, for example, on the maximum discharge or peak stage reached during an event. These variables, available only after the conclusion of a storm event, can then be compared with historic records (where available) of stream discharge in each river. In the Phase I study, an “acceptable” storm event is defined as having a peak flow exceeding the 75% percentile historic discharge value established by the USGS for each gaged tributary. These levels are listed in Table 1.3 for the gaged streams. Hydrologic data for the Lower Anacostia River at Buzzards Point has only been collected since March 2019, the release of discharge data is being withheld until sufficient calibration data is collected by the USGS. Therefore, discharge criteria for storms cannot be set.

Data collected during the Phase I study show the apparent relation between precipitation amount and intensity and maximum discharge reached in the gaged streams. Generally, the historic data show that the flow criteria will be met if rainfall exceeds 1-inch during summer months (over a 24 hours period) and 0.5” during the winter months (equivalent rainfall intensity is 0.05 inches per hour during the winter, and 0.2 inches per hour in the summer). Every effort will be made to collect storm samples when a precipitation event is predicted to meet these criteria. If a storm event is sampled that ultimately does not meet these criteria, samples will be archived, and further sampling will be scheduled.

3) Low-flow events are defined as discharge less than the historic 25% percentile discharge in the gaged streams. Generally, tributaries of interest in this study will be at low-flow conditions whenever there has been no precipitation in the preceding 7 (or more) days before sampling.

4) An acceptable sample contains at least 1 gram of suspended sediment. Parameters determined in the Phase I work will be used in setting the sampling schemes (for example, the timing when aliquots are collected for the LV composite sample) in each river. These schemes provided sufficient sediment at all sites, with up to 7 grams of sediment obtained in some storms.

Due to the small size and flashiness of the smaller tributaries (LBDC, WB, HR, NR) it was sometime necessary to not collect aliquots on set time intervals rather than discharge intervals. At NEB and NWB sufficient mass of sediment was obtained using a discharge interval of 500MG of river flow.

5) Low-flow sampling presents unique difficulties because the SSC in these rivers is typically less than 10 mg/L, and often less than 2 mg/L. To meet the 1gm sediment threshold, a minimum of 1000L of river water will be filtered.

Table 1.3. Summary of tributary discharge criteria for sample collection.

Stream

25% percentile discharge (low-flow) cfs

Mean Discharge cfs

75% percentile discharge (storm-flow) cfs

NE Branch of Anacostia River (NEB) 13 29 36

NW Branch of the Anacostia River (NWB) 5.9 15 16

Lower Beaverdam Creek (LBDC)

Hickey Run (HR) 0.016 0.38 0.79

Watts Branch (WB) 0.57 1.3 1.0

Nash Run* -- -- --

* These tributaries are not gauged. Therefore, meaningful statistics cannot be yet assigned.

Phase II Anacostia River Tributary Sampling QAPP/SOP 7/27/20

1.4 Description of Samples and Data to be Collected

Four types of samples will be collected in this work: (1) discrete samples, (2) large-volume composite samples collected during storms, (3) large-volume samples collected by stream-side filtering during low stream-flow conditions, and (4) bed-sediment samples. A consistent sampling scheme will be used at the permanent stations, the temporary trailer-deployed stations, and in the watercraft, sampling used on the Lower Anacostia. Various water-quality parameters will also be measured continuously during the sampling effort and throughout the study period.

1.4.1 Discrete Samples for Sediment Characterization

Discrete samples will be collected during storms and low-flow to determine the suspended sediment and particulate organic carbon content in the LV samples and to update the models that were developed in the Phase I study used to predict SSC from in-situ turbidity and discharge. Samples will be collected in pairs whenever aliquots are for the large-volume composite sampler are obtained. However, in the small flashy streams (Watts, Hickey, Nash) it may be necessary to collect the SSC/POC samples at 30- or 60-minute intervals. Sampling will be conducted throughout the storm and will continue regardless if the total volume of the LV composite (60-70L) sample has been obtained.12 pairs samples for SSC and POC analysis will be selected to cover the time interval aliquots were collected for LV samples – additional samples may then be processed to cover the entire storm event.

Discrete samples of river water will be collected using automatic samplers (ISCO 6712 or 6700). Samples will be collected in pre-cleaned 1L poly wedge bottles using dedicated inlet line. The discrete sampler is electrically connected to the LV samplers, so one-pair of discrete samples are collected for each aliquot obtained for the LV sample. Bottles will be replaced as necessary to ensure that discrete samples are collected throughout the entire storm event. Date and time of collection for each sample will be determined from the downloaded autosampler data and will be recorded in field notes and on electronic files.

Discrete samples will be also collected during the stream-side filtering of low-flow events. Pairs of 1L discrete samples will be collected at the start of filtering and whenever the filter pads are replaced. Samples will be collected using a 3-way valve installed on the inlet line.

1.4.2 Large-Volume Composite Samples

Two large-volume (60-70 L maximum volume) composite samples will be collected during storm events on the major tributaries. The analytic methods require two LV samples- one for PCB/PAH and one for pesticide analysis. These are produced using two autosamplers operating sequentially to collect aliquots of river water through dedicated 3/8”-ID Teflon-lined poly-tubing inlet lines. Each autosampler delivers river water directly into dedicated large Teflon (PFA) bags held in a covered plastic barrel. Further details are provided in Chapter 2. The LV samples are subsampled in the laboratory to obtain a sample for analysis of total metals.

Sampling is initiated by a preset increase in river stage, and aliquots are collected at either pre-set intervals of discharge (flow-weighted) or time. Flow-weighted sampling can be accomplished at NEB and NWB, where Phase I work helps set the intervals. In the smaller tributaries sampling is more problematic because a large range in storm discharge can occur over a short time span. In the 2017 Phase I sampling, it was found necessary to collect large fewer but volume aliquots (5L).

Sample collection involves a multi-step pump scheme: (1) the sample lines are purged, (2) the sampling line are then rinsed, (3) the lines are re-purged, (4) an aliquot of set volume is collected, and (5) the lines are purged. The typical pump rate for the ISCOs is 1L in 30 seconds (depending upon pumping head), and 30 seconds for each line purge or rinse. In case the lines become plugged with floating debris, the auto-samplers will be programed to re-try collection each aliquot once.

For the NE and NW Branches, the large-volume samplers will collect 1L aliquots at a pre-selected discharge interval (typically 500 MG). Sampling will be initiated upon a 3”-rise in stage associated with the onset of precipitation

In LBDC, Watts, Hickey Run, and Hickey Run, typical rain events produce storm flow lasting between 3 and 6 hours. Phase I work showed that the LV composite samples could only be produced by collecting few aliquots of larger volume. In order to maximize the mass of sediment obtained from these smaller streams, sampling will be initiated upon a 2” rise in stage and collect aliquots of up to 5L volume.

1.4.3 Low-Flow Samples

Suspended-sediment during low-flow conditions represents the finest sized particles transported in rivers.

Most of this material is silt and clay-sized and often has a high percentage (>80%) of organic matter. Thus, while the largest percentage of yearly mass of sediment (and sediment-bound contaminants) are transported during storms, higher concentrations of hydrophobic contaminants are often measured during low-flow. It also may be the case that the makeup of contaminants associated with sediment differs between low-flow and storm flow.

Low-flow samples are collected using two stream-side filtration systems, using the same two identical sampling systems; these systems are described in more detail in Chapter 2. Peristaltic pumps deliver river water through dedicated inlet lines (the same lines used in storm sampling whenever possible) at rates of 100 to 200mL per minute to large diameter (245mm) plate filters equipped with glass fiber filters. The filtered water is collected to determine volume filtered. Filters are replaced whenever the pumping rate slows, indicating the filters have become loaded; this is typically after 100-200L of water has been processed.

Discrete samples of river water for analysis of SSC and POC are collected concurrent with the filtering;

pairs of samples are collected whenever filters are changed. This ensures SSC and POC data are collected from the same location in the stream as is the sediment obtained for chemical analysis. During the low-flow sampling, a pre-calibrated WQ sonde will be deployed near the sampling line inlet to continuously measure temperature, conductivity, and turbidity in the river.

1.4.4 Metal Samples

Samples for total metal (whole-water) and dissolved metals (note: 1 pair of each type sample is counted as 1 sample in table 1.2) will be prepared from the LV samples collected during storm and low-flow conditions from each river. Aliquots for storm will be prepared directly from the LV composite samples as they are processed in the USGS District Field Laboratory. The LV sample bags will be agitated to stir up sediment, then 250mL of water will be transferred to pre-cleaned Teflon bottle using a peristaltic pump and acid-rinsed pump tubing. An acid-washed capsule filter (Whatman, 0.45-micron nominal pore size) will then be attached to the tubing and used to filter a second 250mL of water. Teflon bottles will be supplied pre-cleaned by the laboratory and will be preserved using ultra-pure nitric acid. Samples collected during low-flow for metal analysis will be obtained in the field directly from the inlet line used to obtain suspended sediment. Samples will be filtered and processed within a plastic portable glove box using USGS clean-hands- dirty-hands procedures.

1.4.5 Bed-Sediment Samples

Sediment from the stream will be collected in LBDC after each storm event. Samples will be collected at the location used in Phase I, located upstream in a “velocity shadow”. Sediment is collected from the uppermost 1 to 2” of newly deposited material, sieved through a pre-cleaned 60 mesh sieve into a bowl, homogenized, and transferred directly into glass jars. Three jars will be prepared: 1) for analysis of PCB and PAHs, 2) for analysis of pesticides, and 3) for sediment-bound metals. One set of duplicate samples will be collected for use in estimating the uncertainty associated with concentrations of contaminants in suspended sediment.

1.5 Large-Volume Sampling in the Lower Anacostia River

Samples will be collected during storm and low-flow conditions from the Lower Anacostia River near the USGS gage station at Buzzards Point (figure 2). This site is located at the end of First Street SW, approximately 0.5 miles downstream of the Fredrick Douglass Memorial Bridge, and approximately 1 mile upstream of the confluence with the Potomac River.

Figure 2. Location of the USGS gage station at Buzzard Point, Washington D.C.

The USGS gage-station was installed in March of 2019 on the north-west riverbank and is equipped with a side-looking acoustic Doppler velocity meter (ADVM), stage gaging equipment, a WQ data sonde, and wireless communication system that allows real-time data access. Several cross-channel surveys (in April, November, and August 2019) were conducted to collect data needed to develop the stage-velocity-

LAPB gage and cross-channel sampling location discharge relation for the gage. These cross-channel surveys were made over storm discharges ranging from -221 m3/s to +185 m3/second (- values indicate upstream flow). The 2019-2020 record shows velocities of -0.1 to +0.18 m/s over stage heights ranging from -3.1 ft to +4.1 feet. Discharge has not yet been released to the public for this station because the USGS is currently collecting additional calibration measurements.

1.6.1 Storm Sampling

Sampling at LABP will entail the use of a watercraft (of suitable size) to safely traverse the river during storms and will be equipped with a winch system needed to deploy a weighted sampling line. The USGS Maryland Science Center has recently purchased a watercraft, but as a backup, other watercraft are available, including the boat at the Washington D.C. Aquatic Education Center.

Sampling will be done consistent with USGS methods for producing composite samples in rivers (Edwards and Glyson, 1999). The LV samples will be prepared by collecting vertically integrated aliquots at 10 equal-width intervals (EWI) across the river. The channel at Buzzard Point is approximately 370 meters wide, so each station across will be separated by 37 m. At the approximate center of each EWI, the boat will stop, and vertically integrated samples will be collected using sampling lines attached to a weighted Teflon covered USGS D-95 sampler; the sampler will be slowly transited to the stream bottom. The weighted sampler ensures proper orientation in the flow of the inlet lines.

Following USGS protocols, the transit rate (time necessary to lower and raise the sampler ) is set by the approximate velocity of river (as determined from the gage station monitoring) and pumping rate. The transit rates for various river discharges applicable for the Lower Anacostia are presented in Table 1.4. By keeping the transit and pumping rates constant as the sampler is lowered and raised, the volume of water collected at each EWI station will be proportional to the discharge in the river. The rate at which water is pumped affects the grain-size of materials being captured by the inlet line. Isokinetic sampling is used to minimize bias in grain-size captured and occurs when the velocity of water entering the sampling line equals the velocity of water in the river at the inlet point. Table 1.4 presents the pumping rate through 3/8” ID tubing required for isokinetic sampling as a function of river water velocity- typical velocities during storms are >0.15 m/s. The pump rate will be checked before each aliquot is obtained.

Figure 4 is a schematic of the equipment that is used to collect samples at LABP. Sampling lines are connected to a weighted sampler that is slowly lowered through the water column, while peristaltic pumps are used to transfer water directly into a 10L Teflon churn. Once 10L have been obtained, the churn is agitated to mix the sediment, then 250 mL is transferred to a Teflon sample bottle for metal analysis. The remainder of water and sediment is then transferred directly into one LV Teflon sample bag (PCB/PAH sample). A second vertical transit is then made to collect water for the LV pesticide sample. Once the LV aliquots have been obtained, two additional vertical transits are made to collect (vertically integrated) 1L samples for SSC and POC. Finally, a WQ data sonde is used to obtain a vertical profile of temperature, conductivity, and turbidity in the water column. A summary of the sampling scheme used at each EWI station is shown in figure 4, and is:

• Record location of EWI station from GPS, point watercraft into flow, anchor and allow watercraft to drift downflow until anchor holds.

• The WQ sonde is lowered into the water and allowed 3 minutes (minimum) to equilibrate.

• The depth of river is measure along with vertical profiles of turbidity, SC, and temperature measured at 1-foot intervals from the surface to the river bed.

• The transit rate is calculated and used to set the pump rate. The pump rate is checked while the inlet lines are being purged.

• The inlet lines/weighted sample is then lowered, and the peristaltic pump started transferring 7L of river water into a Teflon churn.

• The churn is agitated slowly and approximately 500 mL of water/sediment is transferred into a Teflon bottle for metal sample. The metal sample is stored on ice until the transect is completed.

The remainder of the sample in the churn is transferred to the LV sample bag for PCB/PAH sample. The churn may be rinsed with a small amount of laboratory grade DI water if necessary.

• The weighted sampler/inlet lines are lowered a second time and 7L of river water is collected in the Teflon churn.

• The churn is emptied into the second LV sample bag for pesticide analysis.

• The inlet lines/weighted sampler sampler is lowered a third time and 7L of river water is collected in the Teflon churn. The churn is agitated and 1L of sample is transferred into a poly-bottle for SSC sample bottle, and 1L is transferred into a glass sample bottle for later processing for POC. SSC and POC bottles are stored on ice in coolers until the transect is completed.

• The sampling lines and sonde are raised and the boat moves to next station where the sampling is repeated.

• Upon completion of the transect, samples will be returned to the shore where LV samples are secured then transferred to field vans, and discrete samples are processed and preserved for dissolved metals and POC.

Ideally, the LV composite sample would consist of aliquots collected during transits on the rising limb, the crest, and again on the falling limb of the storm hydrograph. This would entail making transits of the river that might be spaced hours apart. This procedure will be used if all 3 traverses can be made safely during daylight hours. Conditions and timing of the storm may, therefore, constrain the sampling to 2 of the three legs of the storm hydrograph.

1.6.2 Low-Flow Sampling.

Low-flow samples will be collected using the same equipment described in section 1.7.3 for stream-side sample collection. The boat will be anchored at each of the 10 EWI stations, and the sampling lines lowered to 60% depth of the river; this ensures samples are from the approximate location of maximum velocity in the profile, and that bottom sediment is not stirred into the water column and sampled. Using a flow-through volume meter, either 25L or 50L of river water will be filtered through the LV filtering apparatus, depending on whether 4 transits (25L volume aliquots) or 2 transits (50L aliquots) are used to collect the required 1000L of river water. At each EWI station, discrete samples for SSC and POC will be then collected directly from the pumping line.

1.6.3 Sampling for Calibration of Turbidity Sensor

The continuous turbidity monitor that has been operating at LABP since 2018 measures turbidity at a single point near the north shore of the Lower Ancostia River. In order to start the calibration of turbidity to suspended-sediment concentrations in the river, discrete point samples will be collected during each storm event. An ISCO automatic sampler will be installed at the site. The inlet line for the sampler will be positioned near the turbidity sensor (thus the samples will not be collected from the same relative depth below the water level). The automatic sampler will be programmed to collect discrete samples at 15-minute intervals during the interval when cross-channel LV storm samples are being collected. Upon completion of the cross-channel sampling, the automatic sampler will be reprogrammed to collect samples at 30-minute intervals over the following 12-hours. Upon completion of the storm event, the hydrograph and turbidity trace for the event will be studied, and 12 samples selected for SSC and POC analysis. Samples will be selected to cover the complete range for turbidity values measured during the sampling interval. These additional samples are included in the sample tabulation

Table 1.4. Transit and pumping rates for flow velocities.

Depth in feet 0-5 5 to 10 10 to 15 0-5 Acceptable range Ratio of transit rate divided by mean water velocity for use with D-95 sampler

0.06-0.11 0.11-0.15 0.15-0.175 --

Mean velocity meters/second Velocity feet per second

Transit rate in inches per minute

Pump rate in liters per minute

+/-0.6 1.97 85-156 156-213 213-248 10.3 +/-0.5 1.64 71-130 130-177 177-207 8.6 +/-0.4 1.31 57-104 104-142 142-165 6.8 +/-0.3 0.98 42-78 78-106 106-124 5.1 +/-0.2 0.66 29-52 52-71 71-83 3.4

+/-0.15 0.49 21-39 39-53 53-62 2.6 +/-0.1 0.33 14-26 26-36 36-42 1.7

+/- 0.05 0.16 6.9-13 13-17 17-20 0.9 +/-0.01 0.033 1.4-2.6 2.6-3.6 3.6-4.2 0.2

1. Transit rates are calculated for a D-59 weighted sampler using a 3/16” diameter inlet nozzle.

Maximum useable depth for this sampler is 15 feet.

2. Pumping rates are for 3/8” ID poly tubing, assuming 3 feet between water surface and boat deck, and no head loss in tubing.

Figure 3. Schematic of equipment used to collect large-volume composite samples from the Lower Anacostia River.

Boat

10L Teflon churn

Cable and winch

2- Teflon inlet lines attached to weighted sampler

Weighted D-95 sampler - inlets pointing downflow

Peristaltic pump, 1 for each line

Teflon sample bags in plastic barrels

WQ

data sonde

Synopsis of Sampling at each EWI station.

1. Lower WQ sonde and allow to equilibrate

2. Determine depth of river and conduct vertical profile of turbidity, SC, and temperature.

3. Calculate transit rate, check pump rate, purge inlet lines

4. Lower weighted sampler and inlet lines, collect LV aliquot in Teflon churn

5. Agitate churn, transfer approximately 100mL into

Teflon bottle for metal sample. Transfer remainder of sample from churn to LV sample bag for PCB/PAH sample. Ensure all sediment has been removed.

6. Lower weighted sampler and collect second LV aliquot in Teflon churn.

7. Transfer sample into second LV sample bag for pesticide analysis

8. Lower weighted sampler and collect third sample aliquot into Teflon churn. Agitate churn and transfer 1L into SSC sample bottle, and 1L into glass POC sample bottle.

9. Raise sampling lines and sonde, move to next station

PCB/

PAH

Pesti-cide

Discrete bottle for metals, SSC, and

POC

1.7 In-Situ Water-Quality Data

Water-quality will be measured continuously by deploying sondes at each site to provide context for the LV samples. Sonde data are also used to calibrate the intake point to the cross-section (site calibration) and to add to the turbidity record used as a surrogate for SSC, ultimately used to estimate sediment and contaminant loads.

1. Point-sample calibration. It is important that the sediment concentrations in samples collected at a point in the stream (the location of the autosampler intake line nozzle) can be related to the average SSC across the entire river channel. The point-sample calibration allows differences between the SSC in discrete samples to be corrected to represent cross-channel average concentrations. Calibration involves collecting discrete SSC using the autosampler while a composite SSC sample is collected using the EWI method. Alternatively, the turbidity at the sampling inlet point can be compared with turbidity measured across the channel. Calibration was performed at the sites in the Phase I study and will be rechecked in Phase II.

Cross-channel data can be collected either by wading (during lower flow regimes) or from a nearby bridge where available. After point samples and cross-channel SSC are collected, box coefficients can be calculated and used to adjust the point sample concentrations to represent the entire channel. USGS field techniques involving collecting samples, or measuring turbidity, at equal width intervals (Edwards and Glysson, 1999). In the Phase II work, turbidity will be used to check cross-section SSC homogeneity and if necessary, sampling will be conducted to produce new box coefficients. As discussed in Section 1.8.2, discrete samples for SSC will be collected during dstorm sampling on the Lower Anacostia River for the purpose of relating turbidity to SSC at the Buzzard Point station.

2. In-situ monitoring of river conditions. Water-quality parameters in each tributary will be measured using in-situ data sondes to provide physical context for the contaminant concentrations, and to provide additional input data for models produced in Phase I. Data sondes (YSI 6200 or equivalent) are currently operating at NEB, NWB LBDC, Watts Branch, Hickey Run and LABP;

these measure temperature, turbidity, specific conductance, and other WQ parameters concurrent with stage and discharge. These sondes are maintained by the USGS using standard procedures (Gibs et al, 2007) that include bi-weekly cleaning, calibration checks, and recalibration if needed. A similar data sonde will be deployed in Nash Run for the specific sampled storm events.

1.8 Target Analyte List

The analytes to be measured in this study are listed in Table 1.5. The suspended sediment captured on filter paper will be frozen before shipping to the appropriate analytic laboratory for extraction and analysis.

Samples dedicated for analysis of 209 PCBs congeners and 35 alkylated and parent PAHs will be extracted using EPA method SW-846 and measured for PCB congeners using EPA Method SW-1668A.

PAHs will be measured using a propriety method that directly follows USEPA Method SW-846 Method 8270D. Organo-chlorine pesticides compounds (17 total) will be analyzed using a method based on USEPA SW846 Method 8270D. Metals will be measured in unfiltered (whole-water or total) and filtered samples (dissolved) concentrations using EPA Method SW3050B. The organic contaminants will be measured at Test America Laboratories, in Knoxville TN. Methods, quality assurance and control documents are presented in documents prepared by Tetra Tech and attached to this plan.

Suspended sediment concentration will be measured at the USGS Kentucky Sediment Laboratory using gravimetric techniques with a detection level of 0.5 mg/L. POC will be extracted in the Md Water Science Center. Samples will be processed in triplicate using pressurized filtration (N2-gas) through pre-baked glass fiber filters following standard USGS guidelines. 3-100mL replicates will be processed for each POC sample collected; loaded filters will be individually wrapped in baked aluminum foil, sealed in a zip-lock plastic bag, frozen, and then shipped overnight to the USGS National Water Quality Laboratory in Denver Colorado. POC is reported at a detection level of 0.2 mgC /L sample processed.

Table 1.5. Target Analytes to be measured by Test America, Inc.

PCBs measured using EPA 1668 A PCB-1 2-Chlorobiphenyl PCB-2 3-Chlorobiphenyl PCB-3 4-Chlorobiphenyl PCB-4 2,2'-Dichlorobiphenyl PCB-5 2,3-Dichlorobiphenyl PCB-6 2,3'-Dichlorobiphenyl PCB-7 2,4-Dichlorobiphenyl PCB-8 2,4'-Dichlorobiphenyl PCB-9 2,5-Dichlorobiphenyl PCB-10 2,6-Dichlorobiphenyl PCB-11 3,3'-Dichlorobiphenyl PCB-12/13* 3,4-Dichlorobiphenyl, 3,4'-Dichlorobiphenyl PCB-14 3,5-Dichlorobiphenyl PCB-15 4,4'-Dichlorobiphenyl PCB-16 2,2',3-Trichlorobiphenyl PCB-17 2,2',4-Trichlorobiphenyl PCB-18/30 2,2',5-Trichlorobiphenyl, 2,4,6-Trichlorobiphenyl PCB-19 2,2',6-Trichlorobiphenyl PCB-20/28 2,3,3'-Trichlorobiphenyl, 2,4,4'-Trichlorobiphenyl PCB-21/33 2,3,4-Trichlorobiphenyl, 2,3',4'-Trichlorobiphenyl PCB-22 2,3,4'-Trichlorobiphenyl PCB-23 2,3,5-Trichlorobiphenyl PCB-24 2,3,6-Trichlorobiphenyl PCB-25 2,3',4-Trichlorobiphenyl PCB-26/29 2,3',5-Trichlorobiphenyl, 2,4,5-Trichlorobiphenyl PCB-27 2,3',6-Trichlorobiphenyl PCB-31 2,4',5-Trichlorobiphenyl PCB-32 2,4',6-Trichlorobiphenyl PCB-34 2,3',5'-Trichlorobiphenyl PCB-35 3,3',4-Trichlorobiphenyl PCB-36 3,3',5-Trichlorobiphenyl PCB-37 3,4,4'-Trichlorobiphenyl PCB-38 3,4,5-Trichlorobiphenyl PCB-39 3,4',5-Trichlorobiphenyl PCB-41/40/71 2,2',3,4-Tetrachlorobiphenyl, 2,2',3,3'-Tetrachlorobiphenyl, 2,3',4',6-Tetrachlorobiphenyl PCB-42 2,2',3,4'-Tetrachlorobiphenyl PCB-43/73 2,2',3,5-Tetrachlorobiphenyl, 2,3',5',6-Tetrachlorobiphenyl PCB-44/47/65 2,2',3,5'-Tetrachlorobiphenyl, 2,2',4,4'-Tetrachlorobiphenyl, 2,3,5,6-Tetrachlorobiphenyl PCB-45/51 2,2',3,6-Tetrachlorobiphenyl, 2,2',4,6'-Tetrachlorobiphenyl PCB-46 2,2',3,6'-Tetrachlorobiphenyl PCB-48 2,2',4,5-Tetrachlorobiphenyl PCB-49/69 2,2',4,5'-Tetrachlorobiphenyl, 2,3',4,6-Tetrachlorobiphenyl PCB-50/53 2,2',4,6-Tetrachlorobiphenyl, 2,2',5,6'-Tetrachlorobiphenyl PCB-52 2,2',5,5'-Tetrachlorobiphenyl PCB-54 2,2',6,6'-Tetrachlorobiphenyl PCB-55 2,3,3',4-Tetrachlorobiphenyl PCB-56 2,3,3',4'-Tetrachlorobiphenyl

Table 1.5. Analyte List – continued

PCB-57 2,3,3',5-Tetrachlorobiphenyl PCB-58 2,3,3',5'-Tetrachlorobiphenyl PCB-59/62/75* 2,3,3',6-Tetrachlorobiphenyl, 2,3,4,6-Tetrachlorobiphenyl, 2,4,4',6-Tetrachlorobiphenyl PCB-60 2,3,4,4'-Tetrachlorobiphenyl PCB-61/70/74/76 2,3,4,5-Tetrachlorobiphenyl, 2,3',4',5-Tetrachlorobiphenyl, 2,4,4',5-Tetrachlorobiphenyl, 2,3',4',5'-Tetrachlorobiphenyl PCB-63 2,3,4',5-Tetrachlorobiphenyl PCB-64 2,3,4',6-Tetrachlorobiphenyl PCB-66 2,3',4,4'-Tetrachlorobiphenyl PCB-67 2,3',4,5-Tetrachlorobiphenyl PCB-68 2,3',4,5'-Tetrachlorobiphenyl PCB-72 2,3',5,5'-Tetrachlorobiphenyl PCB-77 3,3',4,4'-Tetrachlorobiphenyl PCB-78 3,3',4,5-Tetrachlorobiphenyl PCB-79 3,3',4,5'-Tetrachlorobiphenyl PCB-80 3,3',5,5'-Tetrachlorobiphenyl PCB-81 3,4,4',5-Tetrachlorobiphenyl PCB-82 2,2',3,3',4-Pentachlorobiphenyl PCB-83/99 2,2',3,3',5-Pentachlorobiphenyl, 2,2',4,4',5-Pentachlorobiphenyl PCB-84 2,2',3,3',6-Pentachlorobiphenyl PCB-85/116/117 2,2',3,4,4'-Pentachlorobiphenyl, 2,3,4,5,6-Pentachlorobiphenyl, 2,3,4',5,6-

Pentachlorobiphenyl

PCB-86/87/97/109/

119/125

2,2',3,4,5-Pentachlorobiphenyl, 2,2',3,4,5'-Pentachlorobiphenyl, 2,2',3,4',5'- Pentachlorobiphenyl, 2,3,3',4,6-Pentachlorobiphenyl, 2,3',4,4',6-Pentachlorobiphenyl, 2,3',4',5',6-Pentachlorobiphenyl

PCB-88/91 2,2',3,4,6-Pentachlorobiphenyl,…