1551844-R8SDMS

Second Five-Year Review Report
for the

Milltown Reservoir /Clark Fork River
Superfund Site

EPA ID MTD980717565

Milltown

Missoula, Granite, Powell, and Deer Lodge Counties, Montana

Prepared By:

United States Environmental Protection Agency
Region 8
Denver, Colorado

September 2016

Date:

Martin Hestmark

Assistant Regional Administrator
Office of Ecosystems Protection
and Remediation


-------
Table of Contents

List of Acronyms	iii

Executive Summary	iv

Five-Year Review Summary Form	v

I.0	Introduction	8

2.0 Site Chronology	9

3.0 Background	10

3.1	Physical Characteristics	10

3.2	Land and Resource Use	11

3.3	History of Contamination	12

3.4	Initial Response	13

3.5	Basis for Taking Action	16

4.0 Remedial Actions	17

4.1	Remedy Selection	17

4.2	Remedy Implementation	23

4.3	Operation and Maintenance (O&M)	28

5.0 Progress Since the Last Five-Year Review	29

6.0 Five-Year Review Process	31

6.1	Administrative Components	31

6.2	Community Involvement	31

6.3	Document Review	31

6.4	Data Review	37

6.5	Site Inspection	42

6.6	Interviews	44

7.0 Technical Assessment	46

7.1	Question A: Is the remedy functioning as intended by the decision

DOCUMENTS?	46

7.2	Question B: Are the exposure assumptions, toxicity data, cleanup levels and

REMEDIAL ACTION OBJECTIVES (RAOs) USED AT THE TIME OF REMEDY SELECTION
STILL VALID? 	47

7.3	Question C: Has any other information come to light that could call into

QUESTION THE PROTECTIVENESS OF THE REMEDY?	48

7.4	Technical Assessment Summary	48

8.0 Issues	49

9.0 Recommendations and Follow-up Actions	50

10.0 Protectiveness Statements	50

II.0	Next Review	51

Appendix A: List of Documents Reviewed	A-l

Appendix B: Press Notice	B-l

Appendix C: Interview Forms	C-l

i


-------
Appendix D-l: MRSOU Site Inspection Checklist	D-l

Appendix D-2: CFROU Site Inspection Checklist	D-8

Appendix E-l: Photographs from MRSOU Site Inspection	E-l

Appendix E-2: Photographs from the CFROU Site Inspection	E-16

Appendix F: MRSOU Monitoring Data	F-l

Appendix G: CFROU Monitoring Data Summary	G-l

Appendix H: MRSOU Sediment Accumulation Areas	H-l

Appendix I: Surface Water Data Evaluation	1-1

List of Tables

Table 1: Chronology of Site Events	9

Table 2: MRSOU Groundwater COC Cleanup Goals	19

Table 3: MRSOU Surface Water COC Cleanup Goals	20

Table 4: CFROU Groundwater COC Cleanup Goals	22

Table 5: CFROU Surface Water COC Cleanup Goals	22

Table 6: Arsenic Soil Cleanup Goals	23

Table 7: Progress on Recommendations from the 2011 FYR	30

Table 8: Previous and Current ARARs for Groundwater COCs	32

Table 9: Previous and Current ARARs for Surface Water COCs	33

Table 10: Institutional Control (IC) Summary Table	35

Table 11: Current Site Issues	49

Table 12: Recommendations to Address Current Site Issues	50

List of Figures

Figure 1: Site Map	14

Figure 2: Detailed Site Map - MRSOU	15

Figure 3. CFROU Reach A Phase Breaks	26

Figure 4: Powell County Overlay District (CFROU)	36

Figure 5: Arsenic Concentrations in Wells 905, 103B, 917B and 107C	38

Figure 6: Arsenic Concentrations in Wells 105C, 11, HLA2, 107A and 11R	39

Figure 7: Arsenic Concentrations in Wells 110B, 907, 922D, 104A, 921 A, TPR-10 and 913A. 39

ii


-------
List of Acronyms

ARAR

Applicable or Relevant and Appropriate Requirement

ARCO

Atlantic Richfield Company

AWQC

Ambient Water Quality Criteria

BMP

Best Management Practice

CaC03

Calcium carbonate

CCC

Criterion Continuous Concentration

CERCLA

Comprehensive Environmental Response, Compensation and Liability Act, as



amended

CFR

Code of Federal Regulations

CFROU

Clark Fork River Operable Unit

CMC

Criteria Maximum Concentrations

COC

Contaminant of Concern

CUP

Conditional Use Permit

EPA

United States Environmental Protection Agency

FYR

Five-Year Review

IC

Institutional Control

MCL

Maximum Contaminant Level

Hg/L

micrograms per liter

mg/kg

milligrams per kilogram

mg/L

micrograms per liter

MCCHD

Missoula City and County Health Department

MDEQ

Montana Department of Environmental Quality

MDL

Method detection limit

MRSOU

Milltown River Sediments Operable Unit

NCP

National Oil and Hazardous Substances Pollution Contingency Plan

NPL

National Priorities List

NRDP

Natural Resource Damage Program

OU

Operable Unit

O&M

Operation and Maintenance

PEC

Probable Effects Concentration

POC

Point of compliance

PRP

Potentially Responsible Party

RAMP

Remedial Action Monitoring Plan

RAO

Remedial Action Objective

RI/FS

Remedial Investigation and Feasibility Study

RIPes

Riparian Evaluation System

ROD

Record of Decision

RPM

Remedial Project Manager

SAA

Sediment Accumulation Area

TBC

To-Be-Considered

USFWS

United States Fish and Wildlife Service

USGS

United States Geological Survey

iii


-------
Executive Summary

The Milltown Reservoir /Clark Fork River site includes about 120 miles of the Clark Fork River
upstream of the former Milltown Dam and Reservoir. The Milltown Dam and Reservoir were
located at the confluence of the Clark Fork and Blackfoot Rivers, a few miles upstream of
Missoula. From the 1860s until well into the 20th century, mineral- and arsenic-laden waste from
mining activities in the region flowed into the headwaters of the Clark Fork River, contaminating
the river and its beds and banks from the Warm Springs Ponds to the Milltown Reservoir. As
contaminated sediments and mine-mill wastes moved downstream, about 6.6 million cubic yards
of these sediments accumulated behind the Milltown Dam over time. These mining activities and
the downstream transport of mining-related wastes contaminated floodplains, sediment, surface
water and groundwater with heavy metals.

This FYR report addresses all site operable units (OUs). OU2 is the Milltown Reservoir
Sediments (MRSOU), including the area encompassed by the former Milltown dam and
reservoir. OU1 (the Milltown Drinking Water Supply OU) is now part of OU2. OU3 is the Clark
Fork River (CFROU) area upstream of the MRSOU and downstream of the Silver Bow
Creek/Butte Area site and the Anaconda Smelter site.

The MRSOU remedy includes construction of a bypass channel at the reservoir; removal of
contaminated reservoir sediment; off-site disposal and use of contaminated sediment as
vegetative cap material; removal of the Milltown Dam; continuation of a replacement water
supply program in the town of Milltown; implementation of temporary groundwater controls
until the Milltown aquifer recovers and other institutional controls; and long-term monitoring of
surface water and groundwater. Remedy construction began in 2006 and is substantially
complete.

The remedy at MRSOU (OU2) currently protects human health and the environment because
potential exposure to contaminated groundwater, surface water and sediment is controlled. For
the remedy to be protective over the long term, the following actions need to be taken:

•	Implement institutional controls for the MRSOU comprehensive institutional control plan
and its components.

•	Determine if additional measures are needed to reduce arsenic concentrations in
groundwater to levels at or below the cleanup goals.

The CFROU remedy includes soil and sediment removal and disposal outside of the OU, some
in-place treatment of soils, revegetation of removed or treated areas, streambank stabilization,
weed control, institutional controls and monitoring. MDEQ started the remedial action
construction with yard removals in Deer Lodge in 2010-2011, the Trestle Project in 2011-2012,
and Eastside Road Pastures, 2012-2013, CFR Reach A, Phase 1 remedial construction on the
river began in 2013. Remedial implementation is ongoing.

The remedy at CFROU (OU3) is expected to be protective of human health and the environment
upon completion of the remedial action. In the interim, exposure pathways that could result in
unacceptable risks are being controlled.

iv


-------
Five-Year Review Summary Form

SITE IDENTIFICATION

Site Name: Milltown Reservoir /Clark Fork River

EPA ID:

MTD980717565

Region: 8

State: MT

City/County: Milltown and Missoula, Granite,
Powell and Deer Lodge Counties

NPL Status: Final

Multiple OUs?

Yes

Has the site achieved construction completion?

No

Lead agency: EPA

If "Other Federal Agency" selected above, enter Agency name:

Author name: Sara Sparks (EPA) and Ryan Burdge and Treat Suomi (Skeo)

Review period: 10/01/2015 - 09/23/2016

Date of site inspection: 11/02/2015-11/04/2015

Type of review: Statutory

Review number: 2

Triggering action date: 09/23/2011

Due date (five years after triggering action date): 09/23/2016

V


-------
Five-Year Review Summary Form (continued)

Issues/Recommendations

OU(s) without Issues/Recommendations Identified in the Five-Year Review:
OU3

Issues and Recommendations Identified in the Five-Year Review:

OU(s): OU2

Issue Category: Institutional Controls

Issue: Institutional controls for MRSOU are not yet implemented for areas
where waste has been left in place and areas where groundwater
contamination is above ROD standards.

Recommendation: Implement institutional controls for the MRSOU
comprehensive institutional control plan and its components.

Affect Current
Protectiveness

Affect Future
Protectiveness

Implementing
Party

Oversight
Party

Milestone Date

No

Yes

PRP

EPA/State

09/30/2017

OU(s): OU2

Issue Category: Remedy Performance

Issue: Groundwater concentrations at MRSOU continue to exceed
arsenic cleanup goals and do not appear to be declining

Recommendation: Determine if additional measures are needed to
reduce arsenic concentrations below the cleanup goals and implement
measures determined to be necessary.

Affect Current
Protectiveness

Affect Future
Protectiveness

Implementing
Party

Oversight
Party

Milestone Date

No

Yes

PRP

EPA/State

09/30/2017

vi


-------
Protectiveness Statements

Operable Unit:	Protectiveness Determination:	Addendum Due Date

OU2	Short-term Protective	(if applicable):

Protectiveness Statement:

The remedy at MRSOU (OU2) currently protects human health and the environment because
potential exposure to contaminated groundwater, surface water and sediment is controlled.
For the remedy to be protective over the long term, the following actions need to be taken:
implement institutional controls for the MRSOU comprehensive institutional control plan and
its components and determine if additional measures are needed to reduce arsenic
concentrations below the cleanup goals.

Operable Unit:	Protectiveness Determination:	Addendum Due Date

OU3	Will be Protective	(if applicable):

Protectiveness Statement:

The remedy at CFROU (OU3) is expected to be protective of human health and the
environment upon completion of the remedial action. In the interim, exposure pathways that
could result in unacceptable risks are being controlled.

7


-------
Second Five-Year Review Report
for

Milltown Reservoir /Clark Fork River Superfund Site
1.0 Introduction

The purpose of a five-year review (FYR) is to evaluate the implementation and performance of a
remedy in order to determine if the remedy is protective of human health and the environment.
FYR reports document FYR methods, findings and conclusions. In addition, FYR reports
identify issues found during the review, if any, and document recommendations to address them.

The United States Environmental Protection Agency (EPA) prepares FYRs pursuant to the
Comprehensive Environmental Response, Compensation and Liability Act as amended
(CERCLA) Section 121, 42 U.S.C. § 9621, and the National Oil and Hazardous Substances
Pollution Contingency Plan (NCP). CERCLA Section 121 states:

If the President selects a remedial action that results in any hazardous substances,
pollutants, or contaminants remaining at the site, the President shall review such remedial
action no less often than each 5 years after the initiation of such remedial action to assure
that human health and the environment are being protected by the remedial action being
implemented. In addition, if upon such review it is the judgment of the President that
action is appropriate at such site in accordance with section [104] or [106], the President
shall take or require such action. The President shall report to the Congress a list of
facilities for which such review is required, the results of all such reviews, and any
actions taken as a result of such reviews.

EPA interpreted this requirement further in the NCP, 40 Code of Federal Regulations (CFR)
Section 300.430(f)(4)(ii), which states:

If a remedial action is selected that results in hazardous substances, pollutants, or
contaminants remaining at the site above levels that allow for unlimited use and
unrestricted exposure, the lead agency shall review such action no less often than every
five years after initiation of the selected remedial action.

Skeo, an EPA Region 8 contractor, conducted the FYR and prepared this report regarding the
remedy implemented at the Milltown Reservoir /Clark Fork River Superfund site (the Site) in
Milltown and Missoula, Granite, Powell, and Deer Lodge Counties, Montana. EPA's contractor
conducted this FYR from October 2015 to September 2016.

EPA is the lead agency for developing and implementing the remedy at OU2 through oversight
of the potentially responsible party (PRP)-financed cleanup at the Site, and coordination with the
State of Montana Natural Resource Damage Program which is performing certain restoration site
activities to, in some cases, accomplish remedial goals and objectives. The Montana Department
of Environmental Quality (MDEQ), as the support agency representing the State of Montana at
OU2, and has reviewed all supporting documentation and provided input to EPA during the FYR
process.

8


-------
MDEQ is the lead agency for implementation of the Remedial Design, the Remedial Action, and
the Operation and Maintenance of the Remedy at the Clark Fork Site, through special account
funding obtained by EPA and the State through an enforcement settlement at OU3. The State of
Montana Natural Resource Damage program is also performing certain natural resource damage
restoration activities at OU3 which in cooperation with MDEQ, to date, have been supplemental
to the remedial implementation. EPA is the support agency for OU3. EPA has prepared this Site-
wide five year review report, in consultation with MDEQ and the State of Montana Natural
Resource Damage Program.

This is the second FYR for the Site. The triggering action for this statutory review is the previous
FYR. The FYR is required due to the fact that hazardous substances, pollutants or contaminants
remain at the Site above levels that allow for unlimited use and unrestricted exposure. The Site
consists of two operable units (OUs). This FYR report addresses all OUs for the site.

OU2 is the Milltown Reservoir Sediments (MRSOU), including the area encompassed by the
former Milltown dam and reservoir. OU1 (the Milltown Drinking Water Supply OU) is now part
of OU2. OU3 is the Clark Fork River (CFROU) area upstream of the MRSOU and downstream
of the Silver Bow Creek/Butte Area site and the Anaconda Smelter site.

2.0 Site Chronology

Table 1 lists the dates of important events for the Site.

Table 1: Chronology of Site Events

l.\cn(

Diilo

1 .ocal public health authorities discovered arsenic contamination in
drinking water wells in Milltown, Montana

1981

EPA added the Site to the Superfund program's National Priorities List
(NPL)

September 08, 1983

EPA issued interim Record of Decision (ROD) for OU1, requiring
construction of a deep well and water tank to serve as an alternative
water supply for Milltown residents. This ROD was amended in 1985.

April 14, 1984

Remedial action construction for OU 1 completed

1986

Atlantic Richfield Company prepared major portions of the final CFROU
remedial investigation and feasibility study (RI/FS). RI/FS work
continued for several years after 1987, including the preparation of a
baseline human health and ecological risk assessment.

1987

RI/FS order on consent for MRSOU issued to Atlantic Richfield
Company (ARCO)

1991

MRSOU RI and baseline human health, ecological and continued release
risk assessments completed

September 16, 1993

PRPs complete Final RI Report for MRSOU

February 15, 1995

Draft FS for MRSOU groundwater released by ARCO. The same year,
unforeseen climatic conditions caused ice scour event, which sent high
levels of metals contamination down river; EPA expanded FS scope and
conducted further risk assessments

1996

EPA issued CFROU ROD

April 2004

MRSOU RI/FS completed; EPA issues MRSOU ROD

December 15, 2004

9


-------
I.mmH

D.ile

Consent Decree for PRP performance of MRSOU remedy and O&M
entered by federal court; this includes requirements for PRP continued
funding of water supply operation and maintenance (O&M) activities.
The Consent Decree also provided for the performance of natural
resource damage actions by the State of Montana at the MRSOU, some
of which are intended to fulfill remedial action requirements.

August 2005

Remedial action at MRSOU begins

February 15, 2006

Initial reservoir drawdown (Stage 1) and start of MRSOU remedial
action

June 01, 2006

Consent Decree for PRP cashout of CFROU remedy and O&M entered
by federal court. This provides for the performance of the CFROU
remedy and O&M by the MDEQ using the cashout money, and funding
and performance of natural resource damage actions by the State of
Montana Natural Resource Damage program.

August 21, 2008

EPA approves Draft Repository O&M Plan and Changes to the Remedial
Action Monitoring Plan (RAMP) for MRSOU

March 2010

MDEQ begins remedial action at CFROU, including irrigated land, Deer
Lodge residential, and Trestle area work.

October 5, 2010

Transfer of reservoir property to State of Montana

December 2010

Clark Fork River bypass channel removal begins

December 2010

EPA completes first five-year review for MRSOU

September 2011

MRSOU remedial activities construction activities were significantly
completed

June 2012

MDEQ begins remedial action at CFROU Reach A, Phase 1.

March 4, 2013

MDEQ completes remedial action at CFROU Reach A, Phase 1. Work at
other Phase areas is ongoing.

April 4, 2014

Remedial action begins at CFROU Phase 5 and 6

July 15,2014

MDEQ submits construction completion report for Phase 1 to EPA

March 25, 2015

EPA and MDEQ release Explanation of Significant Differences for
CFROU

June 12, 2015

Remedial action begins at CFROU Phase 2

June 29, 2015

3.0	Background

3.1	Physical Characteristics

The Clark Fork Basin Superfund complex is made up of four contiguous Superfund sites, each
broken into separate NPL sites. The four Superfund sites are the Silver Bow Creek/Butte Area
site, the Montana Pole site, the Anaconda Smelter site and Milltown Reservoir /Clark Fork River
site. The Anaconda Smelter site, the Silver Bow Creek/Butte Area site and the Milltown
Reservoir /Clark Fork River site are each broken into several OUs.

EPA originally designated three OUs for the Site. There are currently two site OUs.

• OU2 is the Milltown Reservoir Sediments (MRSOU). It includes about 540 acres in
the Clark Fork River and Blackfoot River floodplain (Figure 1). The MRSOU
consists of the area encompassed by the former Milltown dam and reservoir and the
associated groundwater contamination. OU1, an interim remedy, is now part of the

10


-------
MRSOU. It focused on providing a safe water supply to Milltown area residents
through the establishment of a public water supply system in Milltown, Montana.

•	OU3 is the Clark Fork River (CFROU) area upstream of the MRSOU and
downstream of the Silver Bow Creek/Butte Area site and the Anaconda Smelter site
(Figure 1). CFROU consists of about 120 river miles of the Clark Fork River,
including surface water, groundwater, soils, in-stream sediments, sediment deposition
and contaminated property, and air located within and adjacent to the 100-year
historic floodplain of the Clark Fork River.

MRSOU is located at the confluence of the Clark Fork and Blackfoot rivers in Missoula County,
Montana. The Milltown Reservoir was formed by the Milltown Dam, built from 1905 to 1908. It
is located approximately 7 miles upstream of downtown Missoula, Montana.

From its headwaters, the Clark Fork River flows north for approximately 43 river miles past the
towns of Galen, Deer Lodge and Garrison (this stretch is known as Reach A of CFROU). The
river then runs northwest for approximately 77 river miles to the headwaters of the Milltown
Reservoir near Bonner.

To better study and evaluate remedial options, EPA divided the CFROU into three reaches based
on physical features of the landscape, proximity to historic mining and intensity of impacts:

•	Reach A - Deer Lodge Valley Reach: Extends from the southeastern tip of the
CFROU near river mile 0 at Warm Springs Creek to just upstream of Garrison at river
mile 43. Reach A has the broadest extent of the 100-year floodplain and is nearest to
historic mining and milling sites in Butte and Anaconda. There are extensive exposed
tailings and unstable streambanks as well as stressed vegetation in this area.

•	Reach B - Drummond Valley Reach. Extends from immediately upstream of
Garrison, where the Little Blackfoot River enters the Clark Fork, to downstream of
Drummond at river mile 76, for a total of 31 river miles. At the starting point for this
reach, the addition of water from the Little Blackfoot River may, under certain flow
conditions, nearly double the Clark Fork's flow. The floodplain is more narrow and
the gradient higher than Reach A, and exposed tailings are far less extensive.

•	Reach C - Bearmouth Canyon Reach. Extends 47 river miles from Drummond to the
northwest tip of the OU area. Through this reach, the floodplain is constrained by a
narrow valley, roads and railroad grades. Here, the flow is augmented by several
tributaries and the reach is farther away from historic mining sites. No exposed
tailings are evident.

3.2 Land and Resource Use

The former Milltown Dam was owned and operated as a hydroelectric generating facility by
North Western Corporation and its predecessors. The community of Milltown is located a half-
mile east of the former dam and powerhouse. The community of Bonner borders Milltown to the
northeast. About 1,700 people live in Milltown, according to 2010 U.S. Census data. A new

11


-------
public water supply was developed for Milltown under OU1. Private wells in the area are
sampled by the Missoula City and County Health Department (MCCHD).

The MRSOU (OU2) includes the Milltown Reservoir and the adjacent areas of impacted
groundwater and contaminated soils and the upland disposal facilities. Land uses along the Clark
Fork River riparian zone are primarily recreational and agricultural. The Clark Fork River in the
vicinity of MRSOU is used for recreational rafting, kayaking and fishing. The City of Missoula
(population 57,000) is located approximately 7 river miles downstream of Milltown, Montana.

Assisted by an EPA Superfund Redevelopment Initiative pilot grant and EPA support,
communities near the MRSOU developed a reuse plan. The plan called for the creation of a state
park with trails, river access, wildlife habitat and interpretive areas celebrating the region's
history and heritage. In 2010, the State of Montana acquired portions of the MRSOU to become
a new state park. The state allocated funding for the park's development and land acquisitions.
There are several trails in the area and the state has plans to link the new park with the larger
community trail network and the newly renovated pedestrian bridge.

About 16,240 people live in the area of CFROU (OU3) according to 2010 U.S. Census data.
Approximately 28 percent of the population (4,500 people) lives in or near Reach A.
Approximately 89 percent of the land within Reach A is privately owned; the remaining 11
percent of the land is managed by federal and state agencies. Land use in the CFROU consists of
residential use, agricultural use and recreational use. The town of Deer Lodge is located within
and adjacent to the OU.

3.3 History of Contamination

In the Butte area, mining companies routinely disposed of mining and milling wastes containing
various amounts of unrecovered metals and arsenic into local creeks in the headwaters of the
Clark Fork River Basin from the late 1860s to well into the 20th century. These streams
conveyed the mining and milling wastes downstream to the Clark Fork River. With the
introduction of electricity in the early 1900s, milling practices improved and new mining
practices significantly increased ore production and metals recovery rates, and substantially
increased the volume of annual mine and mill tailings. These wastes subsequently mixed with
other stream sediments and were carried down Silver Bow Creek and into the upper Clark Fork.

In 1908, a major flood event mobilized large quantities of metals and arsenic-contaminated
sediments from the upper Clark Fork River channel and floodplain, transporting large quantities
of waste to the recently constructed Milltown Reservoir. Much of the arsenic and metals
contaminated sediment was deposited in the reservoir backwater area created by the dam.

Between 1918 and 1959, a series of settling ponds (known as Warm Springs Ponds, now part of
the Silver Bow Creek Superfund site) were built near the end of Silver Bow Creek, to better
control the contaminated sediments entering the upper Clark Fork River. As a result, the amount
of contaminated sediments from the Butte and Anaconda area reaching the Milltown Dam and
reservoir after 1918 significantly lessened. However, substantial quantities of mine waste

12


-------
continued to be washed downstream towards the reservoir from previously deposited areas
downstream of Warm Springs Ponds and the Anaconda area as well as output from the ponds.

In addition to fluvial deposition of metals-contaminated sediments in the historic 100-year
floodplain, agricultural fields were irrigated with water from the Clark Fork River that at times
contained elevated concentrations of metals in dissolved form and as suspended sediment. This
caused ongoing contamination, at low levels, of the fields. In some instances, irrigation ditches
overflowed or were breached, flooding and contaminating fields downgradient of the ditches
with river water. The irrigated fields are located on terraces above the influence of metals and
arsenic impacts associated with flood deposition.

3.4	Initial Response

In 1981, local public health authorities found arsenic in drinking water wells in the Milltown area
at concentrations exceeding the federal drinking water standard. EPA added the Site to the
National Priorities List (NPL) in 1983. Also in 1983, the Atlantic Richfield Company (ARCO)
suspended its mining activity in Butte after shutting down the Anaconda smelter.

In 1984, EPA issued an interim record of decision (ROD) for OU1. A resulting fund-lead
response action installed a new drinking water system for Milltown (i.e., a water supply well).
However, no institutional controls were put in place at that time. The Montana Power Company,
a predecessor of the Northwestern Corporation, implemented rehabilitation and upgrades to the
Milltown spillway and dam from 1986 through 1990, and 14,500 cubic yards of reservoir
sediments and debris were transported and encapsulated in the Upland Disposal site (near MW
913 A, Figure 2). An earlier disposal site had also been constructed on site by the Montana Power
Company.

In 1989, the United States sued ARCO for reimbursement of response costs at three of the NPL
sites listed above. In 1991, EPA issued an Administrative Order on Consent to ARCO initiating
the remedial investigation and feasibility study (RI/FS) process for the MRSOU.

From 1994 to 1995, EPA issued an Administrative Order on Consent to ARCO initiating the
RI/FS process for the CFROU. In 2000, EPA issues a time-critical removal action memorandum
and a Unilateral Administrative Order to ARCO to address immediate human health risks for
residents of Eastside Road in Deer Lodge, in response in part to an Agency for Toxic Substances
and Disease Registry health consultation and EPA Human Health Risk Assessment action levels.

13


-------
Figure 1: Site Map

llSSOUi

Milltown

Milltown Reservoir

Sediments OU

- -

Clark Fork River OU

Reach C/
Reach B-
Reach A -

Drummond.

DeerJ-'odge

¦Anaconda

Silver Bow
Superfund Site

Butte

Sources: Esri. DigitalGlobe. GeoEye, i-cubed, USDA, AEX, Getmapping. Aerogrid, IGN. IGP. DeLorme, AND. Tele Atlas,
First American, UNEP-WCMC, USGS, the GIS User Community, Clark Fork River Operable Unit Monitoring Report for 2011
and Fourth Five-Year Review Report for Silver Bow Creek/Butte Area Superfund Site.

SOLUTIONS

o

NORTH

Milltown Reservoir Sediments Superfund Site

Milltown, Missoula County, Montana

Disclaimer: This map and any boundary lines within the map are approximate and subject to change. The map is not a survey. The map is for
informational purposes only regarding EPA's response actions at the Site.

14


-------
Figure 2: Detailed Site Map - MRSOU

922D

921A

[917B

H&A2

Right Bank
[Repository! j

ilplOB.

TPR Railroad
jEmbarikment
and Buttress

913A/913B

Tunnel Pond?
Repository

Sources: Esri, DigitalGlobe, GeoEye, Earthstar
Geographies, CNES/Airbus DS, USDA, USGS,
AEX, Getmapping, Aerogrid, IGN, IGP, swisstopo,
the GIS User Community, USDA National
Agriculture Imagery Program, and Water Supply
and Milltown Reservoir Sediments Operable Units of
the Milltown Reservoir Sediments/Clark Fork River
Superfund Site First Five-Year Review Report.

Legend

i—I Milltown Reservoir Sediments OU
' ' Boundary

Clark Fork River OU Boundary

Monitoring Wells

Approximate Historical Extent of
Arsenic Groundwater Plume

Milltown Reservoir Sediments Superfund Site

Milltown, Missoula County, Montana

Disclaimer: This map and any boundary lines within the map are approximate and subject to change. The map is not a survey. The map is for
informational purposes only regarding EPA's response actions at the Site.

15


-------
3.5 Basis for Taking Action

MRSOU

EPA, in consultation with MDEQ, provided oversight of the MRSOU RI/FS activities conducted
by ARCO. The 1993 baseline human health risk assessment for the MRSOU was prepared to
assess potential risks at the Site using standard EPA health risk assessment methods for
residential and recreational uses. EPA determined that the non-carcinogenic and carcinogenic
risks associated with consuming groundwater contaminated with arsenic were unacceptable.
Other exposure pathways for humans - including residential use for existing homes near the
reservoir and recreational use of land surrounding the reservoir - were considered not significant.
If residential use of land immediately surrounding the reservoir occurred, it would be
unacceptable. The analysis of a potential detoxification threshold for ingestion of arsenic
suggested that long-term exposures at the Site, other than through consumption of impacted
groundwater, would not be associated with a greatly increased non-cancer and cancer risk.

The ecological risk assessment determined the water quality downstream exceeded the water
quality criteria and that copper caused an unacceptable acute risk to aquatic life. Additionally,
the ecological risk assessment determined that normal high-flow events may pose an intermittent
low-level chronic risk to fish because of the combined impacts of copper and other metals in the
water column and copper in ingested macroinvertebrates.

CFROU

The primary sources of contamination are tailings and tailings mixed with soil in streambanks
and the historic floodplain. Contaminants move from tailings and impacted soils through the
process of erosion, directly into the river and other surface waters. In addition to erosion of
tailings and impacted soils, metals and arsenic can be leached directly from the tailings and
contaminated soils into groundwater and surface water.

The CFROU 1998 human health risk assessment identified arsenic as the contaminant of concern
(COC) for potential human health risks in Reach A. The RBCs for residential, recreational, and
agricultural exposure are listed below. These RBCs are for arsenic concentrations in soils, as
averaged over exposure units. EPA considers acceptable exposure levels to be concentration
levels that represent an excess upper bound lifetime cancer risk to an individual of between 10"4
(1 in 10,000 probability) to 10"6 (1 in 1,000,000 probability), with 10"6 as the point of departure.
EPA proposed the following arsenic concentrations, which represent a 10-4 excess cancer risk:

Re si denti al	150 mg/kg

Recreational	680 mg/kg (children at Arrow Stone Park and other recreational scenarios)

1,600 mg/kg for fishermen, swimmers and tubers along the river
Rancher/Farmer 620 mg/kg

16


-------
On historically irrigated lands, however, where residential development has occurred or where it
may occur in the future, the risk assessment concluded that risks may be unacceptable.

The CFROU ecological risk assessment found unacceptable risks from the metals contamination
to plants and aquatic life within the several reaches of the CFROU. Soils and vegetation areas
most clearly show the impacts from these risks. In addition, United States Geological Survey
(USGS) studies found excessive rates of erosion along streambanks in the upper reaches of the
CFROU. The studies also identified the possibility of severe erosion of the upper river in large
floods that would cause large inputs of contaminants and sediment into the river.

4.0	Remedial Actions

In accordance with CERCLA and the NCP, remedial actions are required to protect human
health and the environment and to comply with applicable or relevant and appropriate
requirements (ARARs). A number of remedial alternatives were considered for each OU at the
Site, and final selection was made based on an evaluation of each alternative against nine
evaluation criteria that are specified in Section 3 00.43 0(e)(9)(iii) of the NCP. The nine criteria
are:

1.	Overall Protection of Human Health and the Environment

2.	Compliance with ARARs

3.	Long-Term Effectiveness and Permanence

4.	Reduction of Toxicity, Mobility or Volume through Treatment

5.	Short-Term Effectiveness

6.	Implementability

7.	Cost

8.	State Acceptance

9.	Community Acceptance

4.1	Remedy Selection

Milltown Water Supply QUI

EPA issued an interim ROD in 1984, and amended this action in 1985. A resulting response
action installed a new drinking water system for Milltown. This OU1 was combined with OU2.

17


-------
MRS0U2

In December 2004, EPA signed the final ROD for the MRSOU. Media-specific remedial action

objectives (RAOs) include:

Groundwater

•	Return contaminated groundwater to its beneficial use within a reasonable timeframe
and prevent ingestion until drinking water standards are achieved.

•	Comply with state groundwater standards, including non-degradation standards.

•	Prevent groundwater discharge containing arsenic and metals that would degrade
surface waters.

Surface Water

•	Achieve compliance with surface water standards, unless a waiver is justified.

•	Prevent ingestion of or direct contact with water posing an unacceptable human
health risk.

•	Achieve acute and chronic federal Ambient Water Quality Criteria (AWQCs), as well
as State water quality standards.

The selected remedy for the MRSOU consists of the following measures:

•	Initiating the process of progressively dewatering Milltown Reservoir Sediment
Accumulation Area (SAA) I sediments by lowering reservoir surface water levels
through use of the existing radial gate and spillway with panels removed (see
Appendix H for map of SSAs).

•	Isolating SAA I sediments from flowing surface water by excavating a bypass
channel through SAA I and armoring the existing embankment along the Blackfoot
River boundary of SAA I and converting powerhouse inlets to low level outlets
removing the spillway section of the Milltown Dam.

•	Removing the radial gate, powerhouse, dividing block, shop and right abutment
gravity wall sections of Milltown Dam as part of integration with the Natural
Resource Damage Program (NRDP) Trustee Restoration Plan.

•	After a period of dewatering and consolidation, remove down to a predetermined
contour surface the sediments in SAA I through the use of mechanical excavation
techniques, hauling the waste (approximately 90 miles via rail cars), and placing the
sediments removed from SAA I in the Opportunity Ponds at the Anaconda Smelter
site.

•	Reconstructing the Blackfoot River and Clark Fork River channels and banks,
including protection of certain infrastructure and regrading/revegetating the Clark
Fork River/Blackfoot River floodplain to provide stability.

•	Replacement of any drinking water supply that exceeds the drinking water standard
for arsenic of 10 micrograms per liter (|ig/L) due to remedial action implementation
(if appropriate, a temporary controlled groundwater area will be established until the
Milltown aquifer recovers using monitored natural attenuation).

•	Replacement or retrofitting of domestic wells which are deemed unusable by EPA
because of the lowering of the groundwater table.

18


-------
•	Conducting long-term operation, maintenance and monitoring of the areas identified
as the dam rehabilitation sediment/debris repositories established by the Montana
Power Company, the portions of the new Interstate-90 embankment outside the
Montana Department of Transportation's right-of-way, and the area in the lower
Clark Fork River channel (SAA Ill-b) where sediments with elevated concentrations
of arsenic and metals will remain after the remedial action and any other on-site
repositories established during the remedial action and any other waste repositories
established on site.

•	Bridge stability mitigation for certain bridges near the MRSOU.

•	Monitoring and maintenance of borrow and staging areas revegetated during remedial
action.

•	Surface water and groundwater monitoring.

•	Implementation of additional best management practices or engineering controls as
detailed in a contingency plan to be approved by EPA or as otherwise required by
EPA, in consultation with MDEQ, if temporary construction-related surface water
quality standards are exceeded.

•	Implementation of the terms and conditions of the incidental take statement in the
United States Fish and Wildlife Service's (USFWS's) Biological Opinion, and
wetlands mitigation as necessary to meet the no-net-loss requirement as determined
by USFWS.

The OU2 2004 ROD indicates that groundwater standards are expected to be met within four to
10 years following completion of dam and sediment removal. The remedial action construction
was significantly completed in June 2012. Cleanup goals are listed in Tables 2 and 3.

Table 2: MRSOU Groundwater COC Cleanup Goals

(irouiulwiilcr ( ()(

KOI) ( loiiniip (ioiil (ji»/l.);l

Arsenic

10

Cadmium

5

Copper

1,300

Lead

15

Zinc

2,000

Notes:

a. Based on the more stringent of federal or state standards.

19


-------
Table 3: MRSOU Surface Water COC Cleanup Goals



A(|ii;ilic l ilc

1 liniiiiii 1 k;il 111

<¦<>(¦









AciiU'
(iiii/l.l

(h runic
(iiii/l.l

Siiindiird (iiii/l.)

Arsenic

340

150

10 - federal
18 - state

Cadmium

2.10

0.27

5

Copper

13

9

l^OO1

Lead

81

3.2

15

Zinc

119

119

2,000

The ROD also identified the need for groundwater institutional controls for the MRSOU. The
institutional controls would include:

•	Continued funding for maintaining the existing replacement water supply for Milltown
residents (installed under the OU1 remedy).

•	Make contingency funds available to reconfigure, expand or update replacement water
supplies.

•	If needed, establish a controlled groundwater area to ban future wells within or
immediately adjacent to the arsenic plume.

•	The ROD also identified the need for institutional controls to prevent residential use of
the MRSOU and to protect disturbance on-site remedial elements such as disposal units.

CFROU3

In April 2004, EPA signed the final ROD for the CFROU. The 2004 RAOs for floodplain
tailings and impacted soils are:

•	Prevent or inhibit ingestion of arsenic-contaminated soils/tailings where ingestion or
contact would pose an unacceptable health risk.

•	Prevent or reduce unacceptable risk to ecological (including agricultural, aquatic, and
terrestrial) systems degraded by contaminated soils/tailings.

The groundwater RAOs are:

•	Return contaminated shallow groundwater to its beneficial use within a reasonable
period.

•	Comply with state groundwater standards, including nondegradation standards (Table

4)-

•	Prevent groundwater discharge containing arsenic and metals that would degrade
surface waters.

1 The MRSOU ROD acknowledges that a waiver of the State standard for copper in the upstream operable unit, and
allows for consideration of upstream input into the MRSOU in determining compliance with the copper ARAR.

20


-------
For surface waters, the RAOs are:

•	Reduce or eliminate "pulses" of metals to the river, including those caused by
snowmelt and thunderstorm events.

•	Achieve compliance with surface water standards, unless a waiver is justified (Table

5).

•	Prevent ingestion of, or direct contact with, water posing an unacceptable human
health risk.

•	Achieve trout toxicity reference values and acute and chronic federal AWQCs.

•	Comply with storm water ARARs.

The selected remedy will be implemented along the erosive streambanks and the historic 100-
year floodplain of all of Reach A and small, localized areas of Reach B. The remedy for Reach C
is no action.

The remedy is currently under construction (see Section 4.2). The remedial actions will proceed
in localized efforts and require about 15 construction seasons to complete. The sequence of
properties to be remediated throughout Reach A and localized areas of Reach B will be carefully
planned and prepared. While the general approach will be to work from the headwaters down,
EPA and MDEQ believes remediation can be done more quickly and effectively and with less
threat to river stability by working on discontinuous stretches of the river. Thus, properties will
be engaged in a discontinuous manner to prevent jeopardizing the integrity of the floodplain,
should a flood event greater than the annual flood occur during the 15-season remedial action
period. Affected landowners will be involved in setting these schedules and clearly informed of
the sequencing of the work.

Specific components of the remedy, as described in the 2004 ROD, include:

•	In most instances, impacted soils and vegetation, also referred to as impacted areas,
will be treated in place, using careful lime addition and other amendment as
appropriate, soil mixing and revegetation.

•	Some impacted areas will be removed, where depth of contamination prevents
adequate and effective treatment in place, where saturated conditions make in-situ
treatment unimplementable, or where post treatment arsenic levels, after one
retreatment attempt, remain above the human health cleanup level for the current or
reasonably anticipated land use. Severely impacted soils, also known as slickens, will
be removed and revegetated.

•	Residential soils above residential action levels will be removed.

•	The Riparian Evaluation System (RipES) process will be used in remedial design to
identify severely impacted areas and impacted areas, and areas where the exceptions
to removal or in-situ treatment will apply.

•	Streambanks will be stabilized primarily by "soft" engineering (with limited hard
engineering where conditions warrant) for those areas classified and an approximate,
flexible 50-foot riparian buffer zone will be established on both sides of the river.

•	Opportunity Ponds will be used for disposal of all removed contamination.

•	Weed control for in-situ treatment, streambank stabilization, and removal areas is
required.

21


-------
•	Best management practices (BMPs) throughout Reach A and in limited areas of
Reach B are required to protect the remedy and ensure land use practices are
compatible with the long-term protection of the selected remedy.

•	Institutional controls and additional sampling, maintenance and possible
removal or in-situ treatment of contamination, including the Trestle Area, will be
required to protect human health.

•	Monitoring during construction, construction BMPs and post-construction
environmental monitoring are required.

•	The remedy is also modified and expanded for the Grant-Kohrs Ranch National
Historic Site, located in Reach A.

Table 4: CFROU Groundwater COC Cleanup Goals

Cmunriwsilcr ( ()(

KOI) ( loiiiiup (ioiil Uiii/I.)

Arsenic

10

Cadmium

5

Copper

1,300

Iron

300

Lead

15

Zinc

2,000

Table 5: CFROU Surface Water COC Cleanup Goals

Surl'siceWsilerCOC

A(|iisilie
l.il'e -
Acule

A(|iisilic

l.il'e -
Chronic

(UJi/l.)

1 III I11SII1 llcsillh

(llli/l.)

Arsenic

340

150

10 - federal
18 - state

Cadmium

2

0.25

5

Copper

13

9

1,300

Lead

81

3.2

15

Zinc

119

119

2,000

The risk-based soils cleanup goals for arsenic at residential, recreational and agricultural areas
are listed in Table 6. These goals are for arsenic concentrations in soils, as averaged over
exposure units.

22


-------
Table 6: Arsenic Soil Cleanup Goals

l.iliul I sc

KOI) (k';iilli|) (>o;il

(in li/Kii)

Residential

150

Recreational

680 for children at Arrow
Stone Park and other
recreational scenarios

1,600 for fishermen,
swimmers and tubers along
the river

Rancher/Farmer

620

Notes:

mg/kg = milligrams per kilogram

2015 Explanation of Significant Differences (ESD)

A review of post-ROD sampling of the CFROU and the results of EPA's 2007 RipES mapping
for the floodplain tailings and soils component of the remedy led to an ESD for the CFROU in
2015. The ESD provides for the use of the RipES process as a tool in development of the
remedial design. However, sampling and field observations relating to vegetation health and
other factors (groundwater, riparian vegetation, contaminant sampling, ownership, infrastructure,
land use and site specific remedy requirements), showed that use of RipES determination alone
would not lead to implementation of ROD requirements or fully meeting RAOs. This ESD
changed the scope of the floodplain tailings and soils component of the remedy described in the
ROD by adding factors that will be considered during remedial design to determine whether
removal, in-situ treatment or other remediation (e.g., best management practices, institutional
controls) is appropriate for a given area.

4.2 Remedy Implementation

Milltown Water Supply (QUI)

OU1 is now part of the MRSOU (OU2). The Milltown Water Supply OU focused on providing a
safe water supply to area residents through establishment of a public water supply system for the
town of Milltown. EPA funded the replacement of one public water supply used by Milltown
residents as part of the OU1 remedy and provided funding for maintenance of this water supply
well. The PRPs eventually provided permanent maintenance funding to the Milltown Water
User's Association for this system. EPA also funded the MCCHD to distribute arsenic test kits to
interested residents who wanted to test their private well water. If tests showed exceedance of
standards, the Settling Defendants provided for the hookup by these residents to the replacement
water supply. The 2004 MRSOU ROD continued funding for maintaining the existing
replacement water supply for Milltown residents and made contingency funds available to
reconfigure, expand or update replacement water supplies.

23


-------
MRSOU (OU2^

Reservoir Drawdown and Dam Removal

Remedial design began on July 18, 2005. In August 2005, the PRPs signed a Consent Decree,
allowing the project to move out of the planning phase and into remedial action. Remedial action
began on February 15, 2006. The initial remedial activity was to lower the water level in the
reservoir to dewater the SAAI sediments, facilitate dam removal and ultimately enable the use
of mechanical excavation techniques for sediment removal. Removal of the Milltown Dam
spillway and ultimate removal of the rest of the dam took place concurrently with reservoir
drawdown. PRP contractors completed final dam removal in March 2009.

Dam removal lowered the groundwater table in the Milltown area, which raised the possibility
that shallow water supply wells in the Milltown area could go dry. Therefore, EPA managed a
well-replacement program as part of the remedial action starting in 2006. Based on the modeling
results, EPA replaced 82 private and small public water supply wells in the Milltown area and
reconfigured numerous additional wells.

Sediment Dewatering, Removal and Relocation

The RI/FS phase of the project evaluated metals contaminant concentrations in sediments in the
Milltown reservoir. Only those sediments shown to be contributing directly to existing
groundwater degradation (sediments with the highest pore water contaminant concentrations),
and with the potential to contribute to future surface water degradation were removed to meet
remedial objectives. Reservoir sediments were divided into two sections: the upper and lower
reservoir SAAs. These two reservoir sections were further divided into sub-areas based on
sediment accumulation features. The lower reservoir consists of SAAs I, II and III. The upper
reservoir encompasses SSAs IV and V. In 2007, sediments in SAA I were removed and isolated
from the Clark Fork River channel.

To facilitate reservoir sediment removal, EPA required a bypass channel for the Clark Fork
River along the northern boundary of SAA I. Beginning in May 2007, approximately 584,000
cubic yards of reservoir sediment, 40,000 cubic yards of underlying soil material and 57,000
cubic yards of underlying alluvium were excavated to form the bypass channel. Excavated
reservoir sediment was relocated by rail transport to Opportunity Ponds. The bypass channel was
completed in early 2008. The excavation of SAA I sediments finished in September 2009; a total
of 2,331,956 cubic yards of sediment was removed and disposed of at the Opportunity Ponds
disposal area at the Anaconda Smelter site. The Clark Fork River was re-diverted to the
reconstructed channel in December 2010. EPA funded or performed bridge stability actions for
three bridges, and a fourth bridge was addressed by its owner.

The PRPs constructed two repositories to contain debris from the demolition of the dam and
SAA Ill-b and SAA IV sediments. One repository is located just downstream of the removed
right abutment of the dam (the Right Bank Repository). The other is the Tunnel Pond
Repository. Groundwater monitoring of the Tunnel Pond Repository will entail sampling one
well, located downgradient of the repository, at the same frequency and for the same analyte list

24


-------
as the other point of compliance (POC) wells. No groundwater monitoring is required for the
Right Bank Repository.

In addition to the two constructed repositories, two other repositories were present prior to
remedial action. Disposal Site No. 1 was removed as part of the work to place SAA Ill-b
sediments in the Tunnel Pond Repository. At the second, the Upland Disposal site, the State of
Montana built a new repository on top of the Upland Disposal site in which to store a portion of
the sediment excavated during implementation of restoration actions from SAA IV and V.
Maintenance and monitoring of disposal areas remains the responsibility of the PRPs, according
to the 2013 long-term monitoring plan.

Compliance wells are located within the current arsenic plume and were monitored during the
remedial action to track progress in restoring the Milltown alluvial aquifer. A series of early
warning wells located around the fringe of the plume and along the Clark Fork River
downstream of the MRSOU are also monitored to ensure that groundwater in existing drinking
water wells was not unacceptably impacted by construction activities. Finally, MCCHD monitors
certain existing public and private water supply wells as public health monitoring wells. Data
available for this FYR (2013) consistently indicate no arsenic exceedances in sampled wells.

The State of Montana Natural Resource Damage Program followed PRP construction activities
with channel construction, revegetation and reconstruction of the floodplain, revegetation, and
development of wetlands. Some of these actions are required to meet certain remedial goals and
objectives. Operation and Maintenance of this work is ongoing.

CFROU3

The majority of the CFROU is Reach A, a 43-mile stretch of the river from Warm Springs in
Anaconda/Deer Lodge County downstream to Garrison in Powell County. In accordance with
the 2004 ROD, in 2006 and 2007, while Consent Decree discussions were in progress, EPA
performed RipES mapping for the floodplain tailings and soils component. MDEQ began its
remedial design activities in 2008, following entry of the Consent Decree, which designated
MDEQ as lead agency for remedy and O&M implementation using cashout funds received from
the PRP. MDEQ focused its first remedial actions on immediate human health and irrigated
lands concerns and are now proceeding with geographically-defined phases (Figure 3).

25


-------
Figure 3. CFROU Reach A Phase Breaks

Upper Clark Fork
River Reach A

11§ Phase breaks

100 year floodplain

Land Ownership

Private

Local Government
State Government

Federal
Other

Dempsey

_POWELL COUNTY
DEER LODGE~COUN~

Galen

Warm Springs

26


-------
MDEQ, in consultation with EPA, and in accordance with Consent Decree requirements,
performed residential yard removals, necessitated by elevated levels of arsenic and lead, in the
fall of 2010 through the summer of 2011. Confirmation sampling were collected to ensure all
contamination was removed. MDEQ, in consultation with EPA and in accordance with Consent
Decree requirements, performed the Trestle Area cleanup within Reach A in the fall and winter
of 2011-2012, with planting in the spring of 2012. The trestle cleanup involved removal of
residential soils with elevated levels of arsenic and reconstruction and revegetation of 1,000 feet
of streambank. In the fall and winter of 2012, MDEQ performed the remedial action for the
pasture areas historically irrigated with Clark Fork River water.

The Reach A Phase 1 Remedial Action Project began on March 4, 2013, and finished on April 4,
2014. MDEQ, NRDP and EPA performed a pre-final inspection of the project on May 9, 2014.
Additional vegetation was planted in April, May and the fall of 2014. Revegetation activities are
ongoing. Monitoring plans for vegetation and streambanks have been developed to ensure that
the remedy is successful over the long term. MDEQ has prepared the Construction Completion
Reports for Phase 1.

Additional activities underway in Reach A include:

•	Phases 5 and 6 In Progress

MDEQ submitted the final Reach A, Phase 5 & 6 Data Summary Report to EPA on
March 14, 2014. Remedial actions began on July 15, 2014, and are ongoing. Phase 5 and
6 involve two private landowners and cleanup on working ranches. The remediation
project will consist of tailings removal on 4.5 river miles. The work is scheduled to be
completed in the spring of 2016, with revegetation activities in the spring of 2016 and fall
of 2016.

•	Phase 2 -InProgress

MDEQ submitted the Preliminary Design Plan for Reach A, Phase 2 to EPA on July 1,
2014. Construction began in the summer of 2015. Phase 2 involves two private
landowners and State of Montana land. The privately-owned property is actively farmed
and ranched. The remediation project will consist of tailings removal on 1.9 river miles
and is scheduled to be completed by the fall of 2016 with revegetation activities to
follow.

•	Phases 3 and 4 - Preliminary Design

Sampling and characterization of the Phases 3 and 4 project areas, located between
Perkins Lane and Galen Road, was completed in the winter of 2015. The Preliminary
Design Plan has been developed and remedial activities are anticipated to start in fall of
2016.

•	Phases 7, 15 and 16 Preliminary Design

MDEQ is currently working with private landowners, Montana Fish, Wildlife and Parks,
and the Grant-Kohrs Ranch on design plans. These plans begin to lay out the details of
the design and how and where remedial work will be conducted. MDEQ will continue to
provide updates as designs progress.

•	Phase 8 - Sampling and Analysis

Phase 8 is currently in the site characterization phase. Crews are digging test pits and
sampling material to determine the extent and depth of contamination along the river and

27


-------
surrounding corridors. Sampling should be completed in early 2016, and the design team
will then begin the design process for remedial action.

• Eastside Road Pastures

Remedial Action for the Eastside Road Pastures began on November 5, 2012. The
majority of work finished on December 6, 2012; fencing finished in the spring of 2013.
MDEQ conducted additional sampling of this area during the spring and summer of 2014.
After a year of little growth in the Eastside Road pastures south of Deer Lodge, MDEQ
implemented additional revegetation measures in the spring of 2015. Sugar beet lime and
top soil was deep tilled into the existing soil. The area was then reseeded and straw was
crimped into the ground for erosion control. Monitoring of this area is ongoing.

Reach C was determined to require no further action. Remedial design work on Reach B is
expected to occur after work is completed on Reach A. Institutional controls for the CFROU are
discussed in Section 6.3.

MDEQ will develop appropriate operation and maintenance plans and best management practice
ranch plans on a parcel-specific basis as the cleanup proceeds. An Institutional Control
Implementation and Assurance Plan will also be developed.

4.3 Operation and Maintenance (O&M)

MRSOU

The Long-Term Post Remedial Action Construction Monitoring Plan, which is the MRSOU
operation and maintenance plan, was finalized in 2013. The plan outlines the groundwater and
surface water monitoring requirements as well as the long-term maintenance and monitoring for
the constructed repositories and buttress areas. Prior to the 2013 plan, monitoring was performed
under the 2007 Remedial Action Monitoring Plan. Groundwater is to be sampled twice each
year, during high and low flow.

Surface water sampling occurs at three sites, six-to-eight times per year on a USGS schedule
designed to take seasonal and hydrologic variability into account. Suspended-sediment samples
are collected by an observer two to 14 times per week, depending on season and flow conditions.
Bed sediment data is collected once annually during low, stable flow conditions (typically
around August). Biological data is collected once annually, on the same dates as the bed
sediment data collection.

The PRPs are responsible for annual maintenance and monitoring of two repositories - Tunnel
Pond and Right Bank. Annual monitoring and maintenance of the buttress and railroad berm
adjacent to the Tunnel Pond Repository and the Interstate-90 slope and buttress are also the
responsibility of the PRPs. Operation and maintenance costs for MRSOU were not available for
review during this FYR.

28


-------
CFROU

The Interim Comprehensive Long-Term Monitoring Plan for the CFROU established monitoring
activities for sediment, surface water and groundwater that will determine the environmental
effectiveness of remediation and restoration actions within the Site as they are implemented over
the next 15 years. The CFROU remedy is intended to remove threats to human health and the
environment posed by mining related contaminants within the floodplain of the upper Clark
Fork. Monitoring under the Interim Comprehensive Long-Term Plan began in the spring of 2010
at each of six Clark Fork monitoring stations, this was prior to initiation of any remediation and
restoration actions within the CFROU. This plan has been updated yearly.

Eventually, a long term operation and maintenance plan will be developed and implemented by
MDEQ.

A breakdown of CFROU costs from 2008 to 2014 were provided and reviewed. Since remedial
actions are still being designed and implemented at the CFROU, separate O&M costs are not
presented. The remedial action at Phase 1 was completed in fiscal year 2014. The next FYR may
examine O&M costs for ongoing maintenance at this phase and any others completed at that
time.

5.0 Progress Since the Last Five-Year Review

The protectiveness statement from the 2011 FYR for the Site stated:

The remedy at the MRSOU is expected to be protective of human health and the environment
upon completion, and in the interim, exposure pathways that could result in unacceptable risks
are being controlled. The Water Supply Operable Unit is fully implemented andfunded, and is
protective of human health and the environment. The long-term protectiveness of the remedial
action will be verified through review and approval of remedial action completion documents, a
comprehensive O&M Plan, an Institutional Control Plan, and through monitoring of
groundwater for all of the ARARs, and periodic evaluation of the O&M results and the
institutional controls. Streambank reconstruction and area revegetation efforts should be
evaluated in the next FYR Report.

The 2011 FYR included five issues and recommendations for the MRSOU2. This report
summarizes each recommendation and its current status below.

2 Because work at the CFROU was in its initial stages, that OU was not evaluated in the 2011 FYR.

29


-------
Table 7: Progress on Recommendations from the 2011 FYR

Kccommciuliilions

Psirlj

Responsible

Milcslonc
l);Mc

Action T;ikcn iinri
On Iconic

Dsilc of
Action

Implement institutional







Not

controls for the MRSOU
comprehensive

PRP/State/EPA

September
2014

Ongoing.

completed
and carried

institutional control plan
and its components.





over to the
2016 FYR

Develop and implement









O&M requirements









through a comprehensive









O&M plan. This plan
should add a requirement
for routine surveying of



September
2013

Completed. Envirocon
completed the Long-
Term Post-Remedial



the Tunnel Pond

EPA/PRP

Action Construction

03/15/2013

Repository berm to
verify that lateral



Monitoring Plan for
MRSOU.



movement is not









occurring over time.









Other requirements may









also be necessary.









Include monitoring for
all of the groundwater
ARARs, and in a long-
term groundwater and
surface water

EPA/PRP

2012

Completed.

Monitoring of all of
the groundwater
ARARs began in 2013
and additional
parameters included in
the long-term

03/29/2013

monitoring plan.





groundwater and
surface water
monitoring plan.



Remove and









appropriately dispose of
contaminated wood
timbers left after dam

EPA/PRP

Fall 2011

Completed. Timbers
removed.

05/03/2012

removal (currently
scheduled for the fall of







2011).









Reclaim and revegetate
borrow area in









accordance with the









requirements of the
statement of work. The
adequacy of vegetation
at the other borrow area

EPA/PRP

September
2012

Completed. Areas
reseeded in the spring
of 2012.

07/20/2012

and the Tunnel Pond









Repository should also
be reviewed.









30


-------
6.0	Five-Year Review Process

6.1	Administrative Components

EPA Region 8 initiated the FYR in October 2015 and scheduled its completion for
September 2016. EPA remedial project manager (RPM) Sara Sparks led the EPA site
review team, and contractor support was provided to EPA by Skeo. In August 2015, EPA
held a scoping call with the review team to discuss the Site and items of interest as they
related to the protectiveness of the remedy currently in place. The review schedule
established consisted of the following activities:

•	Community notification.

•	Document review.

•	Data collection and review.

•	Site inspection.

•	Local interviews.

•	FYR Report development and review.

6.2	Community Involvement

On November 1, 2015, EPA participated in a radio interview that was broadcast on KQRV in
Deer Lodge, Montana. This interview announced the commencement of the FYR process for the
Site and invited community participation in the FYR process. In June 2016, EPA published a
public notice in the Missoulian and the Missoula Independent newspapers providing contact
information for EPA RPM Sara Sparks and inviting community participation in the FYR process
for the Site. The press notice is available in Appendix B. No one contacted EPA as a result of the
advertisement.

EPA will make the final FYR Report available to the public. EPA will place copies of the
document in the designated site repositories: Grant-Kohrs Ranch National Historic Site, 266
Warren Lane, Deer Lodge, Montana 59722 and Missoula City/County Library, 301 East Main
Street, Missoula, Montana 59802. Upon completion of the FYR, EPA will place a public notice
in the Silver State Post, Missoulian and Missoula Independent newspapers to announce the
availability of the final FYR Report in the Site's document repositories.

6.3	Document Review

This FYR included a review of relevant site-related documents. Appendix A provides a complete
list of the documents reviewed.

ARARs Review

Section 121 (d)(2)(A) of CERCLA specifies that Superfund remedial actions must meet any
federal standards, requirements, criteria or limitations that are determined to be ARARs. ARARs
are those standards, criteria or limitations promulgated under federal or state law that specifically
address a hazardous substance, pollutant, contaminant, remedial action, location or other

31


-------
circumstance at a CERCLA site. To-Be-Considered criteria (TBCs) are non-promulgated
advisories and guidance that are not legally binding, but should be considered in determining the
necessary level of cleanup for protection of human health or the environment. While TBCs do
not have the status of ARARs, EPA's approach to determining if a remedial action is protective
of human health and the environment involves consideration of TBCs along with ARARs.

Chemical-specific ARARs are specific numerical quantity restrictions on individually listed
contaminants in specific media. Examples of chemical-specific ARARs include the maximum
contaminant levels specified under the Safe Drinking Water Act as well as the ambient water
quality criteria enumerated under the Clean Water Act. The remedy selected for the Site was
designed to meet or exceed all chemical-specific ARARs and meet location- and action-specific
ARARs.

Groundwater ARARs

The decision documents established federal Maximum Contaminant Levels (MCLs) and
Montana Water Quality Standards as ARARs for groundwater at the Site. Numerical values
listed in decision documents were compared to current federal and state standards to identify any
changes that could affect protectiveness of the remedy (Table 8). The state standard for arsenic is
now the same as the federal standard, which was selected in the 2004 ROD.

Table 8: Previous and Current ARARs for Groundwater COCs



Siiindiii'ds Idcnlifii-d in 2004 KOI)

201(> Siiindiii'ds

('(impound

Shut (ji»/l.)

l"cdor;d (iiii/l.)

Stale

l"eder;d (uii/l.)1'

Al'bClUC

20

10

10

10

Cadmium

5

5

5

5

Copper

1,300

1,300

1,300

1,300

Lead

15

15

15

15

Zinc

2,000

N/A

2,000

N/A

Notes:

a.	Montana Numeric Water Quality Standards - Circular DEQ-7. February 2012.

b.	Safe Drinking Water Act contaminants and federal MCLs.

Surface Water ARARs

The decision documents established federal AWQCs and Montana Water Quality Standards as
ARARs for surface water at the Site. Numerical values listed in decision documents were
compared to current federal and state standards to identify any changes that could affect
protectiveness of the remedy (Table 9). At the time of the ROD, the State of Montana's surface
water quality standard for arsenic was 18 |ig/L, based on human health, and 20 |ig/L for
groundwater as a drinking water supply. The state standard for arsenic for surface water and
groundwater is now 10 |ig/L, matching the federal standards. No other changes were identified in
this review.

32


-------
Table 9: Previous and Current ARARs for Surface Water COCs



20l(i Surfiicc Wilier Miindiirds

2004 KOI) Siiindiirds

Milled)

ledciiil (2)

Slide (1)

l-cdci'iil (2)

A(|iiiilic Life

11II lllilll
1 k;il 111

(fiii/l.)

CMC
(Aculc)
(3)

CCC
(Chronic)
(4)

\(|iiiilic Life

1 III lllilll
llcilllll

CMC
(Aculc)
(3)

CCC
(Chronic)
(4)

Compound

Aculc

(HS/L)

Chronic
(uii/l.)

(iili/l.l

(Uli/I.l

Aculc

(iili/l.l

Chronic

(iili/l.l

Sliiiuliird

(JiJi/l.)

(iili/l.l

(iili/l.l

Arsenic

340

150

10

340

150

340

150

18

340

150

Cadmium

0.52*

0.097*

5



0.25***

0.52*

0.097*

5



0.25***

Copper

3.79*

2.85*

1,300

N/A

N/A

3.79*

2.85*

1,300

2.337#

1.45#

Iron

N/A

1,000

N/A

N/A

N/A

N/A

1000

300a

N/A

N/A

Lead

13.98*

0.545*

15

65***

2 5***

13.98*

0.545*

15

65***

2 5***

Zinc

37*

37*

2,000

120***

120***

37*

37*

2,000

120***

120***

Notes:

*	= value indicated is for a hardness of 25 milligrams per liter (mg/L) as calcium carbonate (CaC03).

** = value indicated is for a hardness of 50 mg/L as CaC03.

*** = value indicated is for a hardness of 100 mg/L as CaC03.

**** = value indicated is for a hardness of 150 mg/L as CaC03.

#	= standards are hardness dependent. Value indicated is for a hardness of 84.6 mg/L as CaC03. Source:
httD:/A\\\\\.CDa.ao\/\\atcrscicncc/critcria/coDDcr/2007/critcria-rull.Ddr.

a = indicates value is a secondary MCL based on aesthetics (taste, odor, staining).

1.	Montana Numeric Water Quality Standards - Circular DEQ-7. February 2012.

2.	Current National Recommended Water Oualitv Criteria. EPA. htto://www.era.eov/waterscience/criteria/wactable/#mm.

3.	CMC = Criteria Maximum Concentration is an estimate of the highest concentration of a material in surface water to which an aquatic community can be
exposed briefly without resulting in an unacceptable effect.

4.	CCC = Criterion Continuous Concentration is an estimate of the highest concentration of a material in surface water to which an aquatic community can be
exposed indefinitely without resulting in an unacceptable effect.

33


-------
Institutional Control Review

MRSOU

The ROD identified that institutional controls, dealing primarily with groundwater but also
addressing residential use and protection of waste repositories, were required at the Site. To date,
a controlled groundwater area or similar institutional control has not been implemented. Site
regulatory agencies are continuing to discuss the need for this institutional control. A Missoula
County ordinance currently in place appears to preclude installation of new public water wells in
the vicinity of the MRSOU arsenic plume. However, these ordinances do not preclude private
well installation in the plume area (Figure 2). Additional institutional controls may be needed to
control private well installation in the arsenic plume, prevent residential use and protect the
waste repositories and the sediments left in place.

An institutional control preventing river access during certain time periods has been necessary in
the past, and may be needed in the future. The majority of the MRSOU has been designated as a
future Montana State Park. Institutional controls dealing with water consumption, residential use
and the waste repositories will need to be incorporated into the future park design and planning
documents.

CFROU

Institutional controls for the CFROU may include county zoning regulations, deed restrictions,
permanent funding for Arrow Stone Park, and groundwater sampling and use controls.
Environmental monitoring is required during all activities.

The Powell County Overlay District covers the area contaminated by mining and smelting
wastes from operations further upstream in the Butte and Anaconda areas (Figure 4).3 The
Overlay District is intended to ensure that future land use in the Superfund Overlay District is
compatible with the presence of potential contaminants and the various remedial actions required
to remove or isolate those potential contaminants from the environment. Requirements include:

•	Property Development: All use changes and development in the Superfund Overlay
Zone are subject to the securing of a Conditional Use Permit (CUP). All applications
for a CUP or variance in the Superfund Overlay Zone shall include the following
additional information beyond that which is required for any CUP or variance. Where
no remedial structures exist on a site, the application materials shall include arsenic
tests, as required by Powell County, and detailed plans (if necessary) for achieving
compliance with the maximum arsenic level allowed for the proposed use.

•	Groundwater Wells: A development certificate shall be required to drill or dig a well
in the Superfund Overlay Zone. Prior to the issuance of a completion certificate of
any well in this overlay district, the well is required to be tested for coliform bacteria,
arsenic, barium, cadmium, chromium, copper, lead, mercury and nitrate, and the
results of the tests submitted to Powell County. No certificate of compliance shall be

3 http ://powellcountvmt. gov/ez/inner.php?PageID= 1501.

34


-------
issued for any well in which the water exceeds state water quality standards for the
proposed use.

• Notice to Purchasers: Before any parcel or any interest in any parcel in the Superfund
Overlay Zone is conveyed, the following statement shall be placed on the deed,
contract for sale or other instrument of conveyance: "This parcel is within a
Superfund site. A permit must be obtained before any development or construction
covered by these regulations is initiated."

Table 10 lists the institutional controls associated with areas of interest at the Site.

Table 10: Institutional Control (IC) Summary Table

Mcdi;i

ICs
Needed

ICs <"silled
lor in (lie
Decision
Documents

Impsiclcd
Arcsi (s)

l<

()h.jecli\e

Iiislriiiiienl in
I'lsicc

Noles

MRSOU
Groundwater

Yes

Yes

area of
delineated
arsenic
plume

Prevent

consumption

of

contaminated
groundwater.

Missoula County
zoning
ordinances in
place preclude
installation of
new public water
wells in the
vicinity of the
arsenic plume.

Additional
controls may be
needed to
prohibit private
well installation.

MRSOU
Soil

Yes

No

repository
and

sediment
areas

Prevent
activities that
could affect
the integrity of
the remedy.
Prevent
residential use.

None

None

CFROU
Groundwater

Yes

Yes

to be

determined
during
each Phase

Prevent

consumption

of

contaminated
groundwater,
if necessary.

Powell County
Overlay District

ICs could
include county
zoning

regulations, deed
restrictions,
permanent
funding for
Arrow Stone
Park, and
groundwater
sampling and use
controls.

CFROU Soil

Yes

Yes

To be
determined
during
each Phase

Prevent
activities that
could affect
the integrity of
the remedy or
cause

unacceptable
human health
exposures.

Powell County
Overlay District

ICs could
include county
zoning

regulations, deed
restrictions,
permanent
funding for
Arrow Stone
Park.

35


-------
Figure 4: Powell County Overlay District (CFROU)

Disclaimer: This map and any boundary lines within the map are approximate and subject to change. The map is not a survey. The map is for
informational purposes only regarding EPA's response actions at the Site.

36


-------
6.4 Data Review

MRSOU

Groundwater Monitoring

Groundwater monitoring at the MRSOU is designed to meet three objectives: 1) ensure that the
remedy is performing as designed; 2) ensure that the remedy complies with applicable
performance standards; and 3) evaluate the need for additional remedial or O&M activities. In
2013, MRSOU long-term post-remedial action monitoring began, replacing the prior remedial
action monitoring plan. The 2013 monitoring plan revised the number of wells to be monitored
to 12 wells and revised the list of dissolved metals requiring analysis.

Data available for the 10 compliance wells (104A, 921 A, 917B, 922D, 105C, 107A, 110B,
HLA2, 11R and 103B), the Upland Disposal Site monitoring well 913 A, and the Tunnel Pond
Repository monitoring well (TPR10) were sampled during high-flow conditions in June 2015
and low-flow conditions in January 2016. The 2015 and 2016 well samples were only analyzed
for dissolved arsenic in accordance with EPA's April 20, 2015 correspondence, which approved
dropping analysis for the other COCs due to two years of data showing no exceedances of state
standards.

Arsenic concentrations in the compliance wells ranged from 0.867 [j,g/L to 67.4 [j,g/L in the most
recent annual monitoring, with nine (during the June monitoring 2015) and eight (during the
December 2015 monitoring event) of the 12 compliance wells continued to exceed the 10 [j,g/L
groundwater standard. Overall, arsenic concentrations in all wells are lower than historic levels
years (Figures 5-7). The ROD indicates that groundwater standards are expected to be met within
approximately four to 10 years following completion of dam and sediment removal. A waiver of
groundwater standards is not currently proposed. However, the PRPs may seek a waiver of
groundwater cleanup standards if compliance is not achieved and is technically impracticable.

Groundwater monitoring of the Tunnel Pond Repository will entail sampling one well, located
downgradient of the repository, on the same frequency and for the same analyte list as the other
POC wells. No groundwater monitoring is required for the right bank repository. The 2013
monitoring plan identifies that the POC well for the Repository was left as "to be determined"
because some of the past sampling results in the existing monitoring well, TPR10, were above
the pertinent ARAR and the state's 10 ug/L groundwater arsenic performance standard. In a
September 16, 2013 letter, the PRPs proposed using well TPR10 as the Tunnel Pond Repository
POC and evaluating its data using a two-part statistical test to assess potential impacts to
groundwater quality from repository construction and use. The statistics were proposed to
determine if:

1.	The rolling average concentration in the last four samples exceeds the state
groundwater standard.

2.	The Mann-Kendall analysis shows a statistically significant increasing trend in
concentrations in the last eight samples.

37


-------
The results show the rolling average concentration in the last four samples does exceed the
arsenic groundwater standard; the Mann-Kendall analysis does not show a significant increasing
trend in concentrations in the last eight samples. The PRPs continue to recommend statistical
analysis of TPR10 as the Tunnel Pond Repository POC with assessment of potential impacts to
groundwater from repository construction and use.

Figure 5: Arsenic Concentrations in Wells 905,103B, 917B and 107C

5

—J'

38


-------
Figure 6: Arsenic Concentrations in Weils 105C, 11, HLA2,107A and 11R

Figure 7: Arsenic Concentrations in Wells 110B, 907, 922D, 104A, 921A, TPR-10 and 913A

39


-------
Surface Water Monitoring

In 2015, surface water quality samples were collected at all three stations six to eight times on a
USGS schedule designed to describe seasonal and hydrologic variability. Flow was monitored
continuously. An observer collected suspended-sediment samples two to 14 times per week,
depending on season and flow conditions. Bed sediment and biota samples were collected once
in August 2015.

The 2015 surface water quality sample results at the three stations for the five COCs are
summarized on Appendix F. At the downstream Clark Fork River near Missoula station, there
were no exceedances of federal standards and the only exceedances of state standards were for
total recoverable copper in the June 10 sample. Total recoverable copper concentration on this
date was significantly higher at the Clark Fork River at Turah station sample, showing the Site
was not causing the downstream exceedance of state standards. The Consent Decree provides for
the consideration of upstream contamination entering the MRSOU to determine compliance with
surface water standards.

To assist with surface water data evaluation, EPA asked the U.S. Geological Survey (USGS) to
conduct a trends analysis for the Site, using the ongoing data collected by USGS at the Site. The
analysis, title "Water-Quality Trends and Constituent-Transport Analysis for Selected Sampling
Sites in the Milltown Reservoir/Clark Fork River Superfund Site in the Upper Clark Fork Basin,
Montana, Water Years 1996 - 2015" is included as Appendix I.

The primary purposes of this report are to characterize temporal trends in flow-adjusted
concentrations (filtered and unfiltered) of mining-related contaminants and assess those trends in
the context of source areas and transport of those contaminants through the Milltown/Clark Fork
River Superfund Site in the upper Clark Fork Basin. Trend analysis was done on specific
conductance, selected trace elements (arsenic, copper and zinc), and suspended sediment for
seven sampling sites for water years 1996-2015. This report provides an update and supersedes
the trend results reported by Sando and others (2014) for seven sampling sites in the
Milltown/Clark Fork River Superfund Site. This report presents the results and information on
trend-analysis methods, streamflow conditions, and various data-related factors that affect trend
results. This information is presented to assist in evaluation trend results; however, it is beyond
the scope of this report to provide detailed explanations of all observed temporal changes.

Vegetation Inspection and Maintenance

The performance standard for vegetation is to establish on the reclaimed areas a "diverse,
effective and permanent vegetative cover of the same seasonal variety native to the area of land
to be affected and capable of self-regeneration and plant succession at least equal in extent of
cover to the natural vegetation of the area except that introduced species may be used in the
revegetation process where desirable and necessary to achieve the approved post-mining land use
plan. Vegetative cover must be capable of:

• Regenerating under the natural conditions prevailing at the site, including occasional
drought, heavy snowfalls and strong winds.

40


-------
• Preventing soil erosion to the extent achieved prior to the operation."

Another performance standard for vegetation is to control noxious weeds consistent with weed
management criteria developed under MCA 7-22-2109 (2)(b) and to meet the <10 percent
guideline for the amount of cover by noxious weeds.

On June 18, 2015, vegetation performance was assessed on the reclaimed areas for which the
PRPs retain O&M responsibility. The inspection covered over 17 acres and included estimation
of percent vegetative cover, determination of species present (including weed species) and
recommendations for maintenance.

In their approval of the 2014 Annual Report, EPA agreed that vegetation performance standards
had been met for two consecutive years at the Right Bank Repository, the Tunnel Pond
Repository and the Interstate-90 buttress. Observations during the 2015 inspection suggest that
vegetation performance standards for remaining areas (the Bonner Development Group Parcel
and the Sheriff Posse Grounds Parcel) have now also been met for two consecutive years. Based
on this data, the PRPs requested that EPA approve completion of the vegetation performance
monitoring responsibilities, and EPA, in consultation with MDEQ and the State NRD Program
approved of this request.

Repository Inspection and Maintenance

The PRPs visually inspected both repositories, the buttress and railroad berm adjacent to the
Tunnel Pond Repository and the Interstate-90 buttress on June 18, 2015. The PRPs also visually
inspected the Tunnel Pond Repository stormwater conveyance system on May 13, 2015. Overall,
the inspections found the stormwater conveyance systems were clean and functioning and the
repository caps and the Tunnel Pond and Interstate-90 buttresses were in good condition and,
with the exception of a few small subsidence holes observed in the Tunnel Pond railroad
embankment and Right Bank Repository cover, did not show visible impacts from settlement,
subsidence or erosion. Pioneer Technical Services did a geotechnical review of the Tunnel Pond
subsidence holes which determined "these features are not anticipated to impact the geotechnical
stability of the tunnel pond embankment" but they should continue to be observed as part of
annual monitoring.

In addition to the inspections described above, the PRPs also installed settlement monuments in
the crest and toe of the Tunnel Pond Repository embankment in April 2014 as required by the
Monitoring Plan. To support this FYR, the monuments were surveyed on October 28, 2015, to
identify any lateral movement in the embankment. Comparison between the 2014 and 2015
survey results were below the 1-inch trigger for initiating additional review assessment.

Community Well Monitoring

MCCHD monitors certain existing public and private water supply wells as public health
monitoring wells. Data available for this FYR (2013) consistently indicate the groundwater in
these areas remained below the arsenic standard of 10 |ig/L.

41


-------
CFROU

Remediation performance standards were established for the CFROU ROD for surface water,
groundwater and vegetation. No performance standards were established in the CFROU ROD for
aquatic biota (e.g., macroinvertebrates and periphyton), instream sediments or geomorphology.
However, the Sampling and Analysis Plan identifies benchmarks for those environmental media
which may serve to evaluate biological conditions and instream sediment toxicity (Appendix G).
The CFROU monitoring network in 2014 included 14 sites; six mainstem sites and eight
tributary sites. Not all sites were sampled for each environmental medium or for each analyte of
each environmental medium. All of the environmental media monitored in 2014 was to be
monitored in 2015, with the addition of monitoring for birds. Data from 2015 sampling were not
available for this FYR.

Arsenic and copper are the COCs in surface water with regular exceedances. Of 30 samples
collected in the mainstem Clark Fork River in 2014, no samples had zinc concentrations
exceeding the performance goal. One sample had cadmium concentrations exceeding the
performance goal. Four samples had lead concentrations exceeding the performance goal.
However, arsenic commonly exceeded performance goals, particularly in Reach A. Of 24
samples collected in the CFROU in Reach A, 96 percent of them exceeded the dissolved arsenic
performance goal and 46 percent of them exceeded the total recoverable arsenic performance
goal. Mill-Willow Creek and Silver Bow Creek through the Warm Springs Ponds are sources of
arsenic to the Clark Fork River.

Total recoverable copper concentration exceeded the state of Montana chronic aquatic life
standard in the mainstem Clark Fork River sites in 95 percent of the samples collected in the first
and second quarters, but only at Deer Lodge in the third and fourth quarters. The Clark Fork
River reach upstream from Deer Lodge is a major source of copper loading and copper
concentrations throughout the river are strongly related to streamflows.

The highest instream sediment COC concentrations in the mainstem of the Clark Fork River
were typically observed in the uppermost sample sites in Reach A. The lowest concentrations
were typically observed at the downstream-most site at Turah. Concentrations of arsenic, copper,
and zinc exceeded the probable effect concentration (PEC) at all Clark Fork River mainstem
monitoring stations during both sample periods in 2014. Among all sites in the CFROU, arsenic
most commonly exceeded the PEC (88 percent) followed by copper (83 percent), lead (79
percent), zinc (75 percent) and cadmium (50 percent).

6.5 Site Inspection

MRSOU

Site inspection participants included Keith Large from MDEQ, Sara Sparks from EPA, and Treat
Suomi and Claire Marcussen from Skeo. The inspection took place on November 2, 2015. See
Appendix D-l and E-l for the site inspection checklist and photographs.

42


-------
The inspection began at the Milltown Bluff, providing an overall view of the MRSOU remedial
components, including the Tunnel Repository and associated embankment and buttress, Railroad
Grade and Main Repository, the Right Bank Repository, the Interstate-90 slope and buttress, the
Bonner Development Group Parcel and the Sheriff Posse Grounds Parcel. From the bluff,
participants observed areas of sparse vegetation along the gravel road near the Buttress slope; the
area has recently been regraded and seeded to promote growth of vegetation and is flagged for
ongoing monitoring of vegetative growth. The stormwater diversion ditch along the Tunnel
Repository was well maintained. Participants saw that most timber debris from the Milltown
Dam demolition has been removed. However, there were still some timbers near the former dam
area on the north side of the Clark Fork River. EPA later determined that these timbers were
brought in by the Montana Fish, Wildlife and Parks Department for use in park construction.

Participants visited the Right Bank Repository where a relative small area of subsidence was
observed (about 2 square feet) and flagged for ongoing monitoring to ensure the subsidence does
not expand. Participants walked along the Blackfoot River to observe the riprap stabilizing the
banks of the river, and used inclinometers to measure the Interstate-90 bridge settlement. The
riprap was intact. However, a number of timbers were observed below the Interstate-90 bridge
along the banks of the Blackfoot River. These salvaged timbers belong to the Montana Fish,
Wildlife and Parks and will be used for the construction of a State Park near this area.
Participants also viewed monitoring wells 917B and 921 A; both wells were secured with locks.

The participants visited the Bonner Development Group Parcel and the Sheriff Posse Grounds
Parcel. Both parcels appear to have established vegetation. Vegetation was also beginning to
become established along the Clark Fork River southwest of the two parcels. The Sheriff Posse
Grounds Parcel consists of about 3 acres of reclaimed areas. It includes a community park with
picnic tables and trails, a rodeo ground, and a cultural slope area. Apart from the rodeo ground,
all were covered with vegetation. The rodeo ground is currently used for rodeo activities.

CFROU

Site inspection participants included Brian Bartkowiak from MDEQ, Sara Sparks from EPA, and
Treat Suomi and Claire Marcussen from Skeo. The inspection took place on November 3, 2015.
See Appendix D-2 and E-2 for the site inspection checklist and photographs.

The inspection began immediately north of the town of Warm Springs below the Warm Springs
Ponds, the beginning of the Clark Fork River Phase 1 remediation area. The riverbanks have
been remediated and are vegetated. An 8-foot fence was observed; it is intended to keep wildlife
away from the new growth along the riverbank until the vegetation is well established. The
inspection proceeded by car. MDEQ staff noted the location of the Phase 2 remediation area
where remedy construction started on June 2015. Participants observed the Beck Borrow area
where clean fill material is obtained and then mixed with compost for use in filling the excavated
floodplain areas (located west of the Phase 10 area).

Participants proceeded to the town of Deer Lodge to view Arrow Stone Park, which is owned by
the City of Deer Lodge and leased to Powell County. Two removal actions there addressed
arsenic-contaminated soils during installation of utility poles and an outhouse. The park is

43


-------
located in the Phase 13 and Phase 14 remediation areas. Parts of the riverbank were eroded
where the Clark Fork River meanders. The park includes picnic areas and a walking trail system.

The site inspection continued in Deer Lodge where residential and streambank remediation of
arsenic-contaminated areas were observed in the Trestle Area. Participants then visited the large
area of pastureland east of the Phase 13 and 14 areas historically irrigated by a ditch that brought
water from the Clark Fork River to the area. The pastures visited included the Eastside
Pastures/Road area and the Windy Mountain Ranch (also known as the Broken Circle Ranch)
area where large areas of contaminated pasture land were remediated in 2011. The pastures were
vegetated with grass. The inspection proceeded to the Phase 7 remediation area, where Race
Track Pond was observed. Participants then visited the Phase 5 and 6 active remediation area.
Trucks and earthmoving equipment were observed removing contaminated floodplain soils and
filling in excavated areas with soil and compost from the Beck borrow area. The tour ended with
a visit to the Opportunity Pond repository where contaminated soils and sediment are placed.

6.6 Interviews

The FYR process included interviews with parties affected by the Site, including the current
landowners and regulatory agencies involved in site activities or aware of the Site. The purpose
was to document the perceived status of the Site and any perceived problems or successes with
the phases of the remedy implemented to date. EPA reached out to multiple stakeholders to
invite them to participate in the interview process. The interviews with those that were interested
took place in person, over the phone and in writing. All interviews are summarized below.
Appendix C provides the complete interviews.

MRSOU

Jeffrey Johnson: Jeffrey Johnson represents the National Park Service at the Grant-Kohrs Ranch
National Historic Site in Phase 15. Overall, he believes that the remedial activities at the
Milltown Reservoir were completed efficiently and that maintenance activities are sufficient. He
mentioned the current performance of the remedy is performing as expected. He is not aware of
any complaints from residents, new state laws that might affect the protectiveness of the remedy
or any changes in projected land uses at the Site. He is comfortable with the status of institutional
controls at the Site.

Chris Brick: Chris Brick is the director of a local community organization, Clark Fork Coalition
Sciences. Overall, she believes site cleanup has been successful. Vegetation at the former bypass
channel is not coming in very well, leading her to believe the cleanup might not be complete.
Another area of concern is an on-site repository adjacent to the bluff where a waste monitoring
well has had arsenic exceedances. As far as maintenance, the repository has reasonably good
grass but Ms. Brick mentioned that there should be more native shrubs. Ms. Brick is satisfied
with the reuse plans for a park on site. However, there are access issues and these need to be
resolved in order to move forward with redevelopment. Ms. Brick mentioned she has seen
mostly positive effects on the community. There were positive effects from the construction
work and the people like that they can continue to float and fish on the river. Lastly, Ms. Brick
commented that EPA did a great job keeping involved parties informed of site activities while

44


-------
cleanup was ongoing. However, now most of the information comes from Powell County and
applies to the CFROU. She is interested in using the former email list to update the community.

Michael Kustudia: Michael Kustudia is the manager of the Milltown State Park located on the
MRSOU. He has been involved with the MRSOU in various capacities over the last 15 years.
Overall he feels well informed and works closely with other involved agencies. He did not
provide any information regarding issues that might affect the protectiveness of the Site. He had
a few suggestions to keep community members informed on a continual basis, including creating
a fact sheet for area residents and users of the park updating people on the status of the arsenic in
groundwater at the Site and the results of the FYR. He also would like to see growth media
brought in for the top of the buttress near the tunnel pond repository. In addition, Mr. Kustudia
identified a small area of slickens at the site that he will show to EPA on their next field visit. He
indicated this is in a remote, hard to find area of the park.

CFROU

Resident 1: Resident 1 is a nearby resident of the CFROU and represents the local community.
He is aware of the former issues at the Site and believes that cleanup, maintenance and reuse
activities are coming along well. He mentioned the Site has had a positive economic effect on the
community by bringing in outside businesses. He commented that EPA has done a very good job
at keeping involved parties informed of site activities. EPA, MDEQ and the Clark Fork River
Technical Assistance Committee work well together in order to do this. He mentioned they
should keep informing local media of site activities. EPA is also putting him on an email list.
Resident 1 owns a private well south of town near Phases five and 6, which he tests regularly and
has never contained site-related contaminants. Resident 1 wants to be sure communication
between parties stays open.

Jeffrey Johnson: Jeffrey Johnson represents the National Park Service at the Grant-Kohrs Ranch
National Historic Site in Phase 15. Overall, he believes the remedial activities and maintenance
are being conducted efficiently and commented that the remedy is performing as expected. He is
aware that some nearby private landowners have submitted comments to MDEQ. He mentioned
that the National Park Service has provided support for MDEQ in site investigations, the
preliminary design plan and the remedial design. He is not aware of any changes in state laws or
any changes in projected land use at the Site. He is comfortable with the status of institutional
controls at the Site.

Brian Bartkowiak: Brian Bartkowiak represents MDEQ. Overall, he believes MDEQ is
completing the cleanup in an efficient, cost-effective and protective manner, while also ensuring
protection of human health and the environment. As far as the remedy, MDEQ has designed
plans consistent with the requirements of the ROD and Consent Decree and is currently
monitoring completed projects. He commented that some residents are concerned regarding the
scale of cleanup activities and the large-scale disturbances of the floodplain. As lead agency,
MDEQ oversees, manages, coordinates, designs and implements the remedial action for the Site
in collaboration with EPA. The agency also coordinates with Montana's NRDP and the National
Park Service for restoration components of the remedy. He commented that MDEQ also provides
public outreach for the Site, providing newsletter updates, weekly ads in the local newspaper,

45


-------
radio segments providing the public with information on current activities, outreach at various
local events, and providing documents to information repositories. He is not aware of any
changes to state laws or projected land uses at the Site. Institutional controls will be developed as
phases of the cleanup are completed.

Brian Bender: Brian Bender is the Powell County Planning Director. Overall he states that he is
well-informed about the activities at the CFROU by the MDEQ staff. He is not aware of any land
use changes or changes in local regulations that would affect the protectiveness of the remedy.
Mr. Bender indicated that the overlay district works well, even if occasionally it catches
something after work in the area is completed. At that point they involve MDEQ and the
situation is quickly resolved. He thinks information about the overlay district could be better
communicated with the community so they understand they need to get things investigated
before the start a project. Mr. Bender would like both EPA and MDEQ administrators to have
more of a presence in Powell County. He suggested they visit with County officials on a
quarterly, or more regular basis.

7.0	Technical Assessment

7.1	Question A: Is the remedy functioning as intended by the decision documents?

MRSOU

Yes. Review of the data collected during the FYR period and supporting documentation indicates
that the MRSOU remedial action continues to be operating and functioning as designed. The
primary objectives of the remedial action are to reduce or eliminate the groundwater arsenic
plume, and reduce a threat to aquatic life below the dam from the release of contaminated
sediments. The Milltown Dam has been completely removed, contaminated sediments have been
excavated or capped, and the Clark Fork River is flowing in the new channel with no
sedimentation or erosion issues identified. Floodplain vegetation is expected to achieve
performance standards. The SAA Ill-b sediments have been excavated and placed in the Tunnel
Pond Repository, which has been filled and the cover completed. The on-site repositories,
Interstate-90 bank improvements, removal and re-grading of the Bypass Channel, bridge
replacements and strengthening of the Interstate-90 Bridge abutments on the Blackfoot River are
completed and functioning as designed.

Vegetation performance standards have now been met for all areas where the PRPs retained
responsibility for revegetation. The PRPs expect to submit a Construction Completion Report in
2016. EPA and the State NRD program will continue to work cooperatively regarding other
vegetation areas and performance standards. Monitoring of the repositories and groundwater will
continue.

The ROD anticipated the dam removal would restore the aquifer by complying with ARARs for
groundwater approximately four to 10 years after dam removal and construction completion.
However, at the time this report was being drafted, it had only been four years since substantial
construction was completed. Groundwater monitoring indicates arsenic concentrations continue
to exceed the arsenic groundwater standard. However, the statistical analysis does not show a

46


-------
significant increasing trend in concentrations in the last eight samples. This issue requires further
investigation at a minimum until the 10-year period has passed.

The PRPs continue to recommend using statistical analysis of TPR10 as the Tunnel Pond
Repository POC with assessment of potential impacts to groundwater from repository
construction and use.

At the time of this FYR, permanent institutional controls have not been put in place for the
groundwater plume, for the waste repositories, for contaminated sediments left in place or for
site access control/residential use. Site regulatory agencies are continuing to discuss the need for
additional institutional controls. A Missoula County ordinance currently in place appears to
preclude installation of new public water wells in the vicinity of the MRSOU arsenic plume.
However, these ordinances do not preclude private well installation in the plume area. Additional
institutional controls may be needed to control private well installation in the arsenic plume and
with respect to the management of the waste repositories and the sediments left in place. Wells
monitored by MCCHD are consistently below the arsenic standard of 10 |ig/L.

An institutional control preventing river access during certain time periods has been necessary in
the past, and may be needed in the future. The majority of the MRSOU has been designated as a
future Montana State Park. Institutional controls dealing with water consumption, residential use
and the waste repositories will need to be incorporated into the future park design and planning
documents.

CFROU

Yes. Remedy implementation is ongoing. Remediation of Phase 1 of Reach A finished in April
2014. Revegetation activities are still ongoing. Long-term monitoring is underway to assess
groundwater, surface water and vegetation during and after remediation. Additional monitoring
efforts include streambed sediments, macroinvertebrates, periphyton, nutrients and fish
populations.

Institutional controls for CFROU to be implemented may include additional county zoning
regulations, deed restrictions, permanent funding for Arrow Stone Park, and groundwater
sampling and use controls. Environmental monitoring is required during all activities.
Institutional controls currently in place include the Powell Creek Overlay District. The Overlay
District, an existing institutional control, is intended to ensure that future land uses in affected
areas are compatible with the presence of potential contaminants and the remedial actions
required to isolate those potential contaminants from the environment.

7.2 Question B: Are the exposure assumptions, toxicity data, cleanup levels and
remedial action objectives (RAOs) used at the time of remedy selection still valid?

Yes. The exposure assumptions, toxicity data, cleanup levels and RAOs used at the time of
remedy selection remain valid for both the MRSOU and the CFROU.

47


-------
The MRSOU ROD indicates that groundwater standards are expected to be met within
approximately four to 10 years following completion of dam and sediment removal. A waiver of
groundwater standards is not currently proposed. However, the PRPs may seek a waiver of
groundwater cleanup standards if compliance is not achieved and is technically impracticable.

At the time of the ROD, the State of Montana's surface water quality standard for arsenic was 18
|ig/L, based on human health, and 20 |ig/L for groundwater as a drinking water supply. As
reflected in the August 2010 version of DEQ-7 (MDEQ2010), the state standard for arsenic for
surface water and groundwater is now 10 |ig/L, matching the federal standards. This revision to
the state standards does not impact the performance standards for the MRSOU, as the more
stringent federal standards were established in the 2004 ROD. Other groundwater and surface
water cleanup goals are based on federal and state standards that have not changed.

The MRSOU remedy is not expected to achieve compliance at all times with the State's WQB-7
standard for copper because of continued contaminant loading originating upstream of the
reservoir primarily from the CFROU. The ROD confirmed that a waiver of the copper standard,
based on technical impracticability, for the upstream CFROU will carry over into and be applied
to the MRSOU ambient surface water. The Consent Decree provides for the consideration of
upstream contamination in determining surface water ARAR compliance.

The risk-based soil cleanup goals for arsenic in the CFROU remain valid, as the toxicity
characteristics of arsenic have not changed since EPA issued the ROD. Land use in affected
areas has not changed in such a way as to affect the exposure assumptions applied in the
development of these site-specific cleanup goals.

7.3	Question C: Has any other information come to light that could call into question
the protectiveness of the remedy?

No. No other information has come to light that could call into question the protectiveness of the
remedy.

7.4	Technical Assessment Summary

MRSOU

Yes. Review of the data collected during the FYR period and supporting documentation indicates
that the MRSOU remedial action continues to be operating and functioning as designed. The
Milltown Dam has been completely removed, contaminant sediments have been excavated or
capped, and the Clark Fork River is flowing in the new channel with no sedimentation or erosion
issues identified. Vegetation performance standards have now been met at area for which the
PRPs are responsible, and are being monitored and improved in areas where the State NRD
program is responsible. Groundwater monitoring indicates arsenic concentrations continue to
exceed the arsenic groundwater standard. However, compliance may still be possible and
monitoring and further analysis should continue.

48


-------
Permanent institutional controls have not been put in place for the groundwater plume, for the
waste repositories, for contaminated sediments left in place or for site access control. Site
regulatory agencies are continuing to discuss the need for additional institutional controls.
Missoula County zoning ordinances are in place that preclude installation of new public water
wells in the vicinity of the arsenic plume. Additional institutional controls may be needed to
control private well installation in the arsenic plume and with respect to the management of the
waste repositories and the sediments left in place.

CFROU

Yes. Remedy implementation is ongoing. Remediation of Phase 1 of Reach A finished in April
2014. Long-term monitoring is underway to assess groundwater, surface water and vegetation
during remediation.

Institutional controls currently in place include the Powell County Overlay District. The Overlay
District is intended to ensure that future land use in affected areas are compatible with the
presence of potential contaminants and the various remedial actions required to isolate those
potential contaminants from the environment. Additional institutional controls for CFROU areas
may include county zoning regulations, deed restrictions, permanent funding for Arrow Stone
Park, and groundwater sampling and use controls.

8.0 Issues

Table 11 summarizes the current site issues.

Table 11: Current Site Issues

Issue

Affects Current
Protcctivcncss?

Affects Future
Protcctivcncss?

Institutional controls for MRSOU are not yet
implemented for areas where waste has been left in
place and areas where groundwater contamination is
above ROD standards.

No

Yes

Groundwater concentrations at MRSOU continue to
exceed arsenic cleanup goals and do not appear to be
declining.

No

Yes

49


-------
9.0 Recommendations and Follow-up Actions

Table 12 provides recommendations to address the current site issues.

Table 12: Recommendations to Address Current Site Issues

Issue

Recommendation /
Follow-Up Action

Party
Responsible

Oversight
Agency

Milestone
Date

Affects
Protectiveness?

Current

Future

Institutional
controls for
MRSOU are not
yet implemented
for areas where
waste has been
left in place and
areas where
groundwater
contamination is
above ROD
standards.

Implement
institutional controls
for the MRSOU
comprehensive
institutional control
plan and its
components.

PRP/
State/EPA

EPA/MDEQ

09/30/2017

No

Yes

Groundwater
concentrations at
MRSOU
continue to
exceed arsenic
cleanup goals
and do not
appear to be
declining.

Determine if
additional measures
are needed to reduce
arsenic concentrations
below the cleanup
goals.

PRP

EPA/MDEQ

09/30/2017

No

Yes

The following additional item, though not expected to affect protectiveness, warrants additional
follow up:

MRSOU

•	Two areas of the Tunnel Pond Repository showed subsidence of the cover. There was
also some minor erosion of the cover material in spots. Inspection of the one
downgradient monitoring well indicated that a locking well cap had not been put in place.
Envirocon indicated that these areas would be re-graded after the spring runoff was over.

10.0 Protectiveness Statements

The remedy at MRSOU (OU2) currently protects human health and the environment because
potential exposure to contaminated groundwater, surface water and sediment is controlled. For
the remedy to be protective over the long term, the following actions need to be taken:

•	Implement institutional controls for the MRSOU comprehensive institutional control plan
and its components.

•	Determine if additional measures are needed to reduce arsenic concentrations below the
cleanup goals.

50


-------
• Continue monitoring GW for at least six more years and tracking the arsenic trends to see
if concentrations are going down per the discussion in the ROD.

The remedy at CFROU (OU3) is expected to be protective of human health and the environment
upon completion of the remedial action. In the interim, exposure pathways that could result in
unacceptable risks are being controlled.

11.0 Next Review

The next FYR will be due within five years of the signature/approval date of this FYR.

51


-------
Appendix A: List of Documents Reviewed

2010	Milltown Transformation Retrospective, Diane Hammer, U.S. EPA. December 2010.

2011	Milltown Vegetation Monitoring Report. Geum Environmental Consulting, Inc. July 2012.

2011	Trestle Area Remedial Action Project Remedial Action Monitoring Plan. TerraGraphics
Environmental Engineering, Inc. October 2011.

2012	Milltown Vegetation Monitoring Report. University of Montana & Geum Environmental
Consulting, Inc. April 2013.

Clark Fork River Biomonitoring Macroinvertebrate Community Assessments, 2006. McGuire
Consulting. April 2007.

Clark Fork River Cleanup Phase 1 Continued River Closure Factsheet and Map. Montana
Department of Environmental Quality & Montana Department of Justice - Natural Resource
Damage Program. February 2014.

Clark Fork River Cleanup Upcoming Proposed River Closure Areas Factsheet and Map.
Montana Department of Environmental Quality & Montana Department of Justice - Natural
Resource Damage Program. February 2014.

Clark Fork River Closure Memo. Geum Environmental Consulting, Inc. January 30, 2014.

Clark Fork River Consent Decree Quarterly Report No. 25. U.S. EPA. February 2015.

Clark Fork River Consent Decree Quarterly Report No. 26. U.S. EPA. May 2015.

"Clark Fork River Flows into New Channel in Life after Milltown Dam." Missoulian. December
16, 2010.

Clark Ford River Operable Unit (OU#3) Explanation of Significant Differences. U.S. EPA. June
2015.

Clark Fork River Operable Unit Wildlife Monitoring. U.S. EPA. March 2012.

Clark Fork River Review. Montana Department of Environmental Quality & Montana
Department of Justice- Natural Resource Damage Program. October 2011.

Clark Fork River Review. Montana Department of Environmental Quality & Montana
Department of Justice- Natural Resource Damage Program. December 2012.

Construction Quality Assurance Plan Remedial Action Clark Fork River Operable Unit of the
Milltown Reservoir/Clark Fork River Superfund Site. Montana Department of Environmental
Quality. February 2009.

Cost Estimate for Clark Fork River Operable Unit Explanation of Significant Differences.
Bartkowiak, B. April 19, 2013.

A-l


-------
Draft Conceptual Redevelopment Plan for the Confluence of the Clark Fork and Blackfoot rivers
and adjacent communities. Milltown Superfund Site Redevelopment Working Group. February
2005.

Draft Final Construction Quality Assurance Project Plan (CQAPP) Reach A, Phase 1 Clark Fork
River Operable Unit Milltown Reservoir/Clark Fork River NPL Site Deer Lodge County,
Montana. Tetra Tech. July 2012.

Draft Interim Comprehensive Long Term Monitoring Plan for the Clark Fork River Operable
Unit - 2013 with SAP and QAPP. Atkins. March 2013.

EPA Superfund Record of Decision: Milltown Reservoir Sediments, EPA ID: MTD980717565,
OU 1, Milltown, MT. U.S. EPA. April 14, 1984.

EPA Superfund Record of Decision: Clark Fork River, EPA ID: MTD980717565, OU 3,
Milltown, MT. U.S. EPA. April 29, 2004.

EPA Superfund Record of Decision Amendment: Milltown Reservoir Sediments, EPA ID:
MTD980717565, OU 1, Milltown, MT. U.S. EPA. August 7, 1985.

Final Clark Fork River Reach A, Phase 1 Geomorphology and Vegetation Monitoring Plan.
Geum Environmental Consulting, Inc., Applied Geomorphology, Inc. October 2012.

Final Community Involvement Plan, Clark Fork River Operable Unit, Milltown Reservoir/ Clark
Fork River, Superfund Site. Montana Department of Environmental Quality. November 2012.

Final Construction Completion Report Deer Lodge and Eastside Road Residential Remedial
Action Project. TerraGraphics Environmental Engineering, Inc. May 2013.

Final Construction Quality Assurance Plan (CAP) Grant-Kohrs Ranch Bank Stabilization
Project. Tetra Tech. April 2013.

Final Construction Quality Assurance Project Plan (CQAPP) Reach A, Phase 1 Clark Fork River
Operable Unit Milltown Reservoir/Clark Fork River NPL Site Deer Lodge County, Montana.
Tetra Tech. October 2012.

Final Long-Term, Post Remedial Action Construction Monitoring Plan, Milltown Reservoir
Sediments Operable Unit. Envirocon. March 2013.

Final Remedial Action Work Plan Eastside Road Pastures Project Powell County, Montana.
TerraGraphics Environmental Engineering, Inc. October 2012.

First Five-Year Review Report: Milltown Reservoir Sediments, Missoula County, Montana.
Pacific Western Technologies Ltd. September 2011.

Instream Sediment Metal Concentrations in the Clark Fork River Operable Unit. EPA. 2012.

Integrating the "3 R's": Remediation, Restoration and Redevelopment, Milltown Reservoir
Sediments Site Case Study. U.S. EPA. April 2011.

A-2


-------
Invitation for Bid. Montana Department of Environmental Quality. January 2015.

Milltown Dam Removal Monitoring- Fisheries Investigations in 2012. Montana Fish, Wildlife
and Parks. March 2013.

Milltown Reservoir Sediments Operable Unit of the Milltown Reservoir/Clark Fork River
Superfund Site Record of Decision. U.S. Environmental Protection Agency, Region 8. December
2004.

Missoula Valley Water Quality Ordinance. Missoula County. 13.26.090 Protection of water
supply wells, http://www.ci.missoula.mt.us/DocumentCenter/Home/View/1033#

PublicServices 13 26 090

Monitoring Report for 2011, Clark Fork River Operable Unit. Atkins, Rhithron Associates, Inc.
& Montana Fish, Wildlife and Parks. August 2012.

Monitoring Report for 2012, Clark Fork River Operable Unit. Atkins, Rhithron Associates, Inc.
& Montana Fish, Wildlife and Parks. December 2013.

Montana Department of Environmental Quality Letter to Stakeholders. Montana Department of
Environmental Quality. 2014.

November 2012 Notification of Clean-Up Eastside Road Adjacent Form Letter. Montana
Department of Environmental Quality. November 2012.

Opening of Clark Fork River Press Release. Governor Steve Bullock, State of Montana. April
30, 2013.

Post-Construction Notification. 89 Sleepy Hollow Lane. Montana Department of Environmental
Quality. January 2012.

Post-Construction Notification. 90 Sleepy Hollow Lane. Montana Department of Environmental
Quality. January 2012.

Post-Construction Notification. 218 Milwaukee Avenue. Montana Department of Environmental
Quality. January 2012.

Post-Construction Notification. 220 Milwaukee Avenue. Montana Department of Environmental
Quality. January 2012.

Post-Construction Notification. 325 Milwaukee Avenue. Montana Department of Environmental
Quality. January 2012.

Post-Construction Notification. 401 Mitchell Street. Montana Department of Environmental
Quality. January 2012.

Post-Construction Notification. 621-619 Mitchell Street. Montana Department of Environmental
Quality. January 2012.

A-3


-------
Post-Construction Notification. 711 Railroad Street. Montana Department of Environmental
Quality. January 2012.

Post-Construction Notification. 1518 Eastside Road. Montana Department of Environmental
Quality. January 2012.

Post-Construction Notification. 1744 Eastside Road. Montana Department of Environmental
Quality. January 2012.

Post-Construction Notification. 1748 Eastside Road. Montana Department of Environmental
Quality. January 2012.

Public Health Well and Domestic Early Warning Well Monitoring Data, Groundwater
Compliance Monitoring Well Results, Milltown Reservoir Sediments Operable Unit. Missoula
City-County Health Department. May 2013.

Public Health Well and Domestic Early Warning Well Monitoring Data, Public Health
Groundwater Results, Milltown Reservoir Sediments Operable Unit. Missoula City-County
Health Department. 2014.

Quarterly Report of Activities for the Long-Term Clark Fork Monitoring Program (January
through March, 2007). U.S. Geological Survey. April 2007.

Quarterly Report of Activities for the Long-Term Clark Fork Monitoring Program (October
through December, 2006). U.S. Geological Survey. January 2007.

Residential Yard Data Summary for Select Historically Irrigated Areas, Clark Fork River
Operable Unit. Hydrometrics, Inc. January 1999.

Residential Yard Data Summary for Select Historically Irrigated Areas, Clark Fork River
Operable Unit. Hydrometrics, Inc. Revised May 1999.

Restoration of the Clark Foot River and Blackfoot River Near Milltown Dam, Revegetation As-
built Report, 2009-2012. Geum Environmental Consulting, Inc. June 2012.

Results of the December 2011 Analysis of 1750 Eastside Road Domestic Water Supply.
Montana Department of Environmental Quality. March 2012.

Return to Use Initiative, 2010 Demonstration Project. U.S. EPA. March 2014.

Soil Data Summary Report for the Eastside Road Pastures, Powell County, Montana, Clark Fork
River Operable Unit of the Milltown Reservoir/Clark Fork River Superfund Site. TerraGraphics
Environmental Engineering, Inc. December 2010.

Water-Quality, Bed-Sediment, and Biological Data (October 2007 through September 2008) and
Statistical Summaries of Long-Term Data for Streams in the Clark Fork Basin, Montana. Dodge,
K.A., Hornberger, M.I., and Dyke, J.L., U.S. Geological Survey. 2009.

A-4


-------
Water-Quality, Bed-Sediment, and Biological Data (October 2008 through September 2009) and
Statistical Summaries of Long-Term Data for Streams in the Clark Fork Basin, Montana. Dodge,
K.A., Hornberger, M.I., and Dyke, J.L., U.S. Geological Survey. 2010.

Water-Quality, Bed-Sediment, and Biological Data (October 2009 through September 2010) and
Statistical Summaries of Long-Term Data for Streams in the Clark Fork Basin, Montana. Dodge,
K.A., Hornberger, M.I., and Dyke, J.L., U.S. Geological Survey. 2011.

A-5


-------
Appendix B: Press Notice

EPA Five-Year Review Planned for the
Milltown Reservoir/ Clark Fork River
Superfund Site

The U.S. Environmental Protection Agency (EPA) is conducting the second Five-Year Review of
remedial actions performed under the Superfund program at the Milltown Reservoir/ Clark Fork River
Superfund site in Butte, Montana. The purpose of the Five-Year Review is to make sure the selected
cleanup actions remain protective of human health and the environment. The Five-Year Review is
scheduled for completion by September 2016.

The Site consists of three operable units. Operable unit 1 was focused on providing a safe water supply to
Milltown area residents through establishment of a public water supply system for the town of Milltown.
The Milltown Reservoir Sediments operable unit (MRSOU) is operable unit 2 and includes approximately
540 acres in the Clark Fork River and Blackfoot River floodplain. MRSOU consists of the area
encompassed by the former Milltown Dam and Reservoir and the area where arsenic contamination exists
in groundwater. The Clark Fork River Operable Unit consists of approximately 120 river miles of the
Clark Fork River and extends from the confluence of the old Silver Bow Creek channel with the
reconstructed lower Mill-Willow bypass, near Anaconda, to the maximum former Milltown Reservoir
pool elevation east of Missoula. The Milltown Reservoir/ Clark Fork River site is one of four
contamination areas, jointly known as the Clark Fork Basin Sites.

More information is available at the site's information repository and on EPA's website:

EPA Superfund Records Center

Montana Office
10 West 15th Street, Suite 3200
Helena, MT 59626
(406) 457-5046
(866) 457-2690 (toll free)

http://www2.epa.gov/region8/milltown-reservoir-sediments-clark-fork-river

EPA invites community participation in the Five-Year Review process: Community members are
encouraged to contact EPA staff with any information that may help the Agency make its determination
regarding the protectiveness and effectiveness of the remedies at the site.

EPA Region 8

Sara Sparks

Remedial Project Manager
Phone: (406) 782-7415
Email: sparks.sara@epa.gov



B-l


-------
Appendix C: Interview Forms

Milltown Reservoir/ Clark Fork River

Five-Year Review Interview

Form

Superfund Site

Site Name: Milltown Reservoir Sediments EPA ID No.: MTD980717565

OU

Interviewer Name: Sj
Subject Name: J(
Subject Contact
Information:

Time: Not Applicable

Jeffrey Johnson

Self

ieffrev g iohnson@nps.gov

Affiliation: Skeo

Affiliation: National Park Service

Date:

01/28/2016

Interview
Location:

Grant-Kohrs Ranch NHS 266 Warren Lane Deer lodge, MT 59722

Interview Format (circle one): In Person Phone Mail

Othe

Interview Category: Federal Agency

1.	What is your overall impression of the project, including cleanup, maintenance and reuse
activities (as appropriate)?

The remedial activities at the Milltown Reservoir were done efficiently. The maintenance is

good.

2.	What is your assessment of the current performance of the remedy in place at the Site?

The remedy in place is performing within expectations.

3.	Are you aware of any complaints or inquiries regarding site-related environmental issues or
remedial activities from residents in the past five years?

No.

4.	Has your office conducted any site-related activities or communications in the past five
years? If so, please describe the purpose and results of these activities.

Not for the Milltown Reservoir OU.

5.	Are you aware of any changes to state laws that might affect the protectiveness of the Site's
remedy?

No.

6.	Are you comfortable with the status of the institutional controls at the Site? If not, what are
the associated outstanding issues?

C-l


-------
Yes.

7.	Are you aware of any changes in projected land use(s) at the Site?

No.

8.	Do you have any comments, suggestions or recommendations regarding the management or
operation of the Site's remedy?

No.

C-2


-------
Milltown Reservoir/ Clark Fork River	Five-Year Review Interview

Superfund Site	Form

Site Name: Milltown Reservoir Sediments EPA ID No.: MTD980717565
OU

Interviewer Name: Treat Suomi	Affiliation: Skeo

Subject Name:	Chris Brick	Affiliation: Clark Fork Coalition

Sciences

Subject Contact	Director: (406) 542-0539

Information:

Time: 2:00 p.m.	Date: 11/02/2015

Interview	140 South 4th Street West Suite 1 Missoula, MT

Location:

Interview Format (circle one): Qn Persoi^ Phone Mail Other:

Interview Category: Local Community Organization

1.	Are you aware of the former environmental issues at the Site and the cleanup activities that
have taken place to date?

Yes.

2.	What is your overall impression of the project, including cleanup, maintenance and reuse
activities (as appropriate)?

Overall, I think it has been successful. I would rate the cleanup an eight out of 10. The
vegetation at the former bypass channel is not coming in very well. The NRDP has done
testing and my understanding is that the area still has some high metals so the substandard
vegetation leads to a belief that the cleanup might not be complete. So, I think it is 80 to 90
percent effective.

The other area is an on-site repository adjacent to the bluff where 3B waste monitoring well
downstream has had arsenic exceedances. At the AR repository, there are questions about
what to do. I think that this is a red flag and that is one area of concern.

Maintenance: that same repository is getting reasonably good grass but I argued for a long
time that there should be more native shrubs. The Interstate-90 bridge piers are the same
concerns that have been previously voiced.

Reuse: there are great plans for a park. There are access problems for the FWP and
International Paper, though. Last I heard, they might be working on that. The state has
money and plans to do park construction and it has been blocked by the access issue. This
needs to be resolved. This has prevented complete redevelopment.

Great job on the bluff, mainly the side and the former reservoir and the area below.

3. What have been the effects of the Site on the surrounding community, if any?

C-3


-------
There have been a lot of A 's in the community. There were positive effects from the
construction work. It is also beneficial that people have been able to continue to float and
fish. And I understandfish move up the Black Foot and Upper Clark Rivers to spawn. But
any beneficial effects to the community have been stalled due to access issues and slowed
redevelopment. It is beneficial that people have been able to float andfish. I understandfish
move up Black Foot and the Upper Clark Fork in order to spawn.

4.	Have there been any problems with unusual or unexpected activities at the Site, such as
emergency response, vandalism or trespassing?

Not that I am aware of.

5.	Has EPA kept involved parties and surrounding neighbors informed of activities at the Site?
How can EPA best provide site-related information in the future?

EPA did a great job while project was ongoing, but now there is not much to report. Most of
the information comes from Powell County now. I am interested in the vegetation of the
bypass channel and water quality. I am also interested in the using the former email list and
allowing people to opt in for future updates.

6.	Do you own a private well in addition to or instead of accessing city/municipal water
supplies? If so, for what purpose(s) is your private well used?

There are not any near the Site.

7.	Are you aware of any changes in projected land use(s) at the Site?

No, I am not aware of any.

8.	Do you have any comments, suggestions or recommendations regarding any aspects of the
project?

No, other than making sure the issues at the former bypass channel with the revegetation are
solved and the water quality issues resulting from issues with the repository. There may be
other issues I am currently unaware of.

C-4


-------
Milltown Reservoir/ Clark Fork River	Five-Year Review Interview

Superfund Site	Form

Site Name: Milltown Reservoir Sediments

EPA ID No.:

MTD980717565

OU







Interviewer Name:

Treat Suomi

Affiliation:

Skeo

Subject Name:

Resident 1

Affiliation:

Nearby Resident

Time: 10:00 a.m.



Date: 11/04/2015

Interview

7956 East Side Road





Location:







Interview Format (circle one): {In Person)

Phone Mail Other:

Interview Category: Residents

1.	Are you aware of the former environmental issues at the Site and the cleanup activities that
have taken place to date?

Yes.

2.	What is your overall impression of the project, including cleanup, maintenance and reuse
activities (as appropriate)?

I think it is coming along well.

3.	What have been the effects of the Site on the surrounding community, if any?

Economically, it has helped. It has brought some outside businesses here.

4.	Have there been any problems with unusual or unexpected activities at the Site, such as
emergency response, vandalism or trespassing?

Not to my knowledge.

5.	Has EPA kept involved parties and surrounding neighbors informed of activities at the Site?
How can EPA best provide site-related information in the future?

I think they have done a very goodjob at this. I serve on CFRTAC and together, EPA, DEQ
and CFRTAC have done a goodjob; the three organizations work well together. They should
continue to keep it coming to the local media. They have done a good job. EPA is putting me
on an email list, too.

6.	Do you own a private well in addition to or instead of accessing city/municipal water
supplies? If so, for what purpose(s) is your private well used?

C-5


-------
I own a private well south of town, near Phases five and six. I test regularly and have never
seen site-related contaminants.

7. Do you have any comments, suggestions or recommendations regarding any aspects of the
project?

Keep the lines of communication open.

C-6


-------
Milltown River/ Clark Fork River
Superfund Site	

Five-Year Review Interview

Form

Site Name: Clark Fork River OU

EPA ID No.: MTD980717565

Affiliation: Skeo

Affiliation: National Park Service

Interviewer Name: Self
Subject Name: Jeffi
Subject Contact
Information:

Time: Not Applicable

Jeffrey Johnson

ieffrev g iohnson@nps.gov

Date:

01/28/2016

Interview
Location:

Grant-Kohrs Ranch NHS 266 Warren Lane Deer lodge, MT 59722

Interview Format (circle one): In Person Phone Mail

Oth

Interview Category: Federal Agency

1.	What is your overall impression of the project, including cleanup, maintenance and reuse
activities (as appropriate)?

The remedial activities at the Clark Fork River are being conducted efficiently. The maintenance

is good.

2.	What is your assessment of the current performance of the remedy in place at the Site?

The remedy in place is performing within expectations.

3.	Are you aware of any complaints or inquiries regarding site-related environmental issues or
remedial activities from residents in the past five years?

I am aware that some private landowners have commented to MDEQ.

4.	Has your office conducted any site-related activities or communications in the past five
years? If so, please describe the purpose and results of these activities.

MDEQ has completed investigations and prepared the Preliminary Design Plan. They are

currently completing the remedial design. Grant-Kohrs Ranch has supported these activities.

5.	Are you aware of any changes to state laws that might affect the protectiveness of the Site's
remedy?

6. Are you comfortable with the status of the institutional controls at the Site? If not, what are
the associated outstanding issues?

C-7


-------
7.	Are you aware of any changes in projected land use(s) at the Site?

No.

8.	Do you have any comments, suggestions or recommendations regarding the management or
operation of the Site's remedy?

No.

C-8


-------
Milltown River/ Clark Fork River
Superfund Site

Five-Year Review Interview

Form

Site Name: Clark Fork River OU

Interviewer Name:
Subject Name:
Subject Contact
Information:

Treat Suomi
Brian Bartkowiak

EPA ID No.:

Affiliation:

Affiliation:

Time: Not Applicable

1225 Cedar Street
P.O. Box 200901
Helena. MT 59620-0901
(406) 444-0214

Date:

MTD980717565

Skeo

MDEO

12/16/2015

Interview Format (circle one): In Person Phone Mail	Othe^ Email

Interview Category: State Agency

1.	What is your overall impression of the project, including cleanup, maintenance and reuse
activities (as appropriate)?

MDEQ is implementing the project in an efficient, cost-effective and protective manner while
ensuring the protection of human health and the environment and emphasizing worker and
public safety.

2.	What is your assessment of the current performance of the remedy in place at the Site?

MDEQ design teams have developed designs consistent with the requirements of the ROD and
Consent Decree. MDEQ is currently monitoring performance of completed project to ensure
performance metrics, performance targets and performance standards are being met.

3.	Are you aware of any complaints or inquiries regarding site-related environmental issues or
remedial activities from residents in the past five years?

Some residences have concerns regarding the scale of the cleanup activities. Residents have
expressed concerns over the large-scale disturbances of the floodplain.

4.	Has your office conducted any site-related activities or communications in the past five
years? If so, please describe the purpose and results of these activities.

MDEQ, as lead agency, oversees, manages, coordinates, designs and implements the remedial
action for the Site in consultation with EPA. MDEQ coordinates with the State of Montana's
NRDP and the U.S. National Park Service for the implementation and integration of restoration
components into the work. Four primary functions of consultation and coordination among the
agencies for the Site are to: 1) understand and receive the information to be collected; 2)
understand how that information is to be analyzed; 3) provide review and comment; and 4)

C-9


-------
maximize the use of the resources available for and the environmental benefits to the Site in the
successful and cost-effective completion of the work.

MDEQ also conducts significant public outreach, including, but not limited to: monthly
stakeholder and landowner tours during construction, periodic newsletter updates, weekly ads in
the local newspaper and radio providing the public with information on current activities, design
review meetings, outreach at local events, and providing key documents at site information
repositories.

5.	Are you aware of any changes to state laws that might affect the protectiveness of the Site's
remedy?

No.

6.	Are you comfortable with the status of the institutional controls at the Site? If not, what are
the associated outstanding issues?

Yes. The cleanup is underway and individual institutional control plans will be developed as
project phases are completed.

7.	Are you aware of any changes in projected land use(s) at the Site?

No.

8.	Do you have any comments, suggestions or recommendations regarding the management or
operation of the Site's remedy?

No.

C-10


-------
Milltown River/ Clark Fork River	Five-Year Review Interview

Superfund Site	Form

Site Name: Clark Fork River OU	EPA ID No.: MTD980717565

Interviewer Name: Treat Suomi	Affiliation: Skeo

Subject Name: Brian Bender	Affiliation: Powell County Planning

Director

Subject Contact	bbender@powellcountvmt.com I 406-846-9795

Information:

Time: 2:30 p.m.	Date:	06/14/2016

Interview	Phone

Location:

Interview Format (circle one): In Person ^hon^ Mail	Other:	

Interview Category: Local Government

1. Are you aware of the former environmental issues at the Site and the cleanup activities that
have taken place to date?

Yes.

2.	Do you feel well-informed regarding the Site's activities and remedial progress? If not, how
might EPA convey site-related information in the future?

Yes, Ifeel well informed. We get quarterly reports anything site-specific from MDEQ staff
However, EPA has not communicated with us in over three years. County staff would
appreciate regular communications form EPA on the status of the project.

3.	Have there been any problems with unusual or unexpected activities at the Site, such as
emergency response, vandalism or trespassing?

Nothing that was critical. MDEQ staff have indicated that there is an occasional incident of
trespassing but they have not indicated that it has been a serious situation.

4.	Are you aware of any changes to state laws or local regulations that might affect the
protectiveness of the Site's remedy?

No.

5. Are you aware of any changes in projected land use(s) at the Site?

No. I am not aware of any and I do not believe any are being proposed. Through the Powell
County Planning Department, we have a Superfund Overlay District. Someone would have to
initiate any changes they wanted. Occasionally, the Overlay District catches something after
the fact, so maybe information about the Overlay District could be better communicated with
the community so they know they need to have things investigated earlier. However, MDEQ
has been good to communicate with and they come in and haul away waste if needed.

C-ll


-------
6.	Has EPA kept involved parties and surrounding neighbors informed of activities at the Site?
How can EPA best provide site-related information in the future?

MDEQ puts weekly notices in the paper and on the radio. EPA does not really have much of
a local presence. EPA is supposed to help residents and now that work in Powell County has
started, we have not really heardfrom EPA. We used to have a funding mechanism in place
to help fund the Powell County Planning Department and that was abruptly taken away.

7.	Do you have any comments, suggestions or recommendations regarding the project?

EPA needs more presence with property owners and county officials, both formally and
informally. It would also be good if MDEQ could meet quarterly or every six weeks with
county officials. Now that work is in Deer Lodge, the same regular, in-person updates could
be given to City Council. Specifically, it would be good to have MDEQ administrators or
senior officials visit on a regular, maybe quarterly, basis.

C-12


-------
Milltown River/ Clark Fork River
Superfund Site	

Five-Year Review Interview

Form

Site Name: Milltown Reservoir OU
Interviewer Name: Treat Suomi
Subject Name: Michael Kustudia

EPA ID No.: MTD980717565

Affiliation: Skeo

Affiliation: Milltown State Park,

Montana Fish, Wildlife &
Parks, Region Two

Subject Contact
Information:
Time: 11:00 a.m.

MKustudia@mt.gov

Date: 07/26/2016

Interview Format (circle one): In Person

Mail Other:

Interview Category: Local Government

1.	Are you aware of the former environmental issues at the Site and the cleanup activities that
have taken place to date?

Yes. I am the manager of Milltown State Park. And before that I was involved it the TAG. So
I have been involvedfor the last 15 years.

2.	Do you feel well-informed regarding the Site's activities and remedial progress? If not, how
might EPA convey site-related information in the future?

Ifeel like I am well informed. Through EPA, NRD and other sister agencies Ifeel well
informed.

3.	Have there been any problems with unusual or unexpected activities at the Site, such as
emergency response, vandalism or trespassing?

The site has become part of a state park now. WE are working on transferring it from the
NRD program. Tunnel that gets vandalized but nothing that would affect the protectiveness
of the remedy.

4.	Are you aware of any changes to state laws or local regulations that might affect the
protectiveness of the Site's remedy?

I am not.

5.	Are you aware of any changes in projected land use(s) at the Site?

Aside from developing the state park as planned.

6.	Has EPA kept involved parties and surrounding neighbors informed of activities at the Site?
How can EPA best provide site-related information in the future?

C-13


-------
EPA (Sara has kept me informed reasonably well. This interview is an example of that. As far
as the community goes, the remedy is largely complete and there really isn't a needfor the
public meetings we used to have.

7. Do you have any comments, suggestions or recommendations regarding the project?

I have a couple of suggestions. In the spring we went out for an annual visit. This has been
one of my continuing messages. The buttress to the buttress for the tunnel pond repository -
it never got any growth media put down on the top. Getting some growth media on the top
would be my wish. We are pretty good at mobilizing volunteers for plantings and such so if
we could get some topsoil there we could get it planted but it is beyond our budget to bring in
the growth media. I am relatively pleased with the vegetation in the area. The surrounding
areas look great but there are too many weeds for my liking. We are likely to have a Mullen
weed pulling event soon.

There is a small area of slickens upstream from the confluence, approximately 5 — 10 feet
across and 3-4 feet wide (15-30 square feet). It is an isolated spot. It is hard to find. Right
below the confluence there is a red spot/stain with a trickle of water. I did a ph test on it and
it was practically neutral. It does seem seasonal. I noticed the trickle before peak runoff and
by the time the runoff came it was gone and I haven't seen it since.

Now that the work is largely done. In terms of the monitoring that goes on in the wells, where
do we stand? Are we on the right trajectory in terms of the arsenic in groundwater? I think
the public would like to hear an update about that as well. Even a one-page fact sheet after
the Five Year Review would be good. There might not be enough for an actual meeting but
some sort of outreach might be helpful as progress is made.

C-14


-------
Appendix D-l: MRSOU Site Inspection Checklist

FIVE-YEAR REVIEW SITE INSPECTION CHECKLIST

I. SITE INFORMATION

Site name: Milltown Sediments OU

Date of inspection: 11/02/2015

Location and Region: Milltown, Missoula County,
Montana, EPA Region 8

EPA ID: MTD980717565

Agency, office, or company leading the five-year
review: EPA

Weather/temperature: mostly cloudy, low 40s

Remedy Includes: (Check all that apply)

1X1 Landfill cover/containment
1X1 Access controls
1X1 Institutional controls

~	Groundwater pump and treatment

~	Surface water collection and treatment
	Q Other 	

~	Monitored natural attenuation

~	Groundwater containment

~	Vertical barrier walls

Attachments:

Inspection team roster attached

Site map attached (See Figure 2)

II. INTERVIEWS (Check all that apply)

1. O&M staff

Name

Name

Title
Title

Date
Date

Interviewed ~ at site ~ at office ~ by phone Phone no.
Problems, suggestions; ~ Report attached 	

2.

Local regulatory authorities and response agencies (i.e., State and Tribal offices, emergency response
office, police department, office of public health or environmental health, zoning office, recorder of
deeds, or other city and county offices, etc.). Fill in all that apply.

Agency:

Contact

Name Title
Problems; suggestions; ~ Report attached	

Date

Phone No.

4.

Other interviews (optional) ~ Report attached

Jeffrey Johnson, National Park Service; Chris Brick, Clark Fork Coalition Sciences

III. ON-SITE DOCUMENTS & RECORDS VERIFIED (Check all that apply)

O&M Documents

1X1 O&M manual
~ As-built drawings
1X1 Maintenance logs
Remarks:

1X1 Readily available
~ Readily available
1X1 Readily available

1X1 Up to date
~ Up to date
1X1 Up to date

~	n/a

M N/A

~	n/a

D-l


-------
2.

Site-Specific Health and Safety Plan

1 1 Readily available

1 1 Up to date

IXI N/A



~ Contingency plan/emergency response
plan

~ Readily available

~ Up to date

[XI N/A



Remarks:







3.

O&M and OSHA Training Records

Remarks:

153 Readily available

1 1 Up to date

[XI N/A

4.

Permits and Service Agreements









~ Air discharge permit

~ Readily available

~ Up to date

[XI N/A



~ Effluent discharge

~ Readily available

~ Up to date

[XI N/A



~ Waste disposal, POTW

1 1 Readily available

1 1 Up to date

[XI N/A



I-! Other permits

~ Readily available

~ Up to date

[XI N/A



Remarks:







5.

Gas Generation Records

Remarks:

~ Readily available

~ Up to date

[XI N/A

6.

Settlement Monument Records

Remarks:

153 Readily available

153 Up to date

~ n/a

7.

Groundwater Monitoring Records

Remarks:

153 Readily available

[Xl Up to date

~ n/a

8.

Leachate Extraction Records

~ Readily available

~ Up to date

[XI N/A



Remarks:







9.

Discharge Compliance Records









1 1 Air ~ Readily available ~ Up to date



N/A



~ Water (effluent) ~ Readily available ~ Up to date



N/A



Remarks:







10.

Daily Access/Security Logs

Remarks:

1 1 Readily available

1 1 Up to date

IXI N/A

IV. O&M COSTS

1.

O&M Organization

153 State in-house
1 1 PRP in-house

1 1 Contractor for State
153 Contractor for PRP







I~1 Federal Facility in-house

1 1 Contractor for Federal Facility





~







D-2


-------
2. O&M Cost Records

~	Readily available	Q Up to date

~	Funding mechanism/agreement in place	Unavailable

3. Unanticipated or Unusually High O&M Costs During Review Period

Describe costs and reasons: None

V. ACCESS AND INSTITUTIONAL CONTROLS M Applicable ~ N/A

A.	Fencing

1. Fencing damaged	~ Location shown on site map ~ Gates secured ^ N/A

Remarks: Fencing is present in areas where public can get to river only. There are also locked gates on
roads that lead to the remedial action construction areas.

B.	Other Access Restrictions

1. Signs and other security measures	~ Location shown on site map ~ N/A

Remarks: Signage is currently present near the river across from the Site. Primarily concerned with
protecting revegetation areas.

C.	Institutional Controls (ICs)

1. Implementation and enforcement

Site conditions imply ICs not properly implemented
Site conditions imply ICs not being fully enforced

Type of monitoring (e.g., self-reporting, drive by)	

Frequency	

Responsible party/agency

Contact		

Name	Title

Reporting is up-to-date
Reports are verified by the lead agency

Specific requirements in deed or decision documents have been met
Violations have been reported
Other problems or suggestions: ~ Report attached
IC Plan is still in development.

~	Yes ~ No |EI N/A

~	Yes ~ No ^ N/A

mm/dd/vvw





Date



Phone no.

~ Yes

~

No

[XI N/A

I~1 Yes

~

No

|E1 N/A

I~1 Yes



No

~ n/a

I~1 Yes

~

No

|E1 N/A

D-3


-------
2. Adequacy	~ ICs are adequate	^ ICs are inadequate	~ N/A

Remarks: A Missoula County ordinance currently in place appears to preclude installation of new public
water wells in the vicinity of the MRSOU arsenic plume. However, these ordinances do not preclude
private well installation in the plume area. Additional institutional controls may be needed to control
private well installation in the arsenic plume, prevent residential use and protect the waste repositories and
the sediments left in place. An institutional control preventing river access during certain time periods has
been necessary in the past, and may be needed in the future. The majority of the MRSOU has been
designated as a future Montana State Park. Institutional controls dealing with water consumption,
residential use and the waste repositories will need to be incorporated into the future park design and
planning documents.

D. General

1.	Vandalism/trespassing ~ Location shown on site map	No vandalism evident
Remarks: River rafters are not obeying signage and floating down the river in prohibited areas.

2.	Land use changes on site	^ N/A

Remarks: Land formerly owned bv Northwestern Corp. (dam operator) was acquired bv the State of
Montana for future use as a state park.

3.	Land use changes off site	^ N/A

Remarks:	

VI. GENERAL SITE CONDITIONS
A. Roads ~ Applicable ^ N/A

VII. LANDFILL COVERS	^Applicable ~ N/A

A. Landfill Surface

1.	Settlement (Low spots) ~ Location shown on site map ~ Settlement not evident

Arial extent		Depth	

Remarks: Settlement is evident in the Tunnel Pond Repository. Subsidence is evident at the Right
Bank Repository. PRPs are monitoring and have worked on revegetation efforts at the Tunnel Pond
Repository area of settlement. Area appears to have grown in size since the last FYR.

2.	Cracks	~ Location shown on site map EH Cracking not evident

Lengths		Widths		Depths	

Remarks: Cracks are associated with slumping of the ground during settlement. The 2015 Draft
Annual Report noted cracks from settlement in the appendix.

3.	Erosion	I~1 Location shown on site map	E3 Erosion not evident

Arial extent		Depth	

Remarks:

4.	Holes	~ Location shown on site map ~ Holes not evident

Arial extent		Depth	

Remarks: Right Bank Repository had evidence of holes. They were flagged for continued PRP
monitoring.

D-4


-------
5.

Vegetative Cover

1X1 Grass

1 1 Cover properly established



153 No signs of stress

1 1 Trees/Shrubs (indicate size and locations on a diagram)



Remarks: Several minor ruts were observed in the cover where erass has not vet come in.

6.

Alternative Cover (armored rock, concrete, etc.)

IEIn/a



Remarks:





7.

Bulges

Arial extent
Remarks:

~ Location shown on site map

E3 Bulges not evident
Heieht

8. Wet Areas/Water
Damage

^ Wet areas/water damage not evident



n Wet areas

1 1 Location shown on site map

Arial extent



1 1 Ponding

1 1 Location shown on site map

Arial extent



I~1 Seeps

1 1 Location shown on site map

Arial extent



1 1 Soft subgrade

1 1 Location shown on site map

Arial extent



Remarks:





9.

Slope Instability

1 1 Slides

1 1 Location shown on site map



No evidence of slope instability





Arial extent







Remarks: None noted.





B.

Benches ~ Applicable N/A





(Horizontally constructed mounds of earth placed across a steep landfill side slope to interrupt the slope in
order to slow down the velocity of surface runoff and intercept and convey the runoff to a lined channel.)

C.

Letdown Channels I

3 Applicable ~ N/A





(Channel lined with erosion control mats, riprap, grout bags, or gabions that descend down the steep side
slope of the cover and will allow the runoff water collected by the benches to move off of the landfill
cover without creating erosion gullies.)

1.

Settlement (Low spots)
Arial extent
Remarks:

~ Location shown on site map

No evidence of settlement
Depth

2.

Material Degradation

Material tvpe

1 1 Location shown on site map

No evidence of degradation
Arial extent



Remarks:





3.

Erosion

Arial extent
Remarks:

~ Location shown on site map

No evidence of erosion
Depth

D-5


-------
4

Undercutting Q Location shown on site map

153 No evidence of undercutting



Arial extent

Depth



Remarks:



5

Obstructions Tvpe

153 No obstructions



I | Location shown on site map Arial extent





Size





Remarks:



6

Excessive Vegetative Growth Tvpe





15^1 No evidence of excessive growth





~ Vegetation in channels does not obstruct flow





I | Location shown on site map Arial extent





Remarks:



D.

Cover Penetrations ~ Applicable ^ N/A



E.

Gas Collection and Treatment ~ Applicable ^ N/A



F.

Cover Drainage Layer Q Applicable ^ N/A



G.

Detention/Sedimentation Ponds ~ Applicable

M N/A

H. Retaining Walls ^ Applicable ~ N/A

1

Deformations Q Location shown on site map

1 1 Deformation not evident



Horizontal displacement Vertical displacement



Rotational displacement





Remarks: Usins a historic railroad srade as a retaining wall orberm for SAA Illb contaminated



sediments. PRPs have bolstered the toe of the srade to prevent movement. In addition, the PRPs also



installed settlement monuments in the crest and toe of the Tunnel Pond Repository embankment in April



2014 as reauired bv the Monitorine Plan. .



2

Degradation Q Location shown on site map

153 Degradation not evident



Remarks:



I. Perimeter Ditches/Off-Site Discharge ^ Applicable

~ n/a

1.

Siltation Q Location shown on site map

153 Siltation not evident



Area extent

Depth



Remarks:



2.

Vegetative Growth Q Location shown on site map

~ n/a



153 Vegetation does not impede flow





Area extent

Type



Remarks: There is some vesetative erowth in ditch. It does not appear to impede flow.

D-6


-------
3.	Erosion	Q Location shown on site map ^ Erosion not evident

Area extent		Depth	

Remarks:	

4.	Discharge Structure	Functioning	^ N/A
Remarks:	

VIII.	VERTICAL BARRIER WALLS	~ Applicable N/A

IX.	GROUNDWATER/SURF ACE WATER REMEDIES ^Applicable ~ N/A

A.	Groundwater Extraction Wells, Pumps, and Pipelines	~ Applicable ^ N/A

B.	Surface Water Collection Structures, Pumps, and Pipelines ~ Applicable ^ N/A

C.	Treatment System	~ Applicable ^ N/A

D.	Monitoring Data

1.	Monitoring Data

Is routinely submitted on time	£3 Is of acceptable quality

2.	Monitoring data suggests:

153 Groundwater plume is effectively contained ~ Contaminant concentrations are declining

E.	Monitored Natural Attenuation	

1. Monitoring Wells (natural attenuation remedy)

Properly secured/locked	^ Functioning ^ Routinely sampled ^ Good condition

~ All required wells located ~ Needs Maintenance	~ N/A

Remarks: Did not visit all of the compliance wells on site. Those observed were in good condition.

	X. OTHER REMEDIES	

If there are remedies applied at the site and not covered above, attach an inspection sheet describing the physical
nature and condition of any facility associated with the remedy. An example would be soil vapor extraction. Not
applicable.

	XI. OVERALL OBSERVATIONS	

A.	Implementation of the Remedy	

Review of the data collected during the FYR period and supporting documentation indicates that the
MRSOU remedial action continues to be operating and functioning as designed. The Milltown Dam has
been completely removed, contaminant sediments have been excavated or capped, and the Clark Fork
River is flowing in the new channel with no sedimentation or erosion issues identified. Vegetation
performance standards have now been met at areas for which the PRPs are responsible and are
progressing at other areas. Groundwater monitoring indicates arsenic concentrations continue to exceed
	the arsenic groundwater standard.	

B.	Adequacy of O&M	

The Long-Term Post Remedial Action Construction Monitoring Plan for MRSOU was finalized in 2013.
The plan outlines the groundwater and surface water monitoring requirements as well as the long-term
maintenance and monitoring requirements for the constructed repositories and buttress areas. Prior to the

	lOlS^Jan^^	

C.	Early Indicators of Potential Remedy Problems	

D-7


-------
Describe issues and observations such as unexpected changes in the cost or scope of O&M or a high
frequency of unscheduled repairs that suggest that the protectiveness of the remedy may be compromised
in the future. None identified.

D. Opportunities for Optimization	

Describe possible opportunities for optimization in monitoring tasks or the operation of the remedy.

None identified.

Site Inspection Team:

•	Sara Sparks, EPA

•	Brian Bartkowiak, MDEQ

•	Claire Marcussen, Skeo

•	Treat Suomi, Skeo

Appendix D-2: CFROU Site Inspection Checklist

FIVE-YEAR REVIEW SITE INSPECTION CHECKLIST



I. SITE INFORMATION

Site name: Clark Fork River OU

Date of inspection: 11/03/2015

Location and Region: Milltown, Missoula County,

EPA ID: MTD980717565

Montana, EPA Region 8

Agency, office, or company leading the five-year

Weather/temperature: 30°s Fahrenheit, cloudy,

review: EPA

occasional snow

Remedy Includes: (Check all that apply)



1X1 Landfill cover/containment

~ Monitored natural attenuation

~ Access controls

~ Groundwater containment

1X1 Institutional controls

~ Vertical barrier walls

1 1 Groundwater pump and treatment



~ Surface water collection and treatment



IXI Other In-situ treatment of soils and sediments.

Attachments: ^ Inspection team roster attached

153 Site map attached (See Figure 1)

II. INTERVIEWS (Check all that apply)

1. O&M site manager/

mm/dd/vvvv

Local Regulatory Name

Title Date

Authorities and



Response Agencies



Interviewed Q at site Q at office Q bv phone Phone no.

Problems, sueeestions: Q Report attached



2. O&M staff

mm/dd/vvvv

Name

Title Date

Interviewed at site at office bv phone Phone no.

Problems, sueeestions: Report attached



3. Other interviews (optional) I~1 Report attached

D-8


-------
Jeffrey Johnson, National Park Service; nearby resident

III. ON-SITE DOCUMENTS & RECORDS VERIFIED (Check all that apply)

1.

O&M Documents









~ O&M manual ^ Readily available

[XI Up to date

M N/A





~ As-built drawings ^ Readily available

1X1 Up to date

M N/A





~ Maintenance logs £3 Readily available

1X1 Up to date

IE|n/a





Remarks: A breakdown of costs for the CFROU from 2008 to 2014 were provided and reviewed.



Since remedial actions are still beins desiened and implemented at the CFROU. separate O&M costs



are not presented.







2.

Site-Specific Health and Safety Plan

1X1 Readily available

1X1 Up to date K

N/A



1X1 Contingency plan/emergency response
plan

1X1 Readily available

1X1 Up to date K

N/A



Remarks: Safety reauirements are in bid rackaees. Contractors have site-SDCcific health and safety



plans.







3.

O&M and OSHA Training Records

Remarks:

1X1 Readily available

1X1 Up to date K

N/A

4.

Permits and Service Agreements









1 1 Air discharge permit

1 1 Readily available

1 1 Up to date K

N/A



~ Effluent discharge

~ Readily available

~ Up to date ^

N/A



~ Waste disposal, POTW

~ Readily available

~ Up to date K

N/A



I-! Other permits

1 1 Readily available

1 1 Up to date K

N/A



Remarks:







5.

Gas Generation Records

Remarks:

~ Readily available

~ Up to date K

N/A

6.

Settlement Monument Records

Remarks:

1 1 Readily available

1 1 Up to date K

N/A

7.

Groundwater Monitoring Records

Remarks:

~ Readily available

~ Up to date K

N/A

8.

Leachate Extraction Records

~ Readily available

~ Up to date ^

N/A



Remarks:







9.

Discharge Compliance Records









~ Air ~ Readily available

~ Up to date

IE|n/a





1 1 Water (effluent) ~ Readily available

1 1 Up to date

1X1 N/A





Remarks:







D-9


-------
10. Daily Access/Security Logs	Q Readily available ~ Up to date Kl N/A

Remarks:	

IV. O&M COSTS

1.	O&M Organization

153 State in-house	Contractor for State

I I PRP in-house	EH Contractor for PRP

~	Federal Facility in-house	Q Contractor for Federal Facility

	~		

2.	O&M Cost Records

153 Readily available	O Up to date

~ Funding mechanism/agreement in place Q Unavailable

3. Unanticipated or Unusually High O&M Costs During Review Period

Describe costs and reasons: None.

V. ACCESS AND INSTITUTIONAL CONTROLS M Applicable ~ N/A

A.	Fencing

1. Fencing damaged	~ Location shown on site map ~ Gates secured ~ N/A

Remarks: Fencing is located throughout Reach A to protect revegetation efforts from humans and
wildlife.

B.	Other Access Restrictions

1. Signs and other security measures	~ Location shown on site map ~ N/A

Remarks: Signs in Reach A notify people of access restrictions. However, there are no warning signs
anywhere else (including Reaches B and CI.

C.	Institutional Controls (ICs)

D-10


-------
1.

Implementation and enforcement

Site conditions imply ICs not properly implemented Q Yes ^ No D N/A
Site conditions imply ICs not being fully enforced ~ Yes 15^1 No | | N/A
Tvpe of monitorino (e.s.. self-reportine. drive bvl
Frequency

Responsible party /agency
Contact mm/dd/vvw

Name Title Date Phone no.
Reporting is up-to-date ~ Yes EH No N/A
Reports are verified by the lead agency O Yes Q No ^ N/A
Specific requirements in deed or decision documents have been met ^ Yes Q No Q N/A
Violations have been reported O Yes Q No ^ N/A
Other problems or suggestions: Q Report attached

The Powell Creek Overlay District covers the area contaminated bv minine and smeltins wastes from
operations further iiDStrcam in the Butte and Anaconda areas. The Overlay District is intended to
ensure that future land use in the SuDcrfund Overlay District is compatible with the presence of
potential contaminants and the remedial actions reauired to isolate those potential contaminants from
the environment.

2.

Adequacy ^ ICs are adequate ~ ICs are inadequate ~ N/A
Remarks: As additional remedial actions are completed, additional institutional controls mav be needed in



other areas of the Site.

D.

General

1.

Vandalism/trespassing Q Location shown on site map ^ No vandalism evident
Remarks:

2.

Land use changes on site ^ N/A

Remarks:

3.

Land use changes off site ^ N/A

Remarks:

VI. GENERAL SITE CONDITIONS

A.

Roads Applicable ~ N/A

1.

Roads damaged ~ Location shown on site map ^ Roads adequate ~ N/A
Remarks: Roads are kept sraded and wet to limit dust.

B.

Other Site Conditions

Remarks:

VII. LANDFILL COVERS ^Applicable ~ N/A

A.

Landfill Surface

D-ll


-------
1.

Settlement (Low spots)

1 1 Location shown on site map

Settlement not evident



Arial extent



Depth



Remarks:





2.

Cracks

1 1 Location shown on site map

1^1 Cracking not evident



Leneths

Widths

Depths



Remarks:





3.

Erosion

1 1 Location shown on site map

1^1 Erosion not evident



Arial extent



Depth



Remarks:





4.

Holes

~ Location shown on site map

Holes not evident



Arial extent



Depth



Remarks:





5.

Vegetative Cover

1 1 Grass

Cover properly established



I~1 No signs of stress

1 1 Trees/Shrubs (indicate size and locations on a diagram)



Remarks: East Side Road rasture was recently reveeetated and lined.

6.

Alternative Cover (armored rock, concrete, etc.)

IEIn/a



Remarks:





7.

Bulges

~ Location shown on site map

E3 Bulges not evident



Arial extent



Heieht



Remarks:





8.

Wet Areas/Water

^ Wet areas/water damage not evident

Damage







n Wet areas

1 1 Location shown on site map

Arial extent



1 1 Ponding

1 1 Location shown on site map

Arial extent



I~1 Seeps

1 1 Location shown on site map

Arial extent



1 1 Soft subgrade

1 1 Location shown on site map

Arial extent



Remarks:





9.

Slope Instability

1 1 Slides

~ Location shown on site map



1^1 No evidence of slope instability





Arial extent







Remarks:





B.

Benches ~ Applicable ^ N/A





(Horizontally constructed mounds of earth placed across a steep landfill side slope to interrupt the slope in



order to slow down the velocity of surface runoff and intercept and convey the runoff to a lined channel.)

D-12


-------
C.	Letdown Channels	~ Applicable ^ N/A

(Channel lined with erosion control mats, riprap, grout bags, or gabions that descend down the steep side
slope of the cover and will allow the runoff water collected by the benches to move off of the landfill
cover without creating erosion gullies.)

D.	Cover Penetrations	~ Applicable ^ N/A

E.	Gas Collection and Treatment	~ Applicable ^ N/A

F.	Cover Drainage Layer	~ Applicable ^ N/A

G.	Detention/Sedimentation Ponds ~ Applicable	^ N/A

H.	Retaining Walls	~ Applicable ^ N/A

I.	Perimeter Ditches/Off-Site Discharge	~ Applicable ^ N/A

VIII.	VERTICAL BARRIER WALLS	~ Applicable N/A

IX.	GROUNDWATER/SURFACE WATER REMEDIES ~ Applicable	N/A

A.	Groundwater Extraction Wells, Pumps, and Pipelines	~ Applicable ^ N/A

B.	Surface Water Collection Structures, Pumps, and Pipelines ~ Applicable ^ N/A

C.	Treatment System	~ Applicable ^ N/A

X. OTHER REMEDIES

If there are remedies applied at the site and not covered above, attach an inspection sheet describing the physical
nature and condition of any facility associated with the remedy. An example would be soil vapor
extraction.

	XI. OVERALL OBSERVATIONS	

A. Implementation of the Remedy	

Describe issues and observations relating to whether the remedy is effective and functioning as designed.
Begin with a brief statement of what the remedy is to accomplish (i.e., to contain contaminant plume,
minimize infiltration and gas emission, etc.).

Remedy implementation is ongoing. Remediation of Phase 1 of Reach A finished in April 2014. Long-
term monitoring is underway to assess groundwater, surface water and vegetation during remediation.

B. Adequacy of O&M	

Describe issues and observations related to the implementation and scope of O&M procedures. In
particular, discuss their relationship to the current and long-term protectiveness of the remedy.
Not applicable.

C.	Early Indicators of Potential Remedy Problems	

Describe issues and observations such as unexpected changes in the cost or scope of O&M or a high
frequency of unscheduled repairs that suggest that the protectiveness of the remedy may be compromised
in the future.

	None idcntiFicd.	

D.	Opportunities for Optimization	

Describe possible opportunities for optimization in monitoring tasks or the operation of the remedy.

None identified.

Site Inspection Team:

• Sara Sparks, EPA

D-13


-------
Brian Bartkowiak, MDEQ
Claire Marcussen, Skeo
Treat Suomi, Skeo


-------
Appendix E-l: Photographs from MRSOU Site Inspection

"fecv-

r=roij

E-l


-------
Clark Fork river from bluff at Milltown State Park. Includes views of several site repositories, including
the Right Bank Repository and the lnterstate-90 slope.

E-2


-------
A-

a

i, .

t-



>|r*

¦m
1m

r "T ^1'

*

i.-fca&rj XUlSB* .««.'"•¦ r^3PW»W



.j & jk -Mb •-•

**2fSsi rj

i* j

j I '"i I i '• ¦' -

•L -h I -	'



View of the Sheriff Posse Grounds Parcel from Milltown Bluff.

E-3


-------
The Tunnel Pond Repository.

E-4


-------
View of the lnterstate-90 slope.

E-5


-------
Orange cone marking hole in the Right Bank Repository.

E-6


-------
View of the Clark Fork River.

E-7


-------
Flagging marking area of subsidence at the Right Bank Repository.

E-8


-------
IM

E-9


-------
Timbers for use in park construction

E-10


-------
ENVIRONMENTALLY
SENSITIVE AREA

PLEASE
KEEP OUT

Keep Out" sign at the Miiltown Reservoir revegetation area

E-ll


-------
E-12


-------
Rodeo grounds at the Sheriff Posse Grounds Parcel.

E-13


-------
Milltown Bluff and the Tunnel Pond Repository.

E-14


-------
Sundial at Bonner Learning Park.

E-15


-------
Appendix E-2: Photographs from the CFROU Site Inspection

CFROU Phase 1 remediation area

E-16


-------
CFROU Phase 1 remediation area.

E-17


-------
t

Sign for river closure at the CFROU.

E-18


-------
Phase 2 remediation in progress.

E-19


-------
View of Arrow Stone Park.

E-20


-------
Riverbank at Arrow Stone State Park


-------
Residential area in Deer Lodge.

E-22


-------
Trestle area in Deer Lodge.

E-23


-------
KOA property, Clark Fork River and residential trailers.

E-24


-------
East Side Road and pasture remediation revegetation.

E-25


-------
Downstream of Phase 7.

E-26


-------
Racetrack Pond.

E-27


-------
Signage along road due to remedial work.

E-28


-------
Revegetation crew.


-------
Revegetation area.

E-30


-------
Grant-Kohrs Ranch National Historic Site.

E-31


-------
E-32


-------
USGS gauge station along the Clark Fork River.

E-33


-------
View of Clark Fork River along Reach C.

E-34


-------
Appendix F: MRSOU Monitoring Data

Table F-l. Historic Dissolved Arsenic Concentration Data Summary, 2008 to 2015

Compliance
Well
Number

Dissolved Arsenic

(Hg/L)

Mmilh Ye;ir

.lllll-
UX

J;m-
n<)

luii-
O'J

IKv-
O'J

.luii-
lu

.l:m-
1 1

.luii-
1 1

Ikv-
1 1

.Iiiii-
12

.l:m-
1 '

Jllll- 1 '

Ikv-
1 '

Imii- 14

Ikv-
14

Imii-15

J;ni-I<>

105C

' :4

^ l(>

^ :4

: <>x

: v.

: 46

2.29

2.03

2.29

1.71

2.08

1.87

2.03

1.99

1.87

1.85

11R











22.4 2

10.9

23.3

23.3

19.2

22.6

18.8

20.7

21.9

24.8

21.4

HOB

10.40

10.25

10.5

10.6

9.5

9.0

9.17

9.19

9.92

7.70

8.84

7.75

8.47

8.02

9.17

7.79

922D

11.50

12.60

13.8

12.5

12.0

12.4

12.8

11.6

12.8

9.29

12.50

9.91

12.1

10.1

12.2

9.60

107A







66.50

57.6

42.9

39.5

38.4

45.0

27.7

11.5

26.2

31.4

25.2

28.5

22.9

104A









11.4



11.8

9.8

11.3

10.2

9.96

9.25

10.9

11.70

11.9

13.9

103B

25.40

25.90

30.2

27.3

29.0

21.8

22.9

17.7

25.9

16.0

13.7

18.7

23.2

20.6

21.8

18.4

HLA-2

45.50



43.50

35.6

31.8

26.3

27.1

22.3

34.8

34.1

21.2

10.3

11.2

11.0

12.3

10.9

917B

I5S

162

133

148

125

116

108

85

97.6

61.1

47.4

57.5

74.9

55.0

67.4

52.8

921A







7.35

6.63

7.19

8.42

6.41

8.80

6.50

8.02

7.09

10.6

7.85

9.12

9.53

TPR-10









: V,

s::

" "X

16.6

12.7

21.4

12.1

21.1

19.0

25.8

18.8

12.5

913A



















0.649

j

0.814j

1.12j

0.994j

1.13j

0.867J

1.12j

11 l

4

(K.S

^ 42ii

i n

4 4^

u44<>

" "u

2 '()

(. 51

well has boon replaced by " 11R"

104B l







0.18



1 X5







well has been replaced by "104A"

905 l

18.10

76.80

r.6o

46.8

18.1

45.9

X (i 1

DRY

"'4.9

well is damaged, no longer on compliance lisi

107C l

13.40

15.30

14.50

15.8











well has been replaced by "107A"

Notes:

J = analyte detected below the reporting limit

1	= former compliance well

2	= sampled on March 1, 2011

3	= sampled on July 1, 2011

ND = not detected above the method detection limit (MDL)
Bold denotes exceedance of MCL

F-l


-------
Table 2. Surface Water Data Summary, 2015



Arsenic

Arsenic

Cadmium

Cadmium

Copper

Copper

Lead

Lead

Zinc

Zinc

Hardness



DISS

TOT

DISS

TOT

DISS

REC

DISS

REC

DISS

REC

NA

Units

mq/l

mq/l

mq/l

M9/L

mq/l

mq/l

M9/L

mq/l

mq/l

mq/l

mg/L

12340500 - Clark Fork River near Missoula

3/25/2015

1.9

2.5

< 0.030

0.03

1.3

4.7

0.07

0.77

< 2.0

7.7

96.5

4/22/2015

2.2

2.6

< 0.030

< 0.030

1.7

4.2

0.04

0.59

< 2.0

6.5

99.8

5/13/2015

2

1.9

< 0.030

< 0.030

1.6

2.4

0.14

0.30

< 2.0

3.1

89.6

5/28/2015

3.3

5.8

< 0.030

0.06

3

8.6

0.19

1.11

2.1

3.9

95.2

6/10/2015

3.7

4.8

< 0.030

0.10

2.2

10.8

0.07

1.51

< 2.0

13.1

105

7/15/2015

3.1

3.3

< 0.030

< 0.030

1.9

2.8

< 0.040

0.17

< 2.0

3.5

129

8/12/2015

3.3

3.6

< 0.030

< 0.030

1.6

2.8

< 0.040

0.30

< 2.0

4.3

143

10/21/2015

4

4.6

< 0.030

< 0.030

1.3

3.7

< 0.040

0.51

< 2.0

5.8

158

12340000 - Blackfoot River near Bonner

4/22/2015

0.8

0.89

< 0.030

< 0.030

< 0.80

0.9

< 0.040

0.15

< 2.0

< 2.0

93.2

5/13/2015

0.79

0.82

< 0.030

< 0.030

< 0.80

< 0.80

< 0.040

0.11

< 2.0

< 2.0

89.9

5/28/2015

0.96

1.1

< 0.030

< 0.030

0.96

1.3

0.167

0.39

< 2.0

< 2.0

95.9

7/15/2015

1.2

1.3

< 0.030

< 0.030

< 0.80

< 0.80

< 0.040

0.05

< 2.0

< 2.0

132

8/12/2015

1.2

1.4

< 0.030

< 0.030

< 0.80

< 0.80

< 0.040

0.06

< 2.0

< 2.0

133

10/21/2015

1.1

1.4

< 0.030

< 0.030

< 0.80

< 0.80

< 0.040

< 0.04

< 2.0

< 2.0

142

12334550 - Clark Fork River atTurah Bridge

3/25/2015

3.6

5.2

0.03

0.077

2.6

12

0.11

1.85

3.3

16.8

110

4/22/2015

4.1

5.4

< 0.030

0.054

3

9.3

0.09

1.26

2.3

12.7

111

5/13/2015

3.5

3.9

< 0.030

< 0.030

2

4.1

0.06

0.45

< 2.0

4.9

85.1

5/28/2015

7.3

9.9

0.08

0.2

17.9

37.8

2.79

6.16

17.4

44.6

100

6/10/2015

7.4

9.6

< 0.030

0.15

4.3

27.9

0.17

3.94

5.1

30.7

110

7/15/2015

4.8

5.5

< 0.030

< 0.030

2.7

5

< 0.040

0.39

< 2.0

4.9

133

8/12/2015

5.2

6

< 0.030

0.041

2

5.2

0.042

0.56

< 2.0

8.2

155

10/21/2015

5.9

6.7

< 0.030

0.037

1.8

7.2

< 0.040

1.05

2.4

10.5

179

Notes:

Bold values denote exceedance of Montana DEQ-7 surface water standard which are measured as total recoverable concentrations.

Bold italic values denote exceedance of federal standards which are measured as dissolved.

The performance standard for copper is derived from the federal water quality criteria measured as dissolved.

F-2


-------
Appendix G: CFROU Monitoring Data Summary

Surface Water

Arsenic and copper are the site COCs in surface water with regular exceedances. Of 30 samples
collected in the mainstem Clark Fork River in 2014, no samples had zinc concentrations
exceeding the performance goal. One sample had cadmium concentrations exceeding the
performance goal. Four samples had lead concentrations exceeding the performance goal.
However, arsenic commonly exceeded performance goals, particularly in Reach A. Of 24
samples collected in the Clark Fork River in Reach A, 96 percent exceeded the dissolved arsenic
and 46 percent exceeded the total recoverable arsenic performance goals.

Silver Bow Creek and the Mill-Willow Creek appear to be sources of arsenic to the Clark Fork
River; 17 of 18 of the samples from those sites exceeded the dissolved arsenic and 14 of 18
samples from those sites exceeded the total recoverable performance goals. Total recoverable
copper concentration exceeded State of Montana chronic aquatic life standard in the mainstem
Clark Fork River sites in 95 percent of the samples collected in the first and second quarters, but
only at Deer Lodge in the third and fourth quarters. These results support the conclusion that
copper contamination in the upper Clark Fork River is strongly related to streamflow and
contaminant loading occurs primarily in Reach A.

G-l


-------
CFR-116A

Sediment.

The highest instream sediment COC concentrations in the mainstem of the Clark Fork River
were typically observed in the uppermost sample sites in Reach A and the lowest concentrations
were typically observed at the downstream-most site at Turah in 2014. Concentrations of arsenic,
copper, and zinc exceeded the probable effect concentration (PEC) at all of the Clark Fork River
mainstem monitoring stations during both sample periods in 2014. Among all sites in the
CFROU, arsenic most commonly exceeded the PEC (88 percent) followed by copper (83
percent), lead (79 percent), zinc (75 percent) and cadmium (50 percent).

G-2


-------
Geomorphology

Geomorphology data were collected during the third quarter of 2014 in Phase 1 of Reach A in
the CFROU. All monitoring metrics for channel dimension (i.e., cross-sectional area, bankfull
width, mean bankfull depth and width-to-depth ratio), pool density and residual pool depth were
within specified target ranges. The secondary channel stability performance target was also met
because the secondary channel did not carry more than 10 percent of the streamflow of the main
channel when streamflows reached the design bankfull level. Performance targets that were not
met included floodplain connectivity and floodplain stability. Failure to meet the performance
targets for channel connectivity and floodplain stability was the result of an over-connected river
channel and floodplain, which results in increased avulsion risk, rather than the disconnected
pre-project channel and floodplain. Performance targets for channel slope, sinuosity, bank
erosion rate and channel migration rate were not scheduled for monitoring in Year 1 (2014).
They will be evaluated in Year 5 (2018).

Vegetation Monitoring Data

Vegetation monitoring data were collected during the third quarter of 2014 in Phase 1 of Reach
A in the CFROU. The only vegetation monitoring metric applicable to Year 1 monitoring was
for overall floodplain plant survival which was 87.7 percent, exceeding the performance target
for Year 1 (80 percent). However, survival was 17.2 percent lower in in the floodplain riparian
shrub cover type (primarily consisting of swales) compared to the other floodplain cover types
and survival of planted birch trees (Betula occidentalis) was particularly low. Low survival in
swales may have been caused by the relatively deep swale excavation in combination with
prolonged flood inundation which resulted in drowning. Other monitoring metrics with Year 1
performance targets (floodplain total native cover and noxious weed cover) will be monitored in
2015. Some floodplain plant survival monitoring plots will be monitored for plant survival in
2015 in planting units that had not yet been planted at the time of monitoring in 2014.

Macroinvertebrate

Overall biotic integrity of the macroinvertebrate community was either "none" or "slight" at all
Clark Fork River tributary and mainstem sites; overall biointegrity scores throughout the
CFROU ranged from 84.1 to 90.9. For metals sensitivity, index classifications in the mainstem
were "none" at all sites except at Gemback Road which was "slight"; metals sensitivity scores in
the mainstem ranged from 75.0 to 87.5. Metals sensitivity index classifications in the tributary
sites was "moderate" at Racetrack Creek and Warm Springs Creek, "slight" in Silver Bow Creek
and the Little Blackfoot River, and "none" in Mill-Willow Creek and Lost Creek; metals
sensitivity scores in the tributaries ranged from 56.9 to 88.9. Nutrient sensitivity index
classifications were "none" at all CFROU sites, with scores ranging from 81.9 to 100.0.

Periphyton

Periphyton monitoring results revealed that many of the non-diatom algae observed in the
CFROU were tolerant to elevated nutrients, acidity, metals, or combinations of those conditions.
However, diatom algae dominated the periphyton assemblage at all CFROU sites monitored in

G-3


-------
2014 and periphyton samples were scored according to several bioassessment indices.
Impairment from sediment was more likely than not (i.e., >51 percent) in three tributary sites
(Mill-Willow Creek, 93 percent; the Mill-Willow Bypass, 77 percent; and Silver Bow Creek,
81%) and four mainstem sites (near Galen, 88 percent; at Galen Road, 57 percent; at Gemback
Road, 79 percent; and at Deer Lodge, 93 percent). Impairment from metals was more likely than
not (i.e., >51 percent) in one tributary site (Silver Bow Creek, 74 percent) and four mainstem
sites (near Galen, 74 percent; at Galen Road, 88 percent; at Gemback Road, 76 percent; and at
Turah, 94 percent).

Fish

Fish Population

Based on fish population monitoring in the Clark Fork River, brown trout continue to dominate
the trout species assemblage in the upper Clark Fork River. This is presumably due, at least in
part, to their relatively high tolerance to metals compared to other salmonids. Brown trout
populations appear to be moderately increasing since 2011 at monitoring sites in the mid- and
upper-reaches of the Clark Fork River. Trout abundance in the Bearmouth reach remained low in
2014, as in prior years, relative to other reaches of the upper Clark Fork River. It is possible that
above average discharge in 2011 increased the quality and quantity of brown trout spawning and
rearing habitat in the upper Clark Fork River and tributaries, resulting in the modest increase in
trout abundance in 2014.

Caged Fish

Results of survival monitoring of caged juvenile brown trout indicated that, as in previous
survival studies in the upper Clark Fork River, mortality rates varied among sites and among
months. Most of the mortality in 2014 in the caged fish occurred in April, July and August. This
bimodal pattern was consistent with results from caged fish studies in 2012 and 2013. Mortality
tended to be highest during spring runoff and on the descending limb of the hydrograph as water
temperatures increased. Brown trout confined in the cages accumulated both copper and zinc in
their tissues at both mainstem Clark Fork River and tributary sites. Tissue burdens of fish
immediately after release from the hatchery were low compared to fish sampled from cages in
the CFROU. Fish from cages in the mainstem had significantly higher metals burdens compared
to fish from tributaries, but the difference was less pronounced for zinc.

G-4


-------
Appendix H: MRSOU Sediment Accumulation Areas

PART 1: DECLARATION

Approximate Sediment
Accumulation Area Boundary

Sediment Pore Water Arsenic
>0 1 mgA. {Approximate
Source Sediment Area
for alluvial aquifer 0 02 mg/L
arsenic plume)

SOURCE ARCO Remedial Study. 2001
EXHBIT1-2

Key Sediment Accumulation Areas

H-l


-------
PART 2, DECISION SUMMARV: SECTION 12—SELECTED REMEDY

Earthen Barm

Prevents protected
lower CFR
sediments from
slumping into nver

SAA III Sediments
Behind Cofferdam

To be removed
with dam spillway

Radial Gate

Topsail Salvage

SAA I sediment
surface material may be
salvaged lo provide growlh
media for project area
reclamation

'90 Grade Control

. SAA I Sediments To Se Removed

After contaminated sediments
are removed. SAA I will be
backfilled and new nver
channels/floodplains constructed

Sorrow Area

Will be used to
haul sediments to
Opportunity Ponds
for disposal

Borrow Area

Tunnel Pond Area

May be used as a disposal
area for debris and SAA III
sediments generated by
dam removal

Armored Embankment

Isolates sediments from Ihe CFR during

excavatiareconstruction activities
	 • '. jr—TTTTTmmm

Mllltown Dam

U To be removed

Cofferdam

Bypass C flannel

Games the CFR flow
during excavalton'
construction activities

Permits spillway
removal in Ihe dry

- •

Diversion Dike

Protects sediments in
downstream portion of
existing CFR channel
from scour by routing CFR
flow throiKjht bypass channel

New or Refurbished Rail Spurs



Permit excavated sediments
to be hauled on existing rail line

Will be excavated to
provide backfill material
reclaimed for existing trees

Note:

Removal process awl detailed volumes
will be determined during remedial design

EXUBIT 2-31

Conceptual IVbdel of Remedial
Cleanup Han

amended from

ElVWiROCOlU

101 WTERNAT1QNAL WAV
MISSOULA MONTANA 9H06

H-2


-------
Appendix I: Surface Water Data Evaluation

1-1


-------
Prepared in cooperation with the U.S. Environmental Protection Agency

Water-Quality Trends and Constituent-Transport Analysis
for Selected Sampling Sites in the Milltown Reservoir/Clark
Fork River Superfund Site in the Upper Clark Fork Basin,
Montana, Water Years 1996-2015

Scientific Investigations Report 2016-5100

U.S. Department of the Interior
U.S. Geological Survey


-------

-------
Water-Quality Trends and Constituent-
Transport Analysis for Selected Sampling
Sites in the Milltown Reservoir/Clark Fork
River Superfund Site in the Upper Clark
Fork Basin, Montana, Water Years
1996-2015

By Steven K. Sando and Aldo V. Vecchia

Prepared in cooperation with the U.S. Environmental Protection Agency

Scientific Investigations Report 2016-5100

U.S. Department of the Interior
U.S. Geological Survey


-------
U.S. Department of the Interior

SALLY JEWELL, Secretary

U.S. Geological Survey

Suzette M. Kimball, Director

U.S. Geological Survey, Reston, Virginia: 2016

For more information on the USGS—the Federal source for science about the Earth, its natural and living
resources, natural hazards, and the environment—visit http://www.usgs.gov or call 1—888—ASK—USGS.

For an overview of USGS information products, including maps, imagery, and publications,
visit http://store.usgs.gov.

Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the
U.S. Government.

Although this information product, for the most part, is in the public domain, it also may contain copyrighted materials
as noted in the text. Permission to reproduce copyrighted items must be secured from the copyright owner.

Suggested citation:

Sando, S.K., and Vecchia, A.V., 2016, Water-quality trends and constituent-transport analysis for selected sampling
sites in the Milltown Fleservoir/Clark Fork Fliver Superfund Site in the upper Clark Fork Basin, Montana, water years
1996-2015: U.S. Geological Survey Scientific Investigations Fleport 2016-5100, 82 p., http://dx.doi.org/10.3133/
sir20165100.

ISSN 2328-0328 (online)


-------
iii

Contents

Abstract	1

Introduction	2

Purpose and Scope	4

Description of Study Area	4

Data-Collection and Analytical Methods	6

Quality Assurance	6

Overview of Streamflow and Water-Quality Characteristics for Water Years 2011-15	8

General Streamflow Characteristics for Water Years 2011—15	8

Water-Quality Characteristics for Water Years 2011-15	8

Reach 4	18

Reach 5	20

Reach 6	20

Reach 7	21

Reach 8	21

Reach 9	21

Water-Quality Trend- and Constituent-Transport Analysis Methods	22

General Description of the Time-Series Model	22

Selection of Trend-Analysis Time Periods	23

Estimation of Normalized Constituent Loads	23

Factors that Affect Trend Analysis and Interpretation	24

Streamflow Conditions	24

Other Factors	24

Water-Quality Trends and Constituent-Transport Analysis Results	28

Water-Quality Trends in Flow-Adjusted Concentrations	28

Copper	38

Arsenic	38

Suspended Sediment	38

Overview of Water-Quality Trend Results	38

Constituent-Transport Analysis Results	39

Copper	45

Arsenic	45

Suspended Sediment	45

Overview of Constituent-Transport Analysis Results	46

Summary and Conclusions	47

References	48

Appendix 1—Summary Information Relating to Quality-Control Data	52

Appendix 2—Summary of the Time Series Model as Applied in this Study	61

Theoretical and Computational Information	61

Specific Aspects of the Application of the Time-Series Model in this Study	62

Appendix 3—Trend-Analysis Results	65

Appendix 4—Transport-Analysis Balance Calculations for Data-Summary Reaches	76


-------
iv

Figures

1,	Map showing location of study area, selected sampling sites, and data-surnmary
reaches in the upper Clark Fork Basin, Montana; the Milltown Reservoir/Clark Fork
River Superfund Site includes the reaches from sampling site 8 to sampling site 22	3

2,	Graphs showing statistical distributions of selected constituents for selected
sampling sites in the Milltown Reservoir/Clark Fork River Superfund Site in the

upper Clark Fork Basin, Montana, water years 2011-15	19

3,	Graphs showing daily mean streamflow for selected sampling sites in the Milltown
Reservoir/Clark Fork River Superfund Site in the upper Clark Fork Basin, Montana,
water years 1993-2015	25

4,	Graphs showing selected streamflow and constituent concentration information

for Clark Fork near Galen, Montana (sampling site 11), water years 1993-2015	27

5,	Graphs showing flow-adjusted fitted trends for selected constituents for sampling
sites in reach 4, extending from Silver Bow Creek at Warm Springs, Montana
(sampling site 8), to Clark Fork near Galen, Montana (sampling site 11),

water years 1996-2015	32

6,	Graphs showing flow-adjusted fitted trends for selected constituents for
sampling sites in reach 5, extending from Clark Fork near Galen, Montana
(sampling site 11), to Clark Fork at Deer Lodge, Montana (sampling site 14),

water years 1996-2015	33

7,	Graphs showing flow-adjusted fitted trends for selected constituents for
sampling sites in reach 6, extending from Clark Fork at Deer Lodge, Montana
(sampling site 14), to Clark Fork at Goldcreek, Montana (sampling site 16),

water years 1996-2015	34

8,	Graphs showing flow-adjusted fitted trends for selected constituents for
sampling sites in reach 7, extending from Clark Fork at Goldcreek, Montana
(sampling site 16), to Clark Fork near Drummond, Montana (sampling site 18),

water years 1996-2015	35

9,	Graphs showing flow-adjusted fitted trends for selected constituents for
sampling sites in reach 8, extending from Clark Fork near Drummond, Montana
(sampling site 18), to Clark Fork atTurah Bridge near Bonner, Montana

(sampling site 20), water years 1996-2015	36

10,	Graphs showing flow-adjusted fitted trends for selected constituents for
sampling sites in reach 9, extending from Clark Fork atTurah Bridge near Bonner,
Montana (sampling site 20), to Clark Fork above Missoula, Montana

(sampling site 22), water years 1996-2015	37

11,	Pie charts representing geometric mean streamflow and estimated normalized
unfiltered-recoverable copper loads contributed from reach inflow and

within-reach sources for data-summary reaches for selected periods	42

12,	Pie charts representing geometric mean streamflow and estimated normalized
unfiltered-recoverable arsenic loads contributed from reach inflow and
within-reach sources for data-summary reaches for selected periods	43

13,	Pie charts representing geometric mean streamflow and estimated normalized
suspended-sediment loads contributed from reach inflow and within-reach

sources for data-summary reaches for selected periods	44


-------
V

Appendix Figures

1-1, Graphs showing spike recoveries for laboratory-spiked deionized-water blank

samples, water years 1993-2015	59

1-2. Graphs showing spike recoveries for laboratory-spiked stream-water samples,

water years 1993-2015	60

3-1, Graphs showing flow-adjusted fitted trends for selected water-quality constituents and
properties for Silver Bow Creek at Warm Springs, Montana (sampling site 8), water

years 1996-2015	69

3-2. Graphs showing flow-adjusted fitted trends for selected water-quality constituents
and properties for Clark Fork near Galen, Montana (sampling site 11), water years

1996-2015	70

3-3, Graphs showing flow-adjusted fitted trends for selected water-quality constituents
and properties for Clark Fork at Deer Lodge, Montana (sampling site 14), water

years 1996-2015	71

3-4, Graphs showing flow-adjusted fitted trends for selected water-quality constituents
and properties for Clark Fork at Goldcreek, Montana (sampling site 16), water years

1996-2015	72

3-5, Graphs showing flow-adjusted fitted trends for selected water-quality constituents
and properties for Clark Fork near Drummond, Montana (sampling site 18), water

years 1996-2015	73

3-6. Graphs showing flow-adjusted fitted trends for selected water-quality constituents
and properties for Clark Fork atTurah Bridge near Bonner, Montana (sampling site

20), water years 1996-2015	74

3-7. Graphs showing flow-adjusted fitted trends for selected water-quality constituents
and properties for Clark Fork above Missoula, Montana (sampling site 22), water
years 1996-2015	75

Tables

1.	Information for selected sampling sites and data-summary reaches in the Milltown
Reservoir/Clark Fork River Superfund Site in the upper Clark Fork Basin, Montana	5

2.	Properties, constituents, and associated information relating to laboratory and

study reporting levels	7

3.	Statistical summaries of continuous streamflow data for selected sampling sites in
the Milltown Reservoir/Clark Fork River Superfund Site in the upper Clark Fork

Basin, Montana	9

4.	Statistical summaries of water-quality data collected at selected sampling sites in
the Milltown Reservoir/Clark Fork River Superfund Site in the upper Clark Fork

Basin, Montana, water years 2011-15	10

5.	Percentages of samples with unadjusted unfiltered-recoverable concentrations
exceeding water-quality standards for selected sampling sites in the Milltown
Reservoir/Clark Fork River Superfund Site in the upper Clark Fork Basin, water

years 2011-15	18

6.	Summary of flow-adjusted trend results for selected sampling sites and
constituents, water years 1996-2015 	29

7.	Summary of flow-adjusted trend results for Clark Fork above Missoula, Montana
(sampling site 22), for selected constituents, water years 1996-2015	31


-------
vi

8. Drainage area and streamflow information relevant to the transport analysis for
data-summary reaches in the Milltown Reservoir/Clark Fork River Superfund Site in
the upper Clark Fork Basin, Montana, water years 1996-2015	40

Appendix Tables

1-1. Summary information relating to quality-control samples (field equipment blank
and replicate samples) collected atsampling sites in the upper Clark Fork Basin,

Montana, water years 1993-2015	53

1-2. Summary information relating to quality-control samples (laboratory-spiked

deionized-water blank samples) collected atsampling sites in the upper Clark Fork

Basin, Montana, water years 1993-2015	54

1-3. Summary information relating to quality-control samples (laboratory-spiked

stream-water samples) collected at sampling sites in the upper Clark Fork Basin,
Montana, water years 1993-2015	56

1-4,	Aquatic-life standards (based on median hardness for water years 2011-15) for
selected sampling sites in the Milltown Reservoir/Clark Fork River Superfund Site

in the upper Clark Fork Basin, Montana	58

2-1.	Statistical summaries of standard errors of estimates for the trend models	64

3-1,	Flow-adjusted trend results for selected water-quality constituents and properties
for selected sampling sites in the Milltown Reservoir/Clark Fork River Superfund

Site in the upper Clark Fork Basin, Montana, water years 1996-2015	 66

3-2.	Flow-adjusted trend results for selected water-quality constituents and properties

for Clark Fork above Missoula, Montana (sampling site 22), water years 1996-2015	68

4-1,	Constituent-transport analysis balance calculations for sampling sites in reach 4,
extending from Silver Bow Creek atWarm Springs, Montana (sampling site 8), to
Clark Fork near Galen, Montana (sampling site 11), for selected periods, water

years 1996-2015	77

4-2, Constituent-transport analysis balance calculations for sampling sites in reach 5,
extending from Clark Fork near Galen, Montana (sampling site 11), to Clark Fork at
Deer Lodge, Montana (sampling site 14), for selected periods, water years

1996-2015	78

4-3, Constituent-transport analysis balance calculations for sampling sites in reach 6,
extending from Clark Fork at Deer Lodge, Montana (sampling site 14), to Clark Fork
at Goldcreek, Montana (sampling site 16), for selected periods, water years

1996-2015	79

4-4, Constituent-transport analysis balance calculations for sampling sites in reach 7,
extending from Clark Fork at Goldcreek, Montana (sampling site 16), to Clark Fork
near Drummond, Montana (sampling site 18), for selected periods, water years

1996-2015	80

4-5. Constituent-transport analysis balance calculations for sampling sites in reach 8,
extending from Clark Fork near Drummond, Montana (sampling site 18), to Clark
Fork atTurah Bridge near Bonner, Montana (sampling site 20), for selected periods,

water years 1996-2015	81

4-6. Constituent-transport analysis balance calculations for sampling sites in reach 9,
extending from Clark Fork atTurah Bridge near Bonner, Montana (sampling site 20),
to Clark Fork above Missoula, Montana (sampling site 22), for selected periods,
water years 1996-2015	82


-------
Conversion Factors

U.S. customary units to International System of Units

Multiply

By

To obtain

Length

inch (in.)

2.54

centimeter (cm)

inch (in.)

25.4

millimeter (mm)

foot (ft)

0.3048

meter (m)

mile (mi)

1.609

kilometer (km)

Area

square mile (mi2)

259.0

hectare (ha)

square mile (mi2)

2.590

square kilometer (km2)

Volume

gallon (gal)

3.785

liter (L)

Flow rate

cubic foot per second (ft3/s)

0.02832

cubic meter per second (m3/s)

Mass

pound, avoirdupois (lb)

0.4536

kilogram (kg)

Supplemental Information

Specific conductance is given in rnicrosiemens per centimeter at 25 degrees Celsius (pS/cm).

Concentrations of chemical constituents in water are given either in micrograms per liter (pg/Lj
or milligrams per liter (mg/L).

Load estimates are given in kilograms per day (kg/d).

Water year is defined as the 12-month period from October 1 through September 30 of the
following calendaryear. The water year is designated by the calendar year in which it ends. For
example, water year 2010 is the period from October 1,2009, through September 30, 2010.


-------
Abbreviations

AMC

Anaconda Mining Company

FAC

flow-adjusted concentration

LRL

laboratory reporting level

LOWESS

locally weighted scatter plot smooth

NWQL

National Water Quality Laboratory

NWIS

National Water Information System

SEE

standard error of estimate

SRI

study reporting level

ISM

time -series model

USGS

U.S. Geological Survey


-------
Water-Quality Trends and Constituent-Transport Analysis
for Selected Sampling Sites in the Milltown Reservoir/
Clark Fork River Superfund Site in the Upper Clark Fork
Basin, Montana, Water Years 1996-2015

By Steven K. Sando and Aldo V. Vecchia

Abstract

During the extended history of mining in the upper Clark
Fork Basin in Montana, large amounts of waste materials
enriched with metallic contaminants (cadmium, copper, lead,
and zinc) and the metalloid trace element arsenic were gener-
ated from mining operations near Butte and milling and smelt-
ing operations near Anaconda. Extensive deposition of mining
wastes in the Silver Bow Creek and Clark Fork channels and
flood plains had substantial effects on water quality. Federal
Superfund remediation activities in the upper Clark Fork Basin
began in 1983 and have included substantial remediation near
Butte and removal of the former Milltown Dam near Mis-
soula. To aid in evaluating the effects of remediation activities
on water quality, the U.S. Geological Survey began collecting
streamflow and water-quality data in the upper Clark Fork
Basin in the 1980s.

Trend analysis was done on specific conductance,
selected trace elements (arsenic, copper, and zinc), and
suspended sediment for seven sampling sites in the Milltown
Reservoir/Clark Fork River Superfund Site for water years
1996-2015. The most upstream site included in trend analysis
is Silver Bow Creek at Warm Springs, Montana (sampling
site 8), and the most downstream site is Clark Fork above Mis-
soula, Montana (sampling site 22), which is just downstream
from the former Milltown Dam. Water year is the 12-month
period from October 1 through September 30 and is designated
by the year in which it ends. Trend analysis was done by using
a joint time-series model for concentration and streamflow. To
provide temporal resolution of changes in water quality, trend
analysis was conducted for four sequential 5-year periods:
period 1 (water years 1996-2000), period 2 (water years
2001-5), period 3 (water years 2006-10), and period 4 (water
years 2011-15). Because of the substantial effect of the inten-
tional breach of Milltown Dam on March 28, 2008, period 3
was subdivided into period 3 A (October 1, 2005-March 27,
2008) and period 3B (March 28, 2008 September 30. 2010)
for the Clark Fork above Missoula (sampling site 22). Trend

results were considered statistically significant when the statis-
tical probability level was less than 0.01.

In conjunction with the trend analysis, estimated normal-
ized constituent loads (hereinafter referred to as "loads") were
calculated and presented within the framework of a constitu-
ent-transport analysis to assess the temporal trends in flow-
adjusted concentrations (FACs) in the context of sources and
transport. The transport: analysis allows assessment of tem-
poral changes in relative contributions from upstream source
areas to loads transported past each reach outflow.

Trend results indicate that FACs of unfiltered-recoverable
copper decreased at the sampling sites from the start of
period 1 through the end of period 4; the decreases ranged
from large for one sampling site (Silver Bow Creek at Warm
Springs [sampling site 8]) to moderate for two sampling sites
(Clark Fork near Galen, Montana [sampling site 11] and Clark
Fork above Missoula [sampling site 22]) to small for four
sampling sites (Clark Fork at Deer Lodge, Montana [sampling
site 14], Clark Fork at Goldcreek, Montana [sampling site 16],
Clark Fork near Drummond, Montana [sampling site 18], and
Clark Fork at Turah Bridge near Bonner, Montana [sampling
site 20]). For period 4 (water years 2011-15), the most notable
changes indicated for the Milltown Reservoir/Clark Fork
River Superfund Site were statistically significant decreases in
FACs and loads of unfiltered-recoverable copper for sampling
sites 8 and 22. The period 4 changes in FACs of unfiltered-
recoverable copper for all other sampling sites were not statis-
tically significant.

Trend results indicate that FACs of unfiltered-recoverable
arsenic decreased at the sampling sites from period 1 through
period 4 (water years 1996-2015); the decreases ranged from
minor (sampling sites 8-20) to small (sampling site 22). For
period 4 (water years 2011-15), the most notable changes indi-
cated for the Milltown Reservoir/Clark Fork River Superfund
Site were statistically significant decreases in FACs and loads
of unfiltered-recoverable arsenic for sampling site 8 and near
statistically significant decreases for sampling site 22. The
period 4 changes in FACs of unfiltered-recoverable arsenic for
all other sampling sites were not statistically significant.


-------
2 Water-Quality Trends and Constituent-Transport Analysis for Selected Sampling Sites

Trend results indicate that FACs of suspended sediment
decreased at the sampling sites from period 1 through period 4
(water years 1996-2015); the decreases ranged from moderate
(sampling site 8) to small (sampling sites 11-22). For period 4
(water years 2011-15), the changes in FACs of suspended sed-
iment were not statistically significant for any sampling sites.

The reach of the Clark Fork from Galen to Deer Lodge
is a large source of metallic contaminants and suspended sedi-
ment, which strongly affects downstream transport of those
constituents. Mobilization of copper and suspended sediment
from flood-plain tailings and the streambed of the Clark Fork
and its tributaries within the reach results in a contribution
of those constituents that is proportionally much larger than
the contribution of streamflow from within the reach. Within
the reach from Galen to Deer Lodge, unfiltered-recoverable
copper loads increased by a factor of about 4 and suspended-
sediment loads increased by a factor of about 5, whereas
streamflow increased by a factor of slightly less than 2. For
period 4 (water years 2011-15), unfiltered-recoverable cop-
per and suspended-sediment loads sourced from within the
reach accounted for about 41 and 14 percent, respectively, of
the loads at Clark Fork above Missoula (sampling site 22).
whereas streamflow sourced from within the reach accounted
for about 4 percent of the streamflow at sampling site 22.
During water years 1996-2015, decreases in F ACs and loads
of unfiltered-recoverable copper and suspended sediment for
the reach generally were proportionally smaller than for most
other reaches.

Unfiltered-recoverable copper loads sourced within the
reaches of the Clark Fork between Deer Lodge and Turali
Bridge near Bonner (just upstream from the former Mill-
town Dam) were proportionally smaller than contributions
of streamflow sourced from within the reaches; these reaches
contributed proportionally much less to copper loading
in the Clark Fork than the reach between Galen and Deer
Lodge. Although substantial decreases in FACs and loads of
unfiltered-recoverable copper and suspended sediment were
indicated for Silver Bow Creek at Warm Springs (sampling
site 8), those substantial decreases were not translated to
downstream reaches between Deer Lodge and Turah Bridge
near Bonner. The effect of the reach of the Clark Fork from
Galen to Deer Lodge as a large source of copper and sus-
pended sediment, in combination with little temporal change
in those constituents for the reach, contributes to this pattern.

With the removal of the former Milltown Dam in
2008, substantial amounts of contaminated sediments that
remained in the Clark Fork channel and flood plain in reach 9
(downstream from Turah Bridge near Bonner) became more
available for mobilization and transport than before the dam
removal. After the removal of the former Milltown Dam. the
Clark Fork above Missoula (sampling site 22) had statistically
significant decreases in FACs of unfiltered-recoverable copper
in period 3B (March 28, 2008, through water year 2010) that
continued in period 4 (water years 2011-15). Also, decreases
in FACs of unfiltered-recoverable arsenic and suspended sedi-
ment were indicated for period 4 at this site. The decrease in

FACs of unfiltered-recoverable copper for sampling site 22
during period 4 was proportionally much larger than the
decrease for the Clark Fork at Turah Bridge near Bonner
(sampling site 20). Net mobilization of unfiltered-recoverable
copper and arsenic from sources within reach 9 are smaller for
period 4 than for period 1 when the former Milltown Dam was
in place, providing evidence that contaminant source materials
have been substantially reduced in reach 9.

Introduction

Mining in the upper Clark Fork Basin in Montana began
in 1864 when small-scale placer mining operations extracted
gold from Silver Bow Creek and its tributaries in and near
Butte (Freeman, 1900; U.S. Environmental Protection Agency,
2005; fig. 1). By the early 1900s, the small gold mining opera-
tions had transitioned to larger scale underground silver and
copper mining owned by the former Anaconda Mining Com-
pany (AMC). with most of the ore being processed at AMC
milling and smelting facilities near Anaconda (U.S. Environ-
mental Protection Agency, 2005, 2010; Gammons and others,
2006). In 1955, the AMC mining operations began to transi-
tion from underground to open-pit mining, with the opening
of the Berkeley Pit north of Butte. The Berkeley Pit mining
operations and AMC milling and smelting operations contin-
ued until closure in the early 1980s.

During the extended history of mining in the upper Clark
Fork Basin, large amounts of waste materials enriched with
metallic contaminants (cadmium, copper, lead, and zinc)
and the metalloid trace element arsenic were generated from
mining operations near Butte and the milling and smelting
operations near Anaconda (Andrews, 1987; Gammons and
others, 2006). Extensive deposition of mining wastes in the
Silver Bow Creek and Clark Fork channels and flood plains
had substantial effects on water quality. Federal Superfund
remediation activities in the upper Clark Fork Basin began in
1983 and have included substantial remediation near Butte and
removal of the former Milltown Dam near Missoula in 2008
(U.S. Environmental Protection Agency, 2004, 2010; CDM,
2005; Sando and Lambing, 2011). The various Superfund
activities are distributed among three National Priorities List
sites: the Silver Bow Creek/Butte Area Site, the Anaconda
Smelter Site, and the Milltown Reservoir/Clark Fork River
Superfund Site, which are described in the "Description of
Study Area" section of this report.

Water-quality data collection by the U.S. Geological
Survey (USGS) in the upper Clark Fork Basin began during
1985-88 with the establishment of a small long-term monitor-
ing program that has expanded through time and continued
through present (2016). Sando and others (2014) analyzed
the monitoring data and characterized flow-adjusted trends in
mining-relaled contaminants for 22 sampling sites in the Silver
Bow Creek/Butte Area Site, the Anaconda Smelter Site, and
the Milltown Reservoir/Clark Fork River Superfund Site in the


-------
Introduction 3

Former 1
A Milltown
Dam

-Aackfoot4^

Bonner

issoul

Clinton

Clark pt

Reach 7	/

V ^ \	'"5

Drummond

Reach 8

|itife *4

Goldcreel

Garrison

Reach 6

Reach

Butte

Bow Creek

!acetrack

Galen

iportunity

113°30'









\









i V

V-



4

%

WJ?>



liwsEfea

/ <



V '

sH



\

\ J

% J





V\

m-







\

Pv
{ »

|' iWi

46°:30l

IDAHO

/¦t

i

i

j

¦

¦ j ¦jrjLt

J

Mjr1 ajP . j
*

r ¦

[I iBMBfff

f

v m

K5jM£*

) J

r

tWr

L/

jr

tf, ¦

_

.

Base modified from US, farm Services .Agency
National Agricultural Imagery Program (NAIP), 2009
U.S. Census Bureau TIGER, 1:100,000,1993
Univfersal Transverse Mercator projection,zone 12
North American Datum of 1927 (NAD 27)

|	10	15.	20 MILES

i—S	r1	1	1	1

5 10 15 20 KILOMETERS

EXPLANATION

Study area

	 Data-summary reach basin boundary

Main-stem Silver Bow Creek or Clark Fork channel
Sampling site and sampling site number (table 1)

16^-v

\W Sampling site included in trend analysis

Sampling site not included in trend analysis

Figure 1. Location of study area, selected sampling sites, and data-summary reaches in the upper Clark Fork Basin, Montana; the Milltown Reservoir/Clark Fork River Superfund Site
includes the reaches from sampling site 8 to sampling site 22.


-------
4 Water-Quality Trends and Constituent-Transport Analysis for Selected Sampling Sites

upper Clark Fork Basin for water years 1996-2010 (water year
is the 12-month period from October 1 through September 30
and is designated by the year in which it ends). An update of
flow-adjusted water-quality trends for the monitoring data was
needed for seven sampling sites to provide timely information
for the 2016 5-vear review for the Milltown Reservoir/Clark
Fork River Superfund Site. The USGS, in cooperation with
the U.S. Environmental Protection Agency, conducted this
study to test for flow-adjusted trends (water years 1996-2015)
in water quality at seven sampling sites (fig. 1, table 1) in the
Milltown Reservoir/Clark Fork River Superfund Site by using
a joint time-series model (TSM; Vecchia, 2005) for concentra-
tion and streamflow; an eighth site (Clark Fork above Little
Blackfoot River near Garrison. Montana [sampling site 15;
fig. 1. table 1]) was included in the study for the purpose of
statistically summarizing water-quality data collected during
water years 2011-15, but the period of water-quality data col-
lection was insufficient for trend analysis.

Purpose and Scope

The primary purposes of this report arc to (1) character-
ize temporal trends in flow-adjusted concentrations (filtered
and unfiltered) of mining-related contaminants and (2) assess
those trends in the context of source areas and transport of
those contaminants through the Milltown Reservoir/Clark
Fork River Superfund Site in the upper Clark Fork Basin.
Trend analysis was done on specific conductance, selected
trace elements (arsenic, copper, and zinc), and suspended sedi-
ment for seven sampling sites for water years 1996-2015. This
report provides an update of and supersedes the trend results
reported by Sando and others (2014) for seven sampling sites
in the Milltown Reservoir/Clark Fork River Superfund Site.
This report presents the trend results and information on
trend-analysis methods, streamflow conditions, and various
data-related factors that affect trend results. This information
is presented to assist in evaluating trend results; however, it is
beyond the scope of this report to provide detailed explana-
tions for all observed temporal changes.

Description of Study Area

The Clark Fork drains an extensive region in western
Montana and northern Idaho in the Columbia River Basin (not
shown on fig. 1). The main-stem Clark Fork begins at the con-
fluence of Silver Bow and Warm Springs Creeks near Warm
Springs, Montana, and flows about 485 miles (mi) through
Montana and Idaho. The study area (fig. 1) encompasses the
upper Clark Fork Basin in west-central Montana upstream
from Clark Fork above Missoula. Montana (sampling site 22,
table 1), with a drainage area of 5,999 square miles (mi2).
Sando and others (2014) presented somewhat detailed infor-
mation describing the hydrographic. physiographic, climatic,
and geologic characteristics of the upper Clark Fork Basin and
an overview of mining and remediation activities.

Early Federal Superfund activities in the upper Clark
Fork Basin involved designation of three areas as National
Priorities List sites in 1983: the Silver Bow Creek Site, the
Anaconda Smelter Site, and the Milltown Reservoir Site. The
Silver Bow Creek Site was redesignated as the Silver Bow
Creek/Butte Area Site in 1987 and includes remnants from
mining operations near Butte and about 26 river miles of
Silver Bow Creek extending from near Butte to the outlet of
Warm Springs Ponds (U.S. Environmental Protection Agency,
2000; CDM, 2005). The Anaconda Smelter Site includes about
300 mi2, primarily in the Mill, Willow, Warm Springs, and
Lost Creek drainage basins near Anaconda (U.S. Environ-
mental Protection Agency, 2010). Many remediation activities
within the Anaconda Smelter Site are administered within the
Regional Water, Waste, and Soils Operable Unit (Henry Elsen,
U.S. Environmental Protection Agency, written commun.,
January 2016). The Milltown Reservoir Site was redesignated
as the Milltown Reservoir/Clark Fork River Superfund Site
in 1992. The Milltown Reservoir/Clark Fork River Superfund
Site includes two primary operable units: the Milltown Res-
ervoir Operable Unit and the Clark Fork Operable Unit. The
Milltown Reservoir Operable Unit includes about 0.84 mi2
defined by the area inundated by maximum pool elevation of
the former Milltown Reservoir (U.S. Environmental Protec-
tion Agency, 2004). The Clark Fork Operable Unit includes
streamside areas of the 115-mi reach of the Clark Fork
extending from the Warm Springs Ponds outlet to the start of
Milltown Reservoir Operable Unit (Montana Department of
Environmental Quality, 2016).

The specific focus of this study is the Milltown Reser-
voir/Clark Fork River Superfund Site, which includes the
Clark Fork Operable Unit and the Milltown Reservoir Oper-
able Unit, and extends about 123 river miles from the outlet
of Warm Springs Ponds on Silver Bow Creek (represented by
sampling site 8) to the outlet of the former Milltown Reservoir
(represented by sampling site 22, which is about 3 river miles
downstream from the former Milltown Dam). Sampling sites
included in this study are located on the main-stem channels
of Silver Bow Creek and the Clark Fork. Sando and others
(2014) included trend analyses for several sampling sites on
tributaries to Silver Bow Creek or the Clark Fork in the Mill-
town Reservoir Clark Fork River Superfund Site; however,
data collection for most of the tributary sampling sites was dis-
continued in water year 2004. No tributary sampling sites were
included in this study. The sampling site numbers and reach
designations assigned by Sando and others (2014) generally
have been retained to facilitate comparisons. An exception is
Clark Fork above Little Blackfoot River near Garrison (USGS
streamgage 12324400), for which data collection began in
water year 2009. Streamgage 12324400 was not included in
Sando and others (2014). A discontinued tributary sampling
site (Little Blackfoot River near Garrison, Montana; USGS
streamgage 12324590) was designated as sampling site 15 in
Sando and others (2014), but in this study Clark Fork above
Little Blackfoot River near Garrison (USGS streamgage
12324400) is designated as sampling site 15. The period of


-------
Table 1. Information for selected sampling sites and data-summary reaches in the Milltown Reservoir/Clark Fork River Superfund Site in the upper Clark Fork Basin, Montana.

[Water year is the 12-month period from October 1 through September 30 and is designated by the year in which it ends. USGS, U.S. Geological Survey; NA, not applicable]

Sam-
pling site
number1

(fig. D

USGS site
identification
number

USGS site name

Abbreviated sampling site name

Data

summary
reach12

Drainage

area,
in square
miles

Period of
water-quality
data collection

Median annual
sampling frequency,
in samples per year
(range)

Trend
analysis
periods3

11

14

15

16
18

20

22

12323750

12323800
12324200

12324400

12324680

12331800

12334550

Silver Bow Creek at Warm Springs,

Montana

Clark Fork near Galen, Montana	Clark Fork near Galen

Clark Fork at Deer Lodge, Montana	Clark Fork at Deer Lodge

Clark Fork above Little Blackfoot	„ ,

_.	_	- .	Clark F ork near Garrison

River near Garrison, Montana

Clark Fork at Goldcreek, Montana	Clark Fork at Goldcreek

Silver Bow Creek at Warm Springs 3 and 4 473

4	and 5

5	and 6

Clark Fork near Dmmmond, Montana Clark Fork near Drammond

Clark Fork at Turah Bridge near
Bonner. Montana

Clark Fork at Turah Bridge

12340500 Clark Fork above Missoula, Montana Clark Fork above Missoula

651
995

6	1,139

6	and 7	1,704

7	and 8	2,501

8	and 9	3,641

3/1993-8/2015

7/1988-8/2015
3/1985-8/2015

3/2009-8/2015

3/1993-8/2015

3/1993-8/2015

3/1985-8/2015

5.999 7/1986-8/2015

8(6-11)

8(1-13)

8(4-20)

8 (7-8)
8(6-10)
8(6-10)

8(6-23)

8(2-18)

1,2, 3,4

1,2, 3,4
1,2, 3,4

NA"'

1,2,3,4

1,2, 3,4

1,2, 3,4

1,2, 3A,
3B, 4

'For this study, the sampling site numbers and reach designations assigned by Sando and others (2014) generally have been retained to facilitate comparisons.

2Where two reach numbers are shown, the site is both an outflow from the upstream reach and an inflow to the downstream reach.

The numerical designations of the trend analysis periods are defined as
water years 1996-2000;
water years 2001-5;
water years 2006-10;
water years 2011	15.

Because of the substantial effect of the breach and removal of Milltown Dam in 2008, for Clark Fork above Missoula (station 12340500), period 3 was subdivided into period 3A (October 1, 2005	March 27,

analysis . Site was included in the study for the purpose of statistically summarizing water-quality data collected during water years 2011	15.


-------
6 Water-Quality Trends and Constituent-Transport Analysis for Selected Sampling Sites

water-quality data collection is insufficient for trend analysis
for sampling site 15, but this site was included in the study
for the purpose of statistically summarizing water-quality data
collected during water years 2011-15.

Data-Collection and Analytical
Methods

Sando and others (2014) present information concerning
historical aspects of data-collection and analytical methods
used in the monitoring program. Data collected in the monitor-
ing program are published (typically on an annual basis) in
data reports that present the methods of data collection, water-
quality' data, quality-assurance data, and statistical summaries
of the data (for example. Dodge and others, 2015). A brief
overview of field and laboratory data-collection and analytical
methods is presented in the following paragraphs.

The sampling design of the monitoring program provides
information relevant to several objectives, including evaluat-
ing constituent transport, regulatory compliance, and long-
term trends. Since 1993, the sampling frequency of the main-
stem sampling sites in the monitoring program generally has
been consistent, with the sites sampled eight times per year in
most years, in the monitoring program, the seasonal timing of
sample collection placed greater emphasis on the snowmelt
runoff period (typically April-July), when streamflow condi-
tions are high and variable and constituent transport is large.
About 75 percent of samples were collected during April-July.
In general, the frequency and timing of sample collection
throughout the period of data collection among the sites are
reasonably consistent to provide reasonable consistency in
trend-analysis results.

In the monitoring program, water samples were collected
from vertical transits throughout the entire stream depth at
multiple locations across the stream by using standard USGS
depth- and width-integration methods (U.S. Geological Sur-
vey, variously dated). Those methods provide a vertically and
laterally discharge-weighted composite sample that is intended
to be representative of the entire flow passing through the
cross section of a stream (Dodge and others, 2015). Specific
conductance was measured onsite in subsamples from the
composite water samples. Subsamples of the composite water
samples were analyzed at the USGS National Water Qual-
ity- Laboratory (NWQL) in Denver, Colorado, for filtered
(0.45-micrometerpore size) and unfiltered-recoverable
concentrations of the trace-element constituents (table 2) by
using methods described by Garbarino and Struzeski (1998)
and Garbarino and others (2006). Water samples also were
analyzed for suspended-sediment concentrations by the USGS
sediment laboratory in Helena, Montana. All water-quality'
data are available in the USGS National Water Information
System (NWIS; U.S. Geological Survey, 2015).

Quality Assurance

Sando and others (2014) present information concerning
historical aspects of quality-assurance procedures used in the
monitoring program. Quality-assurance data collected in the
monitoring program are reported and statistically summarized
in annual data reports (for example, Dodge and others, 2015).
Selected quality-assurance information relevant to this study is
presented in the following paragraphs.

Analytical results for field quality-assurance samples
(including field blank and replicate samples) that were
collected in the monitoring program during water years
1993-2015 were compiled and statistically summarized
(table 1-1 in appendix 1 at the back of the report). Those data
provide information on the consistency and environmental
representativeness of data collection. Representative sampling
for trace elements in streams is particularly difficult because of
low concentrations in stream waters and ubiquitous presence
in the sampling environment that produce an associated large
potential for contamination.

Summary of analytical results for field blank samples
(table 1-1 in appendix 1 at the back of the report) provides
information on potential effects of contamination during the
sampling process on trend-analysis results. For the trace-
element constituents included in the trend analysis (table 2),
the frequency of detection in field blank samples at concentra-
tions greater than the laboratory reporting level (LRL) at the
time of analysis ranged from 0.5 percent (filtered arsenic) to
10.7 percent (unfiltered-recoverable zinc). Precise statisti-
cal analysis of the analytical results of field blank samples is
difficult because of the multiple LRLs used by NWQL during
the study period (table 2). Also, it is difficult to precisely
quantify- the field blank sample results with respect to the
study datasets because contamination indicated by field blank
samples was routinely monitored in the Clark Fork monitor-
ing program, and stream-sample data judged to be affected by
persistent contamination issues were identified during periodic
reviews of the data and excluded from data analysis. However,
it is important that trend-analysis procedures arc structured
to minimize potential effects of sampling contamination on
low-concentration data included in the trend analysis. Specific
procedures used in application of the trend-analysis method
with respect to handling of low-concentration and censored
data (that is, analytical results reported as less than the LRL;
Helsel, 2005) arc described in the section of this report "Gen-
eral Description of the Time-Series Model."

Summary of analytical results for field replicate samples
(table 1-1 in appendix 1 at the back of the report) provides
information on data precision. For the entire study period, the
relative standard deviations (a measure of overall precision)
for field replicate sample pairs were within 20 percent for all
constituents, indicating reasonable precision (Taylor, 1987;
Dodge and others, 2015).


-------
Quality Assurance 7

Table 2. Properties, constituents, and associated information relating to laboratory and study reporting levels.

[Water year is the 12-month period from October 1 through September 30 and is designated by the year in which it ends. NWQL, U.S. Geological Survey
National Water Quality Laboratory; uS cm. microsiemen per centimeter at 25 degrees Celsius; NA, not applicable; mg/L, milligram per liter; jig L. microgram
per liter]

Property or constituent

Units of
measurement

Number of NWQL
laboratory reporting
levels during water years
1993-2015

Range in NWQL
laboratory reporting
levels

Study reporting level
used in application of the
time-series model1

Specific conductance2

(.iS/cm

NA

NA

NA

pH, standard units

standard units

NA

NA

NA

Calcium, filtered

rag/L

5

0.005-0.022

NA

Magnesium, filtered

mg/L

7

0.002-0.011

NA

Cadmium, filtered

Hg/L

7

0.01-1.0

NA

Cadmium, unfi Itered-reco verable

Hg/L

10

0.007-1.0

NA

Copper, filtered2

Hg/L

4

0.2-1

1.0

Copper, unfiltered-reeoverable2

TO/I-

6

0.3-2

1.0

Lead, filtered

Hg/L

10

0.015-5

NA

Lead, unfiltered-reeoverable

Hg/L

6

0.03-5

NA

Zinc, filtered

Hg/L

7

0.9-20

NA

Zinc, unfiltered-reeoverable2

Hg/L

4

2-31

2.0

Arsenic, filtered2

Hg/L

7

0.022-1

1.0

Arsenic, unfiltered-reeoverable2

Hg/L

7

0.06-1

1.0

Suspended sediment2

mg/L

NA

NA

1

Procedures for determining and applying the study reporting level used in the application of the time-series model are discussed in the section of this report
"General Description of the Time-Series Model."

2Propertv or constituent was analyzed for temporal trends.

Analytical results for laboratory-spiked deioiiized-water
blank samples and stream-water samples that were collected
in the monitoring program during water years 1993-2015
are presented in tables 1-2 and 1-3. respectively, in appen-
dix 1 at the back of the report. Annual mean recoveries for
laboratory-spiked deionized-water blank samples for all
constituents combined have ranged from 82.3 to 118 percent
(mean of 104 percent). Annual mean recoveries for laboratory -
spiked stream-water samples for all constituents combined
have ranged from 84.3 to 114 percent (mean of 105 percent).
Potential effects of temporal variability in spike recoveries on
trend results are described in appendix 1 and also the section
"Specific Aspects of the Application of the Time-Series Model
in this Study" in appendix 2. Based on analysis of all quality-
assurance data, the quality of the study datasets were deter-
mined to be suitable for trend analysis.


-------
8 Water-Quality Trends and Constituent-Transport Analysis for Selected Sampling Sites

Overview of Streamflow and Water-
Quality Characteristics for Water
Years 2011-15

Statistically summarizing recent streamflow and water-
quality characteristics of the study sampling sites (fig. 1,
table 1) is useful for generally describing water quality and in
providing comparative information relevant for interpreting
trend results. Data are summarized for water years 2011-15,
a summary period that represents recent water-quality condi-
tions and the increment of data collected after the study period
1996-2010 reported by Sando and others (2014).

General Streamflow Characteristics for Water
Years 2011-15

To aid in interpreting water-quality characteristics of the
sampling sites, statistical summaries of continuous streamflow
data are presented in table 3. The continuous streamflow data
are available inNWIS (U.S. Geological Survey, 2015). In
general, streamflow conditions during water years 2011-15
were somewhat high. Mean annual streamflows for water
years 2011-15 generally were about 10-20 percent higher than
period-of-record mean annual streamflows.

Water-Quality Characteristics for Water
Years 2011-15

Statistical summaries of water-quality- data (water years
2011-15) for sampling sites in the Milltown Reserv oir/

Clark Fork River Superfund Site in the upper Clark Fork
Basin are presented in table 4. The statistical summaries in
table 4 are based on unadjusted trace-element concentrations
(the observed concentrations before flow adjustment). Flow-
adjustment, described in the sections of this report "General
Description of the Time-Series Model" and "Factors that
Affect Trend Results and Interpretation." is relevant when
interpreting trends in concentrations of water-quality* constitu-
ents that are strongly dependent on streamflow conditions.
However, flow adjustment is not relevant for statistically sum-
marizing the observed water-quality data during water years
2011-15.

In addition to statistical summaries of unadjusted con-
centrations, ratios of median filtered to unfiltered-recoverable
trace-element concentrations arc reported in table 4 to pro-
vide general information on the predominant phase (that is,
dissolved or particulate) of transport. Values of aquatic-life
standards (Montana Department of Environmental Quality,
2012; based on median hardness for each site for water years
2011-15) for cadmium, copper, lead, and zinc are presented
in table 1-4 in appendix 1 at the back of the report; those
values were used for plotting the standards in relation to
statistical distributions of selected trace elements. The arsenic

human-health standard is 10 micrograms per liter (|ig/L;
Montana Department of Environmental Quality, 2012). Per-
centages of samples (water years 2011-15) with unadjusted
unfiltered-recoverable concentrations exceeding water-quality
standards for each site are presented in table 5. The exceed-
ance percentages for the hardness-based aquatic-life standards
for cadmium, copper, lead, and zinc in table 5 were based on
comparison of trace-element concentrations of each individual
sample with the aquatic-life standards that were calculated by
using the hardness for each individual sample.

Statistical distributions of water-quality characteristics of
the sampling sites are illustrated in figure 2 by using boxplots
of selected example constituents (unadjusted specific conduc-
tance and unadjusted concentrations of copper, arsenic, and
suspended sediment); the boxplots provide an overview of
important water-quality characteristics in the upper Clark Fork
Basin. Also shown in figure 2 are applicable water-quality
standards. Specific conductance is presented as an example
because it is an index of ionic strength, is strongly correlated
with hardness (which is used in calculations of aquatic-life
standards), and provides information on the extent of water
contact with geologic materials, types of geologic materials
present in the sampling-site basins, and potential effects of
remediation activities on ionic strength. Copper and arsenic
are presented as examples of trace elements because they
arc constituents of concern with respect to potential toxicity
issues, but they have much different geochemical characteris-
tics. Spatial and temporal variability in copper concentrations
in the upper Clark Fork Basin generally is similar to vari-
ability in other metallic contaminants that tend to adsorb to
particulates in water (Sando and others, 2014) and is consid-
ered generally representative of those constituents. In contrast,
arsenic in the upper Clark Fork Basin tends to largely exist in
the dissolved phase and does not exhibit the same variability
as metallic contaminants (Sando and others, 2014). Suspended
sediment is presented because it provides information on
transport of particulate materials, which is a factor that can
strongly affect transport of metallic contaminants.

To assist in the presentation of results, Sando and others
(2014) divided Silver Bow Creek and the Clark Fork into
nine data-summary reaches based on the location of sampling
sites along the main-stems of those streams. The sampling site
numbers and reach designations assigned by Sando and others
(2014) generally have been retained to facilitate comparisons,
and water-quality characteristics for sampling sites in six
reaches (reaches 4-9) are presented. Water-quality charac-
teristics within the six reaches are affected by environmental
characteristics w ithin the delineated reach basin boundaries
(fig. 1). Water-quality characteristics of the sampling sites are
described for each of the data-summary reaches. Emphasis
is placed on describing spatial differences in observed water
quality in the Milltown Reservoir/Clark Fork River Super-
fund Site in the upper Clark Fork Basin during water years
2011-15.


-------
Table 3. Statistical summaries of continuous streamflow data for selected sampling sites in the Milltown Reservoir/Clark Fork River Superfund Site in the upper Clark Fork
Basin, Montana.

[Water year is the 12-month period from October 1 through September 30 and is designated by the year in which it ends. ftVs, cubic foot per second; POR, period of record]

Statistical summaries of daily mean streamflow,

in ft3/s

Drainage

Sampling site number Abbreviated sampling site name area,
(fig. 1, table 1)	(table 1)	in square

miles

Analysis period.







Mean
(also referred to
as "mean annual
streamflow")





in water years
(number of years)

Minimum

25th
percentile

Median

75th
percentile

Maximum

....... ^

0

1

Ni

22
15

41



96
88

97
88

1 060
1.060

2011-15(5)

35

92

i 30

172

174

1,390

POR: 1989-2015 (27)

13

70

100

143

152.

1.390

2011-15(5)

55

187

237

283

302

1,960

POR: 1979-2015 (37)

22

159

219

257

298

2,390

2011-15(5)

61

198

263

3 i 5

331

2.560

POR: 2010-15 (6)

61

2.09

2.67

32.3

334

2.560

2011-15(5)

1 12

320

409

570

583

6.100

OO
00

¦s-s

280

380

519

556

9.100

2011-15(5)

185

461

595

771

813

7.740

POR: 1994-2015 (2.2)

	

419

563

718

758

8.430

2011-15(5)

250

790

990

1,490

1,560

12,700

POR: 1985-2015 (31)

177

678

870

1,260

1,260

12,700

2011-15(5)

500

1,400

1.730

3.330

3.760

28.100

POR: 1930-2015 (86)

340

1,270

1,650

2.930

2.960

30.800

14

15

16

20

n

Silver Bow (.'reek al Warm Sprirms 47/

Clark Fork near Galen

Clark Fork at Deer Lodge
Clark I 'ork near Garrison
Clark l ork aS Gokk:reek
Clark I 'ork near Drninniond

Clark Fork at Turah Bridge
Clark I 'ork above Missoula

651

995

1.139

.704

2.501

3,641

5.999


-------
Table 4. Statistical summaries of water-quality data collected at selected sampling sites in the Milltown Reservoir/Clark Fork River Superfund Site in the upper Clark Fork Basin,
Montana, water years 2011-15.

I Water year is the 12-month period from October 1 through September 30 and is designated by the year in which it ends. ftVs, cubic foot per second; NA, not applicable; [iS cm. microsiemen per centimeter at
25 degrees Celsius; CaCOv calcium carbonate; jug/L, microgram per liter; mg/L, milligram per liter]

Statistical summaries of water-quality data1

Constituent or property,
unadjusted
(not flow adjusted)
units of measurement

Number of samples
(values in parentheses
indicate number of
censored values)

Minimum
uncensored
value2

25th

percentile

Median

Mean

75th
percentile

Maximum

filtered to median
unfiltered-recoverable
concentrations for
trace elements,
in percent3



Silver Bow Creek at Warm Springs, Montana

(sampling site E

i, fig. 1,table 1)







Streamflow, instantaneous, ft3/s

40

20

66

89

146

161

1,030

NA

Specific conductance, (iS/cm

40

182

342

394

407

489

577

NA

pH, standard units

40

8.1

8.5

8.8

NA

9.1

9.4

NA

Hardness, filtered, mg/L as CaCO,

40

74.9

136

170

169

203

253

NA

Calcium, filtered, mg/L

40

22.5

39.7

48.4

48.7

58.6

73.3

NA

Magnesium, filtered, mg/L

40

4.52

9.10

11.8

11.5

14.4

16.9

NA

Cadmium, filtered, ug/L

40 (4)

0.023

0.031

0.038

0.044

0.054

0.096

45

Cadmium, unfiltered-recoverable, (ig/L

40

0.027

0.065

0.085

0.119

0.125

0.567



Copper, filtered, (.ig/L

40

1.6

2.6

3.5

4.3

4.7

21.4

51

Copper, unfiltered-recoverable, ug/L

40

2.8

5.0

6.8

9.5

11.2

35.2



Iron, filtered, ug/L

40

7.0

16.2

30.0

30.0

38.7

63.0

13

Iron, unfiltered-recoverable, ug/L

40

61.1

159

225

256

313

839



Lead, filtered, (.ig/L

40

0.044

0.103

0.158

0.162

0.186

0.566

14

Lead, unfiltered-recoverable, (ig/L

40

0.37

0.81

1.16

1.80

2.07

6.39



Manganese, filtered, |ig/L

40

27.1

42.7

61.2

72.6

84.7

208

64

Manganese, unfiltered-recoverable, ug/L

40

60.1

77.5

95.2

116

130

332



Zinc, filtered, ug/L

40(11)

1.5

1.7

2.8

2.8

3.3

6.1

33

Zinc, unfiltered-recoverable, ug/L

40 (2)

2.3

5.5

8.6

13.3

14.1

69.8



Arsenic, filtered, (ig/L

40

8.4

13.4

19.2

20.9

28.0

38.1

86

Arsenic, unfiltered-recoverable, ug/L

40

10.4

16.9

22.4

22.8

28.8

37.9



Suspended sediment, mg/L

40

1

3

6

6

7

21

NA

Suspended sediment, percent fines4

40

60

84

88

87

92

98

NA


-------
Table 4. Statistical summaries of water-quality data collected at selected sampling sites in the Milltown Reservoir/Clark Fork River Superfund Site in the upper Clark Fork Basin,
Montana, water years 2011-15.—Continued

[Water year is the 12-month period from October 1 through September 30 and is designated by the year in which it ends, ftYs, cubic foot per second; NA, not applicable; j.iS cm. microsiemen per centimeter at
25 degrees Celsius; CaCOv calcium carbonate; j.ig L. microgram per liter; mg/L, milligram per liter]

Statistical summaries of water-quality data'

Constituent or property,
unadjusted

(not flow adjusted)
units of measurement

Number of samples
(values in parentheses
indicate number of
censored values)

Minimum
uncensored
value2

25th
percentile

Median

Mean

75th
percentile

Maximum

filtered to median
unfiltered-recoverable
concentrations for
trace elements,
in percent3



Clark Fork near Galen, Montana (sam

pling site 11, fig. 1, table 1)







Streamflow, instantaneous, ft3/s

40

38

110

175

249

284

1,380

NA

Specific conductance, (.iS/cm

40

182

292

367

360

434

498

NA

pH, standard units

40

8.2

8.4

8.6

NA

8.7

9.1

NA

Hardness, filtered, mg/L as CaCO,

40

76.4

125

164

158

191

225

NA

Calcium, filtered, mg/L

40

23.2

37.1

47.9

46.4

55.4

65.1

NA

Magnesium, filtered, mg/L

40

4.44

7.75

10.6

10.2

12.7

15.1

NA

Cadmium, filtered, (ig/L

40(2)

0.020

0.037

0.041

0.044

0.049

0.111

42

Cadmium, unfiltered-recoverable, |ig/L

40

0.034

0.076

0.098

0.115

0.160

0.287



Copper, filtered, tig/L

40

1.4

3.1

3.7

4.3

4.7

19.8

31

Copper, unfiltered-recoverable, ug/L

39

4.8

9.2

11.9

15.4

17.5

51.6



Iron, filtered, |ig/L

40

7.5

11.7

20.0

20.2

27.1

43.0

8

Iron, unfiltered-recoverable, jig/L

40

67.5

167

248

297

370

860



Lead, filtered, (ig/L

40

0.037

0.074

0.112

0.116

0.132

0.387

7

Lead, unfiltered-recoverable, (.ig/L

40

0.40

1.10

1.51

2.06

2.82

6.33



Manganese, filtered, ug/L

40

13.1

37.8

41.8

54.7

63.8

130

48

Manganese, unfiltered-recoverable, ug/L

40

40.9

73.0

87.5

102

122

220



Zinc, filtered, ug/L

40 (7)

1.4

1.8

2.6

2.8

3.3

9.4

24

Zinc, unfiltered-recoverable, (ig/L

40

2.8

7.1

10.7

13.5

18.0

45.1



Arsenic, filtered, ug/L

40

7.0

10.4

12.7

13.8

18.0

27.5

82

Arsenic, unfiltered-recoverable, ug/L

40

8.9

12.4

15.4

16.0

19.0

31.5



Suspended sediment, mg/L

40

2

5

8

12

12

59

NA

Suspended sediment, percent fines4

40

32

68

76

75

87

96

NA


-------
Table 4. Statistical summaries of water-quality data collected at selected sampling sites in the Milltown Reservoir/Clark Fork River Superfund Site in the upper Clark Fork Basin,
Montana, water years 2011-15.—Continued

[Water year is the 12-month period from October 1 through September 30 and is designated by the year in which it ends. ftYs, cubic foot per second; NA, not applicable; uS/cm, microsiemen per centimeter at
25 degrees Celsius; CaC03, calcium carbonate; ug/L, microgram per liter; mg/L, milligram per liter]

Statistical summaries of water-quality data1

Constituent or property,
unadjusted
(not flow adjusted)
units of measurement

Number of samples
(values in parentheses
indicate number of
censored values)

Minimum
unceiisored
value2

25th
percentile

Median

Mean

75th
percentile

Maximum

filtered to median
unfiltered-recoverable
concentrations for
trace elements,
in percent3



Clark Fork at Deer Lodge,

Montana (sam

pling site 14, fig. 1, table 1)







Strcamflow, instantaneous, ft3/s

40

44

197

265

353

357

2,000

NA

Specific conductance, (.iS/cm

40

228

346

436

412

481

525

NA

pH, standard units

40

7.9

8.2

8.3

NA

8.4

8.9

NA

Hardness, filtered, mg/L as CaC03

40

97.1

154

200

183

214

231

NA

Calcium, filtered, mg/L

40

29.1

46.0

58.8

54.0

62.8

68.8

NA

Magnesium, filtered, mg/L

40

5.92

9.56

13.1

11.8

13.7

15.5

NA

Cadmium, filtered, |ig/L

40

0.035

0.049

0.065

0.069

0.072

0.280

43

Cadmium, unfiltered-recoverable, |ig/L

40

0.046

0.094

0.152

0.203

0.221

0.784



Copper, filtered, ug/L

40

3.4

5.6

7.0

8.3

7.7

45.9

25

Copper, unfiltered-recoverable, ug/L

40

9.4

15.2

27.6

46.3

49.3

220



Iron, filtered, |ig/L

40

5.5

11.7

18.5

18.7

24.9

45.8

4

Iron, unfiltered-recoverable, ,ug/L

40

63.0

224

436

708

788

4,290



Lead, filtered, ug/L

40(1)

0.041

0.082

0.142

0.152

0.189

0.372

4

Lead, unfiltered-recoverable, ug/L

40

0.55

1.61

3.28

5.70

6.63

32.8



Manganese, filtered, ug/L

40

11.7

22.6

30.0

32.6

38.7

70.8

36

Manganese, unfiltered-recoverable, (ig/L

40

22.9

57.4

82.9

97.5

115

364



Zinc, filtered, ug/L

40

1.6

3.6

5.5

6.5

6.6

50.6

23

Zinc, unfiltered-recoverable, jig/L

40

5.0

15.3

23.2

34.9

37.6

164



Arsenic, filtered, fig/L

40

in

10.3

13.3

14.0

16.2

36.6

81

Arsenic, unfiltered-recoverable, |ig/L

40

9.7

13.8

16.4

19.2

20.3

46.6



Suspended sediment, mg/L

40

2

8

17

33

31

218

NA

Suspended sediment, percent fines4

40

39

72

81

77

86

96

NA


-------
Table 4. Statistical summaries of water-quality data collected at selected sampling sites in the Milltown Reservoir/Clark Fork River Superfund Site in the upper Clark Fork Basin,
Montana, water years 2011-15.—Continued

[Water year is the 12-month period from October 1 through September 30 and is designated by the year in which it ends, ftYs, cubic foot per second; NA, not applicable; j.iS cm. microsiemen per centimeter at
25 degrees Celsius; CaCOv calcium carbonate; j.ig L. microgram per liter; mg/L, milligram per liter]

Statistical summaries of water-quality data1 n . .
	 1	 Ratios of median

Constituent or property,
unadjusted
(not flow adjusted)
units of measurement

Number of samples
(values in parentheses
indicate number of
censored values)

Minimum
uncensored
value2

25th
percentile

Median

Mean

75th
percentile

Maximum

filtered to median
unfiltered-recoverable
concentrations for
trace elements,
in percent3





Clark Fork above Little Blackfoot River near Garrison,

Montana (sam

pling site 15, fig. 1, table 1)





CD

Streamflow, instantaneous, ft3/s

39

71

227

289

410

418

2,310

NA

a

5*
§
o

•Hn

Specific conductance, (.iS/cm

39

249

363

449

421

479

527

NA

pH, standard units

39

7.9

8.2

8.4

NA

8.6

8.9

NA

GO

=T

Hardness, filtered, mg/L as CaCO,

39

107

162

202

186

213

228

NA

CP
m

1
©
<

Calcium, filtered, mg/L

39

31.9

47.4

58.8

54.1

61.7

66.5

NA

Magnesium, filtered, mg/L

39

6.65

10.4

13.4

12.3

14.4

15.5

NA

<
m
3

Cadmium, filtered, jig/L

39(1)

0.024

0.050

0.065

0.067

0.072

0.227

42

a.

s

w

s

Cadmium, unfiltered-recoverable, |ig/L

39

0.027

0.117

0.155

0.227

0.272

0.835



Copper, filtered, (.ig/L

39

2.8

6.2

7.9

9.2

9.7

40.6

25

i

Copper, unfiltered-recoverable, ug/L

39

10.0

19.1

31.9

51.3

54.0

222



m

3

o

T°

Iron, filtered, |ig/L

38

5.2

9.2

15.7

19.0

25.2

64.4

3

Iron, unfiltered-recoverable, j_ig/I.

38

40.7

256

505

806

823

3,860



m
m

Lead, filtered, ug/L

39(1)

0.048

0.086

0.135

0.181

0.247

0.715

4

a

CP

Lead, unfiltered-recoverable, ug/L

39

0.33

2.08

3.74

6.40

6.63

32.3



51'

Manganese, filtered, (ig/L

39

8.6

20.7

27.2

29.4

35.7

65.1

32

o

m

5*

Manganese, unfiltered-recoverable, ug/L

39

13.4

63.4

84.5

105

129

344



i

Zinc, filtered, p,g/L

39 (2)

1.9

3.1

4.9

5.6

6.9

37.1

18

w

S

Zinc, unfiltered-recoverable, ug/L

39(1)

3.2

15.9

26.7

43.9

44.5

181



<
m

Arsenic, filtered, |ig/L

39

7.8

10.9

15.2

15.0

17.3

36.7

87

m
m

Arsenic, unfiltered-recoverable, jig/L

39

10.5

15.2

17.4

20.3

21.2

46.0



s

Suspended sediment, mg/L

39

1

11

21

37

37

205

NA

T

m

Suspended sediment, percent fines4

39

46

72

79

77

83

92

NA



w


-------
Table 4. Statistical summaries of water-quality data collected at selected sampling sites in the Milltown Reservoir/Clark Fork River Superfund Site in the upper Clark Fork Basin,
Montana, water years 2011-15.—Continued

[Water year is the 12-month period from October 1 through September 30 and is designated by the year in which it ends. ftYs, cubic foot per second; NA, not applicable; uS/cm, microsiemen per centimeter at
25 degrees Celsius; CaC03, calcium carbonate; ug/L, microgram per liter; mg/L, milligram per liter]

Statistical summaries of water-quality data1

Constituent or property,
unadjusted
(not flow adjusted)
units of measurement

Number of samples
(values in parentheses
indicate number of
censored values)

Minimum
unceiisored
value2

25th
percentile

Median

Mean

75th
percentile

Maximum

filtered to median
unfiltered-recoverable
concentrations for
trace elements,
in percent3



Clark Fork at Goldcreek, Montana (sam

ipling site 16,fig. 1, table 1)







Strcamflow, instantaneous, ft3/s

40

137

393

522

820

902

4,450

NA

Specific conductance, (.iS/cm

40

216

297

364

353

411

456

NA

pH, standard units

40

7.9

8.1

8.3

NA

8.6

9.1

NA

Hardness, filtered, mg/L as CaC03

40

98.5

131

165

158

186

211

NA

Calcium, filtered, mg/L

40

29.6

38.7

48.2

46.5

55.0

62.1

NA

Magnesium, filtered, mg/L

40

5.96

8.21

10.6

10.2

12.2

13.6

NA

Cadmium, filtered, |ig/L

40 (3)

0.020

0.031

0.041

0.044

0.050

0.124

40

Cadmium, unfiltered-recoverable, |ig/L

40

0.021

0.072

0.102

0.158

0.209

0.530



Copper, filtered, ug/L

40

2.1

4.3

5.1

6.1

6.4

23.3

27

Copper, unfiltered-recoverable, ug/L

40

5.6

11.4

18.6

32.1

41.3

133



Iron, filtered, |ig/L

40(1)

3.8

8.8

18.6

25.9

36.0

93.7

5

Iron, unfiltered-recoverable, ,ug/L

40

31.8

182

360

699

922

2,940



Lead, filtered, ug/L

40 (2)

0.035

0.056

0.111

0.141

0.170

0.677

5

Lead, unfiltered-recoverable, ug/L

40

0.14

1.31

2.24

4.33

5.99

19.9



Manganese, filtered, ug/L

40

5.5

12.8

16.1

18.3

20.0

45.1

24

Manganese, unfiltered-recoverable, (ig/L

40

9.3

46.9

67.4

84.4

107

253



Zinc, filtered, ug/L

40 (5)

1.8

2.3

3.5

4.1

5.7

17.7

20

Zinc, unfiltered-recoverable, jig/L

40(1)

2.9

11.0

17.3

29.9

41.7

113



Arsenic, filtered, fig/L

40

5.6

7.9

9.0

9.9

11.5

22.5

79

Arsenic, unfiltered-recoverable, |ig/L

40

7.5

9.7

11.4

13.3

14.4

28.4



Suspended sediment, mg/L

40

2

8

16

35

40

176

NA

Suspended sediment, percent fines4

40

56

71

82

78

87

94

NA


-------
Table 4. Statistical summaries of water-quality data collected at selected sampling sites in the Milltown Reservoir/Clark Fork River Superfund Site in the upper Clark Fork Basin,
Montana, water years 2011-15.—Continued

[Water year is the 12-month period from October 1 through September 30 and is designated by the year in which it ends, ftYs, cubic foot per second; NA, not applicable; j.iS cm. microsiemen per centimeter at
25 degrees Celsius; CaCOv calcium carbonate; j.ig L. microgram per liter; mg/L, milligram per liter]

Statistical summaries of water-quality data'

Constituent or property,
unadjusted

(not flow adjusted)
units of measurement

Number of samples
(values in parentheses
indicate number of
censored values)

Minimum
uncensored
value2

25th
percentile

Median

Mean

75th
percentile

Maximum

filtered to median
unfiltered-recoverable
concentrations for
trace elements,
in percent3



Clark Fork near Drummond, Montana (sam

pling site 18,fig, 1, table 1)







Streamflow, instantaneous, ft3/s

40

248

563

781

1,040

1,090

5,540

NA

Specific conductance, (.iS/cm

40

243

346

417

403

458

560

NA

pH, standard units

40

7.9

8.1

8.1

NA

8.2

8.5

NA

Hardness, filtered, mg/L as CaCO,

40

109

158

190

184

211

265

NA

Calcium, filtered, mg/L

40

32.6

45.1

54.3

52.7

59.7

74.9

NA

Magnesium, filtered, mg/L

40

6.75

10.7

13.2

12.9

15.0

19.0

NA

Cadmium, filtered, (ig/L

40(2)

0.021

0.032

0.043

0.045

0.053

0.101

35

Cadmium, unfiltered-recoverable, |ig/L

40(1)

0.026

0.072

0.124

0.168

0.241

0.536



Copper, filtered, tig/L

40

1.9

3.9

4.8

5.6

6.2

19.8

24

Copper, unfiltered-recoverable, ug/L

40

5.4

9.8

19.4

29.9

36.7

107



Iron, filtered, |ig/L

40 (2)

3.6

9.0

15.0

20.7

26.9

88.7

3

Iron, unfiltered-recoverable, jig/L

40

24.8

180

440

710

979

3,170



Lead, filtered, (ig/L

40(2)

0.039

0.059

0.115

0.142

0.152

0.592

4

Lead, unfiltered-recoverable, (.ig/L

40

0.17

1.27

3.02

4.61

6.29

19.8



Manganese, filtered, ug/L

40

4.2

12.4

15.5

17.4

22.0

37.7

20

Manganese, unfiltered-recoverable, ug/L

40

9.9

49.7

76.6

96.0

121

294



Zinc, filtered, ug/L

40(1)

2.2

3.5

4.3

4.7

5.3

13.2

19

Zinc, unfiltered-recoverable, (ig/L

40(1)

4.6

12.0

22.7

35.0

48.3

134



Arsenic, filtered, ug/L

40

6.3

8.0

9.7

10.1

11.3

23.9

86

Arsenic, unfiltered-recoverable, ug/L

40

7.7

10.6

11.3

13.3

13.9

30.7



Suspended sediment, mg/L

40

2

9

22

40

57

216

NA

Suspended sediment, percent fines4

40

42

68

79

76

86

93

NA


-------
Table 4. Statistical summaries of water-quality data collected at selected sampling sites in the Milltown Reservoir/Clark Fork River Superfund Site in the upper Clark Fork Basin,
Montana, water years 2011-15.—Continued

[Water year is the 12-month period from October 1 through September 30 and is designated by the year in which it ends. ftYs, cubic foot per second; NA, not applicable; uS/cm, microsiemen per centimeter at
25 degrees Celsius; CaC03, calcium carbonate; ug/L, microgram per liter; mg/L, milligram per liter]

Statistical summaries of water-quality data1

Constituent or property,
unadjusted
(not flow adjusted)
units of measurement

Number of samples
(values in parentheses
indicate number of
censored values)

Minimum
unceiisored
value2

25th
percentile

Median

Mean

75th
percentile

Maximum

filtered to median
unfiltered-recoverable
concentrations for
trace elements,
in percent3

Clark Fork at Turah Bridge near Bonner, Montana (sampling site 20, fig. 1, table 1)

Strcamflow, instantaneous, ft3/s

40

462

1,050

1,500

2,230

2,640

10,600

NA

Specific conductance, (.iS/cm

40

140

214

285

277

340

385

NA

pH, standard units

40

7.8

8.0

8.1

NA

8.2

8.4

NA

Hardness, filtered, mg/L as CaC03

40

60.1

97.6

132

127

156

186

NA

Calcium, filtered, mg/L

40

17.3

27.8

35.9

35.6

43.5

52.8

NA

Magnesium, filtered, mg/L

40

4.11

6.76

9.57

9.27

11.6

13.1

NA

Cadmium, filtered, jig/L

40(12)

0.017

0.019

0.027

0.031

0.037

0.083

37

Cadmium, unfiltered-recoverable, |ig/L

40 (3)

0.025

0.048

0.073

0.104

0.132

0.404



Copper, filtered, ug/L

40

1.3

2.2

2.9

3.8

3.9

17.9

27

Copper, unfiltered-recoverable, ug/L

40

3.8

5.9

10.5

16.8

20.1

61.9



Iron, filtered, |ig/L

40 (3)

3.3

7.1

20.4

29.7

34.0

359

6

Iron, unfiltered-recoverable, jig/L

40

47.7

132

316

507

527

2,450



Lead, filtered, jig/L

40 (6)

0.030

0.039

0.069

0.134

0.137

2.79

4

Lead, unfiltered-recoverable, ug/L

40

0.20

0.58

1.67

2.67

3.33

11.9



Manganese, filtered, |ig/L

40

3.0

5.4

6.9

9.6

9.8

48.6

16

Manganese, unfiltered-recoverable, jig/L

40

9.5

26.9

43.5

57.7

66.7

212



Zinc, filtered, (.ig/L

40 (4)

1.5

2.3

3.3

3.9

4.7

17.4

23

Zinc, unfiltered-recoverable, ug/L

40

3.8

8.4

14.2

22.9

27.3

109



Arsenic, filtered, |ig/L

40

2.7

4.5

5.6

5.6

6.0

14.2

90

Arsenic, unfiltered-recoverable, ug/L

40

3.0

5.6

6.2

7.3

8.2

21.0



Suspended sediment, mg/L

40

3

7

16

32

30

186

NA

Suspended sediment, percent fines4

40

44

66

78

74

85

91

NA


-------
Table 4. Statistical summaries of water-quality data collected at selected sampling sites in the Milltown Reservoir/Clark Fork River Superfund Site in the upper Clark Fork Basin,
Montana, water years 2011-15.—Continued

I Water year is the 12-month period from October 1 through September 30 and is designated by the year in which it ends. ftVs, cubic foot per second; NA, not applicable; [iS cm. microsiemen per centimeter at
25 degrees Celsius; CaCO . calcium carbonate; fig/L, microgram per liter; mg/L. milligram per liter]

Statistical summaries of water-quality data1 n . , ..
	 	 Ratios of median

Constituent or property,
unadjusted
(not flow adjusted)
units of measurement

Number of samples
(values in parentheses
indicate number of
censored values)

Minimum
uneensored
value2

25th
percentile

Median

Mean

75th
percentile

Maximum

filtered to median
unfiltered-recoverable
concentrations for
trace elements,
in percent3



Clark Fork above Missoula, Montana (

sampling site 22, fig. 1, table 1 j







Streamflow, instantaneous, IP/s

40

910

1,710

4,100

5,530

7,240

22,900

NA

Specific conductance, p.S/cm

40

148

189

230

239

288

341

NA

pH, standard units

40

8.0

8.2

8.3

NA

8.4

8.7

NA

Hardness, filtered, mg/L as CaC(J3

40

70.7

88.5

109

113

141

163

NA

Calcium, filtered, mg/L

40

19.3

23.8

29.5

30.3

36.9

44.9

NA

Magnesium, filtered, mg/L

40

5.30

6.98

8.48

9.06

11.3

12.9

NA

Cadmium, filtered, |ig/L

40 (25)

0.017

0.014

0.018

0.019

0.023

0.046

47

Cadmium, unfiltered-recoverable, ug/L

40 (12)

0.020

0.021

0.038

0.056

0.067

0.345



Copper, filtered, ug/L

40

1.0

1.5

1.7

2.1

2.1

7.0

35

Copper, unfiltered-recoverable, ug/L

40

1.9

3.2

4.8

9.0

9.4

53.1



Iron, filtered, (ig/L

40(1)

3.7

6.6

17.3

22.6

34.2

60.5

7

Iron, unfiltered-recoverable, ug/L

40

40.9

96.3

255

370

344

2,030



Lead, filtered, jig/L

40 (10)

0.026

0.031

0.054

0.068

0.089

0.212

7

Lead, unfiltered-recoverable, (ig/I,

40

0.13

0.41

0.80

1.40

1.57

8.04



Manganese, filtered, ug/L

40

3.5

5.5

6.5

7.7

8.5

20.0

24

Manganese, unfiltered-recoverable, (.ig/L

40

8.8

19.7

27.5

36.5

41.4

155



Zinc, filtered, fig/L

40(16)

1.4

1.4

1.9

2.0

2.4

5.5

26

Zinc, unfiltered-recoverable, |ig/L

40 (4)

2.3

4.9

7.2

12.2

12.2

84.4



Arsenic, filtered, ug/L

40

1.2

2.5

3.1

3.1

3.6

7.1

85

Arsenic, unfiltered-recoverable, jig/L

40

1.4

2.9

3.7

4.1

4.8

13.2



Suspended sediment, mg/L

40

2

5

13

26

20

176

NA

Suspended sediment, percent fines4

40

61

73

82

79

85

95

NA

distributional parameters affected by censored observations (that is, concentrations reported as less than the laboratory reporting level) were estimated by using adjusted maximum likelihood estimation (Colin, 1988).
2Minimum uneensored value refers to the smallest concentration reported as delected above any of the various laboratory reporting levels applicable for a given constituent,

3Ratio of median filtered to unfiltered-recoverable concentration greater than 100 percent affected by low median concentrations near minimum laboratory reporting levels (table 2) and small bias in filtered concentrations,
'Percent lines refers to the percentage of suspended sediment smaller than 0.062-millimeter diameter.


-------
18 Water-Quality Trends and Constituent-Transport Analysis for Selected Sampling Sites

Table 5. Percentages of samples with unadjusted unfiltered-recoverable concentrations exceeding water-quality standards for
selected sampling sites in the Milltown Reservoir/Clark Fork River Superfund Site in the upper Clark Fork Basin, water years 2011-15,

[Water year is the 12-month period from October 1 through September 30 and is designated by the year in which it ends. CaCO , calcium carbonate]

Sampling

site
number

(fig. 1.

table 1)

Abbreviated sampling site name
(table 1)

Arsenic
human-
health

standard

Percentage of samples exceeding indicated standard

Aquatic-life standards

Cadmium

Copper

Lead

Zinc

Acute Chronic Acute Chronic Acute Chronic Acute Chronic

8

Silver Bow Creek at Warn Springs

100

0

3

8

18

0

3

0

0

11

Clark Fork near Galen

98

0

0

26

41

0

8

0

0

14

Clark Fork at Deer Lodge

95

0

15

58

75

0

23

3

3

15

Clark Fork near Garrison

100

0

18

59

79

0

23

3

3

16

Clark Fork at Goldcreek

68

0

18

48

60

0

28

0

0

18

Clark Fork near Drummond

80

0

15

38

58

0

25

3

3

20

Clark Fork at Turah Bridge

13

0

13

28

48

0

25

0

0

22

Clark Fork above Missoula

3

0

5

15

23

0

13

0

0

Reach 4

Reach 4 extends about 2 river miles from Silver Bow
Creek at Warm Springs, Montana (sampling site 8), to Clark
Fork near Galen, Montana (sampling site 11). Within the
reach, water from Warm Springs Ponds mixes and geochemi-
callv reacts with water contributed from the Mill-Willow
Bypass and Warm Springs Creek; thus, complex water-quality
processes are possible in the short reach.

The Warm Springs Ponds system was originally con-
structed during 1908-17 (and expanded during the 1950s)
to trap sediment enriched in trace elements (CDM, 2005). In
about 1967, the AMC started introducing a lime and water
suspension into Silver Bow Creek upstream from Warm
Springs Ponds to raise pH and promote precipitation and
deposition of metals in Warm Springs Ponds (U.S. Environ-
mental Protection Agency, 2000). The Mill-Willow Bypass
was constructed in about 1969 to capture streamllow s of Mill
and Willow Creeks near their mouths and divert the combined
streamfiows (believed to be relatively clean water; U.S. Envi-
ronmental Protection Agency, 2000) around Warm Springs
Ponds and into Silver Bow Creek between the outlet from the
Warm Springs Ponds and sampling site 8 (CDM, 2005). Warm
Springs Creek originates in the mountains west of the AMC
Smelter, flows generally east through areas adjacent to the
AMC Smelter and various tailings piles and ponds, and joins
Silver Bow Creek to form the Clark Fork near Warm Springs.
The Warm Springs Creek Basin is affected by pollution from
milling and smelting operations of the AMC Smelter. Thick
tailings deposits are extensive in the Silver Bow Creek and
Clark Fork flood plain near Warm Springs (Smith and others,
1998) and provide a source of sediment enriched with metallic
contaminants within reach 4.

In reach 4, the mean annual streamflow for water years
2011 -15 increased by about 79 percent from 96 cubic feet
per second (ftVs) at sampling site 8 to 172 ftVs at sampling
site 11 (table 3) primarily because of contributions from Warm
Springs Creek and also ephemeral gulches and groundwater
inflow. Near the end of reach 4, Warm Springs Creek and
Silver Bow Creek join to form the Clark Fork.

Silver Bow Creek at Warm Springs (sampling site 8) is
about 0.2 river mile downstream from Warm Springs Ponds,
which were designed to trap suspended sediment and metallic
contaminants by physical deposition and treatment (liming;
U.S. Environmental Protection Agency, 2000). Median con-
centrations of unfiltered-recoverable copper and zinc (6.8 and
8.6 ng/L, respectively) and suspended sediment (6 milligrams
per liter [mg/L]) are lower than median concentrations of
most downstream main-stem Clark Fork sampling sites (fig. 2,
table 4). The median concentration of unfiltered-recoverable
arsenic (22.4 p.g/L) at sampling site 8 is higher than median
concentrations at the downstream main-stem Clark Fork sam-
pling sites. The high median arsenic concentration at sampling
site 8 is affected by contributions of water with high arsenic
concentrations from the Mill-Willow Bypass and by complex
hydrologic and limnologic factors that affect arsenic biogeo-
chemical processing in Warm Springs Ponds (Chatham, 2012).
The median pH for sampling site 8 is 8.8 standard units,
which is higher than the median pH of the downstream main-
stem Clark Fork sampling sites (table 4). High pH in Warm
Springs Ponds (a result of a combination factors, including
liming and nutrient processing by aquatic vegetation; Cha-
tham, 2012) promotes arsenic solubility and mobilization
(Stumm and Morgan, 1970). Exceedances of most water-
quality standards were infrequent (that is, less than or equal
to 20 percent of samples) for sampling site 8; however, the


-------
1,000

~S 3

00-—

.3.o.i s

gS

= - - S
S"g

2 „ M 1-
^ ° C 05
'	' u Q) Q)

"o u P"a

2=

S'5-55^
#g.E ™

^ iZ S=

03 CD O

E

° .9

—.as	a> ,i=i

"a »-	o —

CD CD	<— i_

S? =	o £

pi g] q)
^ a3 ^'

J -t—' TO

a)

E =

E

£ S.eJ

lift

l£l

c/5 c/5
—, —5 CU

I A. Specific conductance

100
200

100

~I	1-

reach
4

reach
5



reach
6

reach
7

reach

— B. Copper

~l	1-

3

1

—i	r~

15 16

? t

20

200
100

200
100

— D. Suspended sediment

14
i

15
i

16
i

18
i

20
i







v

y/ w

> y ^r

\s>



Vi"

Sampling site

<<**>
'S

reach
9

<^>V

22

I



EXPLANATION

[Water year is defined as the
12-month period from October 1
through September 30 and is
designated by the year in which
it ends]

90th percentile
75th percentile

<

i	

1

senic

? a

i i

i fi

I

^ ^ d

i

^ d

i

i

i

=



T

1

i i

1 1

i i

4 15 1

i

^ ~

3 1

i

8 2

i

r

0 2

i

J .

2

1

Inter-

median quartile

range

25th percentile—I
10th percentile

22 Sampling site number (fig. 1,
table 1)— Site numbers
that lie on top of reach
dividing lines indicate that
the site is both an outflow
from the upstream reach
and an inflow to the
downstream reach

Aquatic-life standards for
copper based on median
hardness for the given site
for water years 2011-15
(Montana Department of
Environmental Quality, 2012)

—	Acute

—	Chronic

	Human-health standard for

arsenic (Montana Department
of Environmental Quality, 2012)

For copper and arsenic,
green boxplots denote filtered
concentrations

For copper and arsenic,
magenta boxplots denote
u nfi Ite red-re cove ra bl e
concentrations



Figure 2. Statistical distributions of selected constituents for selected sampling sites in the Milltown Reservoir/Clark Fork River Superfund Site in the upper Clark Fork Basin,
Montana, water years 2011-15. A, specific conductance; B, copper; C, arsenic; and D, suspended sediment.

o

<
CD

2

o

CO

CD
Q>

3

a»
CD
¦

O
c

Q>

-<
o

3"

Q)

O)
O
CD

Ui'

O*

«/>

a»
CD

CD
a»

«/>
rsj

CJl


-------
20 Water-Quality Trends and Constituent-Transport Analysis for Selected Sampling Sites

arsenic human-health standard was exceeded in 100 percent of
samples (table 5).

Clark Fork near Galen (sampling site 11) is about 2 river
miles downstream from sampling site 8 and about 1 river mile
downstream from the start of the Clark Fork at the conflu-
ence of Silver Bow Creek and Warm Springs Creek. Spatial
changes in water quality between sampling sites 8 and 11 in
water years 2011-15 include increases in median concentra-
tions of unfiltered-recoverable metallic trace elements and
suspended sediment, as well as decreases in median concentra-
tions of unfiltered-recoverable arsenic (fig. 2, table 4). Factors
that might contribute to the patterns include mobilization of
materials from flood-plain tailings deposits near Warm Springs
and complex processes as water from Warm Springs Ponds
mixes and geochemically reacts with water contributed from
the Mill-Willow Bypass and Warm Springs Creek. Exceed-
ances of most water-quality standards were somewhat infre-
quent for sampling site 11. but the acute aquatic-life standard
for copper was exceeded in 26 percent of samples, the chronic
aquatic-life standard for copper was exceeded in 41 percent of
samples, and the arsenic human-health standard was exceeded
in 98 percent of samples (table 5).

Reach 5

Reach 5 extends about 21 river miles from Clark Fork
near Galen (sampling site 11) to Clark Fork at Deer Lodge,
Montana (sampling site 14), and meanders through a broad
valley with extensive flood-plain tailings deposits. Lost Creek
(a tributary to the Clark Fork in reach 5) originates in the
mountains northwest of the AMC Smelter and flows generally
east to its confluence with the Clark Fork near Galen. The Lost
Creek Basin is affected by pollution from milling and smelting
operations of the AMC Smelter (U.S. Environmental Protec-
tion Agency, 2010). In reach 5, the mean annual streamflow
for water years 2011-15 increased by about 65 percent from
172 ff'/s at sampling site 11 to 283 ftVs at sampling site 14
(table 3) partly because of contributions from Lost Creek
and also numerous other tributaries, ephemeral gulches, and
groundwater inflow.

Spatial changes in water quality between sampling
sites 11 and '14 in water years 2011-15 include substantial
increases in median concentrations of unfiltered-recoverable
metallic trace elements and suspended sediment (fig. 2,
table 4). Mobilization of mining wastes from extensive flood-
plain tailings deposits and stream banks contribute to the
pattern. Exceedances of water-quality standards were frequent
for sampling site 14: the acute aquatic-life standard for copper
was exceeded in 58 percent of samples, the chronic aquatic-
life standard for copper was exceeded in 75 percent of sam-
ples, the chronic aquatic-life standard for lead was exceeded in
23 percent of samples, and the arsenic human-health standard
was exceeded in 95 percent of samples (table 5).

Reach 6

Reach 6 extends about 26 river miles from Clark Fork
at Deer Lodge (sampling site 14) to Clark Fork at Goldcreek,
Montana (sampling site 16). Clark Fork above Little Black-
foot River near Garrison (sampling site 15), is in reach 6 and
is located about 14 river miles downstream from sampling
site 14 and about 12 river miles upstream from sampling
site 16. Water-quality data collection for sampling site 15
began in water year 2009 (tabic 1); thus, water-quality data
for sampling site 15 are suitable for summarizing water years
2011 -15 water-quality characteristics but are not adequate for
trend analysis.

The Clark Fork meanders through a broad valley from
Deer Lodge to Garrison, in which flood-plain tailings along
the Clark Fork are present to a similar extent as in the valley
upstream from Deer Lodge (Smith and others, 1998). The
Little Blackfoot River (a tributary to the Clark Fork in reach
6) drains a basin with moderate density of agricultural and
historical mining activity (in comparison with other tributar-
ies downstream from Deer Lodge) and discharges into reach
6 near Garrison (about 1 river mile downstream from sam-
pling site 15) where the Clark Fork Valley begins to narrow.
Downstream from Garrison, flood-plain tailings are less
extensive than in the valley upstream. In reach 6, the mean
annual streamflow for water years 2011-15 increased by about
11 percent from 283 ft3/s at sampling site 14 to 315 ft3/s at
sampling site 15 and then by about 81 percent to 570 ftVs at
sampling site 16 (table 3). The overall increase in streamflow
from sampling site 14 to sampling site 16 was about 101 per-
cent, mostly because of contributions from the Little Blackfoot
River and also numerous other tributaries, ephemeral gulches,
and groundwater inflow.

Spatial changes in water quality between sampling
sites 14 and 16 in water years 2011-15 include decreases in
median concentrations of unfiltered-recoverable metallic trace
elements, unfiltered-recoverable arsenic, and suspended sedi-
ment, despite small increases in most of these values between
sampling sites 14 and 15. Water-quality changes in reach 6
primarily were affected by transport of mining wastes from
upstream source areas in combination with streamflow inputs
from areas with less mining effects (including the Little Black-
foot River). Dispersion and dilution of mining wastes gener-
ally result in decreasing water-quality effects with distance
downstream from primary source areas. Exceedances of water-
quality standards were frequent for sampling site 15: the acute
aquatic-life standard for copper was exceeded in 59 percent
of samples, the chronic aquatic-life standard for copper was
exceeded in 79 percent of samples, the chronic aquatic-life
standard for lead was exceeded in 23 percent of samples, and
the arsenic human-health standard was exceeded in 100 per-
cent of samples (table 5). Exceedances of water-quality stan-
dards were somewhat frequent for sampling site 16: the acute
aquatic-life standard for copper was exceeded in 48 percent


-------
Overview of Streamflow and Water-Quality Characteristics for Water Years 2011-15 21

of samples, the chronic aquatic-life standard for copper was
exceeded in 60 percent of samples, the chronic aquatic-life
standard for lead was exceeded in 28 percent of samples, and
the arsenic human-health standard was exceeded in 68 percent
of samples (table 5).

Reach 7

Reach 7 extends about 31 river miles from Clark Fork at
Goldcreck (sampling site 16) to Clark Fork near Driimmoiid,
Montana (sampling site 18). In reach 7, channel meandering
and exposed flood-plain tailings are less extensive than in
upstream reaches (Lambing, '1998; Smith and others, 1998).
Flint Creek (a tributary that discharges to the Clark Fork in
reach 7 near Drummond) drains a basin with high density of
agricultural and historical mining activity (in comparison with
other tributaries downstream from Deer Lodge). Downstream
from Drummond, the Clark Fork Valley narrows further, and
meandering of the Clark Fork decreases further in association
with the narrow valley and presence of highway and railroad
embankments (Lambing, 1998; Smith and others, 1998). In
reach 7, the mean annual streamflow for water years 2011-15
increased by about 35 percent from 570 ftVs at sampling
site 16 to 771 ff'/s at sampling site 18 (table 3) mostly because
of contributions from Flint Creek and also numerous oilier
tributaries, ephemeral gulches, and groundwater inflow.

Spatial changes in water quality between sampling
sites 16 and 18 in water years 2011-15 include generally small
increases in median concentrations of unfiltered-recoverable
metallic trace elements and suspended sediment. Although
the increases were not large, they contrast with the pattern of
decreasing water-quality effects with distance downstream
from primary mining-waste source areas in the upper Clark
Fork Basin. The spatial changes in water quality between
sites 16 and 18 probably were affected by streamflow contri-
butions from the Flint Creek Basin, which lias high density of
agricultural and historical mining activity (in comparison with
other tributaries downstream from Deer Lodge). The Clark
Fork flood plain and stream banks downstream from Flint
Creek probably also contain mining-waste deposits sourced
from the Flint Creek Basin. Exceedances of water-quality stan-
dards were somewhat frequent for sampling site 18: the acute
aquatic-life standard for copper was exceeded in 38 percent
of samples, the chronic aquatic-life standard for copper was
exceeded in 58 percent of samples, the chronic aquatic-life
standard for lead was exceeded in 25 percent of samples, and
the arsenic human-health standard was exceeded in 80 percent
of samples (table 5).

Reach 8

Reach 8 extends about 34 river miles from Clark Fork
near Drummond (sampling site 18) to Clark Fork at Turah
Bridge near Bonner, Montana (sampling site 20). In reach 8,
the Clark Fork flows through a narrow flood plain (generally
less than 1 mi wide) with little or no visible mining tailings.
Rock Creek (a tributary to the Clark Fork in reach 8) drains
a heavily forested basin with low density of agricultural and
historical mining activity (in comparison with other tributar-
ies downstream from Deer Lodge) and discharges into reach 8
near Clinton, Montana. In reach 8, the mean annual stream-
flow for water years 2011-15 increased by about 93 percent
from 771 ftVs at sampling site 18 to 1,490 ft3/s at sampling
site 20 (table 3) primarily because of contributions from
Rock Creek, as well as numerous other tributaries, ephemeral
gulches, and groundwater inflow.

Spatial changes in water quality between sampling
sites 18 and 20 in water years 2011-15 include generally
substantial decreases in median concentrations of unfiltered-
recoverable metallic trace elements, unfiltered-recoverable
arsenic, and suspended sediment. Water-quality changes in
reach 8 were affected by dilution from Rock Creek. Exceed-
ances of most water-quality standards were somewhat: infre-
quent for sampling site 20, but the acute aquatic-life standard
for copper was exceeded in 28 percent of samples, the chronic
aquatic-life standard for copper was exceeded in 48 percent
of samples, and the chronic aquatic-life standard for lead was
exceeded in 25 percent of samples (table 5).

Reach 9

Reach 9 extends about 9 river miles from Clark Fork at
Turah Bridge (sampling site 20) to Clark Fork above Mis-
soula, Montana (sampling site 22). Reach 9 includes the
former Milltown Reservoir where large amounts of min-
ing wastes had been deposited. The former Milltown Dam
was removed in 2008. The Blackfoot River (a tributary that
discharges to the Clark Fork in reach 9 near Bonner) drains
a largely forested basin with low density of agricultural and
historical mining activity (in comparison with other tributar-
ies downstream from Deer Lodge). In reach 9, mean annual
streamflow increased by about 123 percent from 1,490 ftVs
at sampling site 20 to 3,330 ft3/s at sampling site 22 (table 3)
primarily because of contributions from the Blackfoot River.

Spatial changes in water quality between sampling
sites 20 and 22 in water years 2011-15 include generally
substantial decreases in median concentrations of unfiltered-
recoverable metallic trace elements, unfiltered-recoverable
arsenic, and suspended sediment. Water-quality changes in
reach 9 were affected by dilution from the Blackfoot River.
Exceedances of most water-quality standards were infrequent
for sampling site 22. but the chronic aquatic-life standard for
copper was exceeded in 23 percent of samples (table 5).


-------
22 Water-Quality Trends and Constituent-Transport Analysis for Selected Sampling Sites

Water-Quality Trend- and Constituent-
Transport Analysis Hethods

This section of the report describes methods used to
analyze trends in flow-adjusted concentrations of water-quality
constituents. Normalized loads (as defined in the section of
this report "Estimation of Normalized Constituent Loads")
were estimated to evaluate temporal changes in relative con-
tributions of selected trace elements and suspended sediment
from upstream source areas to the outflows of each data-
summary reach. Methods used for estimation of normalized
constituent loads also are described.

General Description of the Time-Series Model

The TSM for streamflow and constituent concentra-
tion (Vecchia, 2005) was used to detect water-quality trends.
Details on theory and parameter estimation for the model are
presented in Vecchia (2005). and the model is summarized
in appendix 2 of this report. Specific information concerning
suitability of application of the TSM to the study datasets and
procedures for determination of statistical significance and
magnitude of trends also are presented in appendix 2.

The TSM analyzes trends in flow-adjusted concentrations
(FACs); that is, the TSM computes FACs, estimates unbiased
best-fit trend lines that represent temporal changes in FACs,
and determines statistical significance of changes. Flow adjust-
ment is necessary because concentrations of many water-
quality constituents are strongly dependent on stream (low
conditions, which are primarily affected by climatic variability
in the study area. The intent of flow adjustment is to identify
and remove streamflow-related variability in concentrations
and thereby enhance the capability to detect trends indepen-
dent from effects of climatic variability. Flow-adjustment
procedures produce FACs that arc estimates of constituent
concentrations after removing effects of streamflow variability.

The TSM uses multiple flow-related variables computed
from concurrent (same day as the concentration sample) and
antecedent (days before the concentration sample) daily mean
streamflow in the flow-adjustment process. The TSM FACs
provide detailed accounting by incorporating interannual, sea-
sonal, and short-term streamflow variability (Vecchia, 2005),
which compensates for interannual, seasonal, and short-term
hysteresis processes that affect concentration and streamflow
relations (Colby, 1956; Chanat and others, 2002; Vecchia,
2005). Detailed analysis of continuous streamflow data pro-
vides definition of the context of streamflow conditions associ-
ated with a given water sample, handling of temporal variabil-
ity in sampling frequency, and interpolation of trend patterns
to periods when water-quality data are sparse or absent. The
TSM inherently accounts for effects of serial correlation.

The TSM incorporates base-10 logarithm (hereinafter
referred to as "'log") transformation of the concentration and
streamflow data. As such, the fitted trends in FACs quantify

temporal changes in central tendency represented by the geo-
metric mean of concentration in reference to log-transformed
streamflow. The geometric mean is the mean of the logs trans-
formed back into their original units.

All of the study datasets (except for Clark Fork near Gar-
rison [sampling site 15], which was not analyzed for trends)
met the data criteria for applying the TSM, which include
at least 15 years of continuous streamflow data and at least
15 years of water-quality data with at least 60 total water-
quality samples and at least 10 samples total in each 3-month
season (Vecchia, 2005). A limitation of the TSM is that it does
not handle censored data in a rigorous manner. In the TSM,
a single value is substituted for all censored data for a given
constituent; thus, criteria must be set to specify the allowable
amount of censored data and a consistent substitution value
for each constituent. Based on analysis of trial datasets with
artificially imposed variable levels of censoring, the TSM
generally can be applied to datasets with about 10 percent or
less censored data without substantial effects on trend results
(Vecchia, 2003). Multiple LRLs (table 2) in the datasets of
the Clark Fork monitoring program complicate the task of
setting consistent substitution values. In applying the TSM to
the study datasets. study reporting levels (SRLs; table 2) were
established to set consistent substitution values for each trace-
element constituent based on investigation of the time frame
during which various NWQL LRLs were used, the frequency
of censoring that resulted from each LRL, and field blank
sample data that provided information on potential contami-
nation bias of low concentrations. The SRLs were applied to
the study datasets by (1) substituting one-half the SRL for all
censored observations with LRLs equal or close to the SRL,
(2) substituting one-half the SRL for all reported uncensored
concentrations (analyzed during times when the LRL- was less
than the SRL) that were less than the SRL, and (3) excluding
censored data with LRLs substantially larger than the SRL.
Any analytical result that was revised by either substitution
or exclusion was considered to be affected by the recensor-
ing procedures used in applying the SRL. The study datasets
largely were unaffected by recensoring for the trace-element
constituents included in the trend analysis (table 2); unfi he red-
recoverable zinc was the only affected constituent, and no
sampling site had more than 8.5 percent of values affected
by the recensoring procedures. Further, for individual con-
stituents, the maximum frequency of detection in field blank
samples at concentrations greater than the SRL was 2.7 per-
cent (for unfiltcrcd-recovcrable zinc; table 1-1).

The TSM accounts for many hydrologic factors that
contribute to complexity in concentration and streamflow
relations. In this study, the TSM was applied as consistently
as possible among sampling-site and constituent combinations
and is considered to be a useful tool for simplifying the envi-
ronmental complexity in the upper Clark Fork Basin to pro-
vide a large-scale evaluation of general temporal changes in
FACs and constituent transport independent from streamflow
variability. As such, the TSM provides a consistent relational
framework for evaluating temporal water-quality changes


-------
Water-Quality Trend- and Constituent-Transport Analysis Methods 23

among the sampling sites. The TSM best-fit trend lines were
considered to provide important information beyond the strict
statistical characteristics of the trend results (in terms of sta-
tistical probability levels [p-values] and levels of significance)
because they aid in comparing and summarizing large-scale
patterns among sampling sites.

Selection of Trend-Analysis Time Periods

Appropriate selection of trend-analysis time periods is
important because the results of trend analyses are dependent
on how the time periods are structured. Factors considered
in selection of trend-analysis time periods included provid-
ing capability to (1) compare trend results among sampling
sites with different periods of data collection, (2) distinguish
somewhat short-term timing of changes in concentration
and streamflow relations during the long study period, and
(3) allow periodic future updates of trend analyses for evalu-
ation of effects of remediation activities. Based primarily on
those factors, trend-analysis periods were defined as sequential
5-year periods that extended from near the start of long-term
data-collection activities for most sampling sites in the upper
Clark Fork Basin to the end of water year 2015. Thus, four
trend-analysis lime periods were defined: period 1 (water years
1996-2000), period 2 (water years 2001-5). period 3 (water
years 2006-10), and period 4 (water years 2011-15).

The TSM-fitted trends for a given trend-analysis period
are monotonic trends that are smoothed to produce gen-
erally consistent slopes across the middle section of the
trend-analysis period that become flatter near the ends of the
trend-analysis period. The flatter slopes near the ends provide
gradual transition between adjacent trend-analysis periods. In
some cases, the fitted trends in a given trend-analysis period
do not precisely follow the patterns in FACs, and there are
short-term (about 1-2 years) trend patterns in FACs that are
unresolved in the fitted trends. In those cases, better temporal
resolution might have been attained by defining two or more
trend-analysis periods in a given 5-vear trend-analysis period.
This approach generally was avoided because it would have
required detailed tend analysis for potentially inconsistent-
time periods among the various sampling-site and constituent
combinations. An important consideration in the design of the
trend-analysis structure of this study was making general com-
parisons among the sampling-site and constituent combina-
tions to evaluate large-scale effects of mining and remediation
activities for consistent time periods. In general, when unre-
solved trending was apparent, more complicated trend models
(with additional trend-analysis periods) were tested, and the
more complicated models did not change the general findings
and conclusions of this report; that is. the overall fitted trends
in the affected trend-analysis periods were consistent with
overall patterns in FACs in the period. However, because of
the substantial effect of the intentional breach of the former
Milltown Dam on March 28, 2008, an exception to consis-
tent trend-analysis periods was made. For Clark Fork above

Missoula (sampling site 22), period 3 was subdivided into
period 3A (October 1, 2005-March 27, 2008) and period 3B
(March 28, 2008-September 30, 2010). The intentional breach
of the former Milltown Dam was part of an extensive remedia-
tion effort from about 2006-8 that resulted in the removal of
the former Milltown Dam (Sando and Lambing, 2011).

Estimation of Normalized Constituent Loads

Normalized constituent loads were estimated to assess the
temporal trends in FACs of mining-related contaminants in the
context of sources and transport. The fitted tends are unbi-
ased best-fit lines through the FACs, which are independent of
streamflow variability. The FAC tends at individual sampling
sites are important descriptors of water-quality changes in the
upper Clark Fork Basin, but without consideration of differ-
ences in streamflow magnitudes among different sampling
sites, the trends do not provide direct information on resultant
changes in contaminant source-area contributions and trans-
port: characteristics. Combining the FAC trends with a station-
ary streamflow index (that maintains relative differences in
streamflow magnitudes among sampling sites but normal-
izes streamflow for a given sampling site to a constant value
through time) allows assessment of how the temporal changes
in FACs translate into relative temporal changes in source and
transport of mining-related contaminants in the upper Clark
Fork Basin. Thus, normalized loads were estimated to conduct
a transport analysis.

Normalized loads were estimated for each of the four
5-year trend-analysis periods. The stationary streamflow index
used in estimating normalized loads was the geometric mean
streamflow for each sampling site for water years 1996-2015.
The geometric mean was selected as a measure of central
tendency in streamflow to maintain consistency with the TSM
analysis, which is conducted on log-transformed data.

For each sampling-site and constituent combination and
each of the 5-year periods, the normalized load was estimated
by multiplying the mean annual fitted trend FAC during the
5-year analysis period times the geometric mean streamflow
for water years '1996-2015 and a units conversion factor,
according to the following equation:

LOAD = MAC*GMQ*K	(1)

where

LOAD is the estimated normalized constituent load
(in kilograms per day) for the indicated
5-vear period;

MAC is the mean annual filled trend FAC (in

micrograms per liter for trace elements
or milligrams per liter for suspended
sediment) for the indicated 5-year period;

GMO is the geometric mean of daily mean

streamflow for water years 1996-2015, in
cubic feet per second; and


-------
24 Water-Quality Trends and Constituent-Transport Analysis for Selected Sampling Sites

K is a units conversion constant (0.00245 for
concentrations in micrograms per liter or
2.45 for concentrations in milligrams per
liter) to convert instantaneous constituent
discharge (in mass units per second) to
an equivalent daily constituent load (in
kilograms per day).

The MIC is calculated by temporally averaging (in each
of the four 5-year periods) the fitted trend FACs that quantify
temporal changes in central tendency based on the geometric
mean. It is notable that the M4C is referred to as a "mean
annual value"; this terminology indicates temporal averag-
ing of geometric mean concentrations. The temporal averag-
ing of geometric mean concentrations in each 5-year period
effectively results in the M4C representing the center of the
5-year period, which introduces a conservative approach to the
transport analysis. The geometric mean generally is closely
associated with the median of the original untransformed
units for data that are approximately log-normally distributed.
Thus, because of effects of analysis of log-transformed data,
the estimated normalized loads generally represent quantifi-
cation with respect to near-median conditions. As such, the
estimated normalized loads do not represent actual magnitudes
of total mass transport, but rather provide information on
relative temporal changes in constituent transport character-
istics of the study sampling sites quantified with respect to
near-median conditions.

Factors that Affect Trend Analysis and
Interpretation

Several factors affect temporal trends in water quality.
Climatic variability (interannual and seasonal) is indicated
in variability in streamflow conditions, which strongly affect
concentration and streamflow relations. Investigating stream-
flow conditions during the study period is relevant to inter-
preting trend results. Other factors relating to data assessment
or treatment that also are relevant to understanding trend-
analysis procedures and interpreting trend results include
relations between unadjusted concentrations and FACs, and
data transformation.

Streamflow Conditions

Daily mean streamfiows for water years 1993-2015
for selected sampling sites in the Milltown Reservoir/Clark
Fork River Supcrfund Site in the upper Clark Fork Basin arc
presented in figure 3. Locally weighted scatter plot smooth
(LOWESS; Cleveland and McGill, 1984; Cleveland, 1985)
lines through the daily mean streamilows also are presented in
figure 3 to represent temporal variability in the moving central
tendency of streamflow. The geometric mean streamfiows

for water years 1996-2015 are presented to represent overall
central tendency of streamflow during the period of trend
analysis. Silver Bow Creek at Warm Springs (sampling site 8),
Clark Fork at Deer Lodge (sampling site 14), and Clark Fork
at Turah Bridge (sampling site 20) were selected as examples
for showing hydrologic patterns (fig. 3) that generally apply to
the other sampling sites.

Temporal variability in streamflow conditions during
the study period generally is similar among sampling sites.
In about water year 1993, streamflow conditions generally
increased to above the geometric mean streamfiows dur-
ing a period of several years. Streamfiows were high during
water years 1996-97, near the start of period 1 (water years
1996-2000). During period 1, streamfiows above the geomet-
ric mean streamfiows generally persisted through water year
1999 and then decreased substantially to below the geometric
mean streamfiows during water year 2000. High streamfiows
were prevalent during most of period 1 and are evident in
annual maximum streamfiows being higher than maximums
of most other years and also in annual minimum streamfiows
being higher than minimums of most other years (fig. 3).
Streamflow during water year 1997 was particularly unusual
in that the receding limb of snowmelt runoff was less abrupt
and less variable than in most years, and post-runoff base
streamfiows generally were above or near the geometric mean
streamflow. Further, the post-runoff base streamilows in water
year '1997 at sampling site 14 (fig. 3B) sometimes exceeded
annual maximum streamfiows during the low streamflow years
2000-2002. During period 2 (water years 2001-5), stream-
flows generally were below the geometric mean streamfiows.
During period 3 (water years 2006-10), streamfiows gradu-
ally increased from below the geometric mean streamfiows
in water year 2006 to above the geometric mean streamfiows
in water year 2010. During period 4 (water years 2011-15),
streamfiows generally were above the geometric mean stream-
flows in water years 2011-12 and then decreased to near the
geometric mean streamfiows in water year 2013. Streamfiows
in water year 2011 were especially high and generally similar
to streamfiows in water year 1997.

Other Factors

Factors relating to data requirements, treatments, and
assessment that affect trend analysis and interpretation of
results include relations between unadjusted concentrations
and FACs, and data transformation. Unadjusted concentrations
are the observed concentrations before flow adjustment.

The FACs are estimates of constituent concentrations
after removing effects of streamflow variability; thus, FACs
ty pically have less variability' than unadjusted concentra-
tions, although the strength of this pattern is variable among
sampling-site and constituent combinations, and also can be
variable through time for a given sampling-site and con-
stituent combination. Time-series streamflow, unfi he red-
recoverable copper, unfiltered-recoverable arsenic, and


-------
Factors that Affect Trend Analysis and Interpretation 25

10,000

A. Silver Bow Creek at Warm Springs (sampling site 8, fig. 1, table 1)

B. Clark Fork at Deer Lodge (sampling site 14, fig. 1, table 1)

1,000 r

S 100 r

10,000 r	r

1,000 r

100

C. Clark Fork at Turah Bridge (sampling site 20, fig. 1, table 1)

EXPLANATION

[Wateryear is defined as the
12-month period from October 1
through September 30 and is
designated by the year in which
it ends]

Daily mean streamflow

Geometric mean streamflow
for water years 1996-2015

Locally weighted scatter plot
smooth (LOWESS; Cleveland,
1985; Cleveland and McGill,
1984) line for daily mean
streamflow

Wateryear (October-September)

Figure 3. Daily mean streamflow for selected sampling sites in the Milltown Reservoir/Clark Fork River Superfund Site in the upper
Clark Fork Basin, Montana, water years 1993-2015. A, Silver Bow Creek at Warm Springs, Montana; 6, Clark Fork at Deer Lodge,
Montana; and C, Clark Fork at Turah Bridge near Bonner, Montana.


-------
26 Water-Quality Trends and Constituent-Transport Analysis for Selected Sampling Sites

suspended-sediment data for Clark Fork near Galen (sampling
site 11) are presented in figure 4 to provide examples for
discussion of relations between unadjusted and flow-adjusted
concentrations.

Similarities among the LOWESS lines for streamflow
(fig. 44) and unadjusted suspended-sediment concentrations
(fig. 4D) illustrate the direct relations between streamflow and
unadjusted suspended-sediment concentrations. Unadjusted
suspended-sediment concentrations tend to be higher during
high streamflow conditions than during low streamflow condi-
tions. During high streamflow conditions, with associated
high hydraulic energy, particulate material is mobilized and
transported in the stream. During low streamflow conditions,
streams have less capacity for transporting particulate materi-
als. Flow-adjustment procedures account for the response of
suspended-sediment concentrations to variations in streamflow
and produce FACs that represent temporal variability in con-
sistent streamflow conditions. In the Clark Fork, suspended-
sediment FACs in high streamflow conditions are less vari-
able and lower than unadjusted concentrations (for example,
fig. 41), water years 1996-99). Suspended-sediment FACs in
low streamflow conditions are less variable and generally cen-
tered within unadjusted concentrations (for example, fig. 4D.
water years 2000-2001).

Unfiltered-recoverable copper has concentration and
streamflow relations that are similar to suspended sediment
because of adsorption on inorganic and organic particulate
materials; these same relations generally apply to other metal-
lic elements. As a result, patterns in unadjusted concentra-
tions and FACs for unfiltered-recoverable copper (fig. 4B) are
similar to those of suspended sediment (fig. 41)).

Arsenic in streams in the upper Clark Fork Basin typi-
cally is mostly in dissolved phase and has less variability
and a weaker direct relation with streamflow than is the case
for metallic elements. Arsenic has been widely dispersed in
the upper Clark Fork Basin as a result of deposition of flue
dust and smelter emissions with resultant large-scale soil and
groundwater contamination (U.S. Environmental Protection
Agency, 2010). Further, arsenic generally is more soluble
than metallic elements in the geochemical conditions that
are prevalent in the upper Clark Fork Basin. These factors
result in high arsenic concentrations in groundwater in some
areas and also mobilization of arsenic to stream channels
for a large range of streamflow conditions. Thus, patterns in
unadjusted concentrations and FACs for unfiltered-recoverable
arsenic (fig. 4C) generally are less variable than for unfiltered-
recoverable copper (fig. 4B) and suspended sediment (fig. 41)).
Also, unadjusted concentrations of unfiltered-recoverable arse-
nic have less correspondence with streamflow than unfiltered-
recoverable copper and suspended sediment.

Similarities among the LOWES S lines for streamflow
(fig. 4/1), unfiltered-recoverable copper (fig. 4B), and sus-
pended sediment (fig. 41)) indicate that temporal variability
in streamflow might confound interpretation of temporal
variability in unadjusted constituent concentrations. Examina-
tion of temporal variability during water years 1993-2015

indicates that, in all cases, the LOWESS lines for stream-
flow (fig 4.4), unfiltered-recoverable copper (fig. 4B), and
suspended sediment (fig. 4D) are highest about 1996-97
and lowest about 2000-2001, then variably increase during
2002-11 and generally decrease during 2012-15. Because
of the strong association between constituent concentrations
and streamflow, interpreting temporal changes in unadjusted
constituent concentrations during specific time periods is dif-
ficult. For example, in water years 2000-2002, mean annual
streamflow was low (about 60 percent of the long-term mean
annual streamflow). Annual mean streamflow in water year
2003 somewhat increased to near-normal conditions (about
90 percent of the long-term mean annual streamflow). Associ-
ated with the increase in streamflow in 2003 were somewhat
abrupt increases in unadjusted concentrations of unfiltered-
recoverable copper and suspended sediment that are reflected
by somewhat abrupt increases in the LOWESS lines for those
constituents. The somewhat abrupt increases in unadjusted
concentrations of unfiltered-recoverable copper and suspended
sediment in water year 2003 probably were affected by the
near-normal streamflow conditions of water year 2003 imme-
diately following the low streamflow conditions of water years
2000-2002. During water years 2000-2002. low streamflow
conditions might have promoted storage of particulate materi-
als in the basin; the stored particulate materials might have
been readily mobilized during water year 2003. Beginning in
water year 2005, streamflow conditions gradually transitioned
from generally low streamflow conditions to high streamflow
conditions in water year 2011. The gradual transition might
have affected the response in unadjusted concentrations of
unfiltered-recoverable copper and suspended sediment to the
high streamflow conditions of water year 2011, particularly
in comparison with the more abrupt increase in streamflow in
water year 2003. Thus, various complexities in concentration
and streamflow relations contribute to difficulties in interpret-
ing temporal patterns in unadjusted constituent concentrations.
Temporal variability in streamflow strongly confounds the
ability to interpret temporal variability in unadjusted constitu-
ent concentrations.

The TSM flow-adjustment procedure analyzes concentra-
tion and streamflow relations on multiple timescales (interan-
iuial. seasonal, and short-term) and accounts for streamflow
variability. In contrast to the LOWESS lines through the
unadjusted constituent concentrations, the TSM-fitted trends
in figure 4 indicate consistent decreases in FACs of unfiltered-
recoverable copper and suspended sediment. The dissimilar
patterns between unadjusted concentrations and FACs indicate
the importance of flow-adjusted trend analysis for identifying
actual patterns in constituent concentrations independent from
variability in streamflow conditions.

An important consideration in interpreting trend results
relates to the trend-analysis methods incorporating log trans-
formation of constituent concentrations. Log transformation
results in datasets that are approximately normally distributed
and allows analysis using rigorous parametric procedures;
however, log transformation decreases variability in the data


-------
Factors that Affect Trend Analysis and Interpretation 27

100

6. Unfiltered-recoverable copper





1,000

100

C. Unfiltered-recoverable arsenic

* > *



1,000

;

D.Suspended sediment







-

~

A

a A A

A

A

-

:

A

-• . .

A
A

A

A A
A

A

-

•

• A

M.



A *

A

. 4 *

If . >	



r



• • x A



a - -		 .

TSso * -

:



«• ' A#

t * * i. .*

a«aa#« A
• A • A |A* A»#

¦ f*i* 'JL,%m

A M • A V

% #^7—f



. -Is

• A

A * A

, * *

; •: «









• *



03

aj	aj

O	,-M

s	s

¦M	Q"

£=	CO

05	p

100

Water year (October-September)

EXPLANATION

[Water year is defined as the
12-month period from October 1
through September 30 and is
designated by the year in which
itends]

	 Daily mean streamflow

— — Geometric mean streamflow
fo r wate r yea rs 1996-2015

Locally weighted scatter plot
smooth (LOWESS;
Cleveland, 1985;

Cleveland and McGill,
1984) line for daily mean
streamflow

•	Flow-adjusted concentration
Flow-adjusted fitted trend

*	Unadjusted (not flow

adjusted) concentration

LOWESS line for unadjusted
concentrations

Figure 4. Selected streamflow and constituent concentration information for Clark Fork near Galen, Montana (sampling site 11), water
years 1993-2015. A, streamflow; 6, unfiltered-recoverable copper; C, unfiltered-recoverable arsenic; and D, suspended sediment.


-------
28 Water-Quality Trends and Constituent-Transport Analysis for Selected Sampling Sites

relative to the original untransformed units representative
of actual environmental variability. In general, the statistical
distributions of constituent concentrations and streainflow (in
original untransformed units) for sampling sites in the upper
Clark Fork Basin are right skewed, indicating that the extent
of data higher than the median is greater than the extent of
data lower than the median. Log transformation results in
expansion of the lower end of the distribution and compres-
sion of the higher end of the distribution. Compression of the
higher end of the distribution has a relatively larger effect than
expansion of the lower end of the distribution. This factor is
important in interpreting trend results with respect to various
regulatory issues, including compliance with human-health or
aquatic-life standards. Trends inFACs represent changes in
central tendency quantified as changes in the geometric mean
in reference to log-transformed streamflow. Thus, the trends
in FACs provide general information on overall temporal
changes (in terms of directions and relative magnitudes) in
concentrations but lack the specificity to indicate compliance
or noncompliance with various regulatory standards. Effects
of data transformation, however, do not negatively affect
the primary purpose of this study in determining temporal
water-quality trends through time and using the trend results
to evaluate relative changes in constituent transport charac-
teristics among sampling sites. In the trend analyses, all data
(high as well as low values) affect changes in FAC geometric
means; thus, the fitted trends appropriately represent unbiased
estimates of overall changes in central tendency.

Water-Quality Trends and Constituent-
Transport Analysis Results

This section of the report presents water-quality trend
and transport-analysis results for selected sampling sites in the
data-summary reaches in the Milltown Reservoir/Clark Fork
River Superfund Site for water years 1996-2015. Results are
presented for all constituents investigated, but emphasis is
placed 011 copper, arsenic, and suspended sediment in the fol-
lowing subsections.

Water-Qua I ity Trends in Flow-Adjusted
Concentrations

For all constituents investigated, detailed results for
trend magnitudes, computed as the total percent changes in
FAC geometric means from the beginning to the end of each
5-year period, are presented in appendix 3 in tables 3-1 (for
most sampling sites) and 3-2 (for Clark Fork above Missoula
[sampling site 22J). Detailed trend results are graphically pre-
sented in figures 3-1 through 3-7 in appendix 3. The detailed
graphical presentations in appendix 3 present fitted trends for

all constituents and allow evaluation of the fitted trends for a
given sampling site in conjunction with FACs.

Fitted trend values (that quantify the temporal changes
in FAC geometric means in terms of concentration units) are
summarized in tables 6 (for most sampling sites) and 7 (for
Clark Fork above Missoula [sampling site 22J) and graphi-
cally summarized in figures 5-10. The summary graphical
presentations in figures 5-10 show side-by-side fitted trends
for the adjacent sampling sites in a given reach and allow
comparisons in temporal patterns between the reach inflow
and outflow; these comparisons facilitate interpretation of the
constituent-transport analysis results.

In this report, qualitative observations are described for
the overall trend magnitude (percent change) from the start of
period 1 to the end of period 4. Overall trend magnitude was
considered to be (1) large, if the absolute value was greater
than about 60 percent; (2) moderate, if the absolute value was
in the range of about 40-60 percent; (3) small, if the absolute
value was in the range of about 20-40 percent; and (4) minor,
if the absolute value was less than about 20 percent.

Trend-magnitude and fitted trend values are considered
semiquantitative estimates determined by complex statistical
analysis. Throughout this report, trend-magnitude and fitted
trend values frequently are mentioned in figures, tables, and
discussion of temporal and spatial changes in water quality
(reported to two significant figures for all constituents except
specific conductance, which is reported to three significant
figures). Reference to specific trend-magnitude and fitted trend
values is intended to facilitate presentation and discussion of
relative spatial and temporal differences between values but is
not intended to represent absolute accuracy at two significant
figures. The /rvalues and levels of significance (a p-value less
than 0.01 is considered statistically significant in this report)
associated with the trend results are indicated in the tables and
figures that present trend results. Significance levels were not
the only factor in evaluating the substance of the trends, but
rather were considered in conjunction with trend directions
and relative magnitudes, and patterns among sites and con-
stituents. In this study, the TSM is considered to be a useful
tool for simplifying the environmental complexity in the upper
Clark Fork Basin to provide a large-scale evaluation of general
temporal changes in FACs and constituent transport indepen-
dent from streamflow variability. Thus, the TSM best-fit trend
lines are considered to provide important information beyond
the strict statistical characteristics of the trend results (in terms
of /^-values and levels of significance) because they aid in
comparing and summarizing large-scale patterns among the
sampling sites. Factors affecting temporal variability in water
quality in the upper Clark Fork Basin are complex. Much
information on changes in water quality is presented herein,
but it is beyond the scope of this report to provide detailed
explanations for all of the changes or to link specific trends
with specific remediation activities.


-------
Water-Quality Trends and Constituent-Transport Analysis Results 29

Table 6. Summary of flow-adjusted trend results for selected sampling sites and constituents, water years 1996-2015.

[Water year is the 12-month period from October 1 through September 30 and is designated by the year in which it ends. Gray shading indicates a statistically
significant (p-value less than 0.01) trend for the trend period before the shaded value, p-value, statistical probability level; jj.S/cm, microsiemen per centimeter at
25 degrees Celsius; ng/L, microgram per liter; mg/L, milligram per liter]

Fitted trend values

Percent change from
start of
period 1

1996	2001	2006	2011	2015	through end of

(start of (start of (start of (start of (end of	period 4

period 1) period 2) period 3) period 4) period 4)

Constituent or property, flow-adjusted Start of Start of Start of Start of	End of

units of measurement	water year water year water year water year water year

Silver Bow Creek at Warm Springs, Montana (sampling site 8, fig. 1, table 1)

Specific conductance, (iS/cm

521

514

501

513

446

-14

Copper, filtered, (ig/L

8.9

4.6

4.1

3.8

2.9

-67

Copper, unfiltered-recoverable, (ig/L

15

9.3

7.9

7.0

5.0

-67

Zinc, unfiltered-recoverable, (ig/L

35

16

8.4

9.8

6.1

-83

Arsenic, filtered, (ig/L

19

19

20

21

17

-11

Arsenic, unfiltered-recoverable, (ig/L

22

22

23

23

19

-14

Suspended sediment, mg/L

5.3

6.3

4.6

2.7

3.1

-42

Clark Fork near Galen, Montana (sampling site 11, fig. 1, table 1)

Specific conductance, (iS/cm

447

454

415

443

388

-13

Copper, filtered, (ig/L

7.6

4.2

4.0

3.3

3.4

-55

Copper, unfiltered-recoverable, (ig/L

15

11

11

11

8.1

-46

Zinc, unfiltered-recoverable, (ig/L

30

13

9.0

12

7.1

-76

Arsenic, filtered, (ig/L

12

11

13

10

11

-8

Arsenic, unfiltered-recoverable, (ig/L

15

14

15

12

14

-7

Suspended sediment, mg/L

5.2

5.8

4.7

5.1

3.8

-27

Clark Fork at Deer Lodge, Montana

(sampling site 14, fig. 1, table 1)



Specific conductance, (iS/cm

479

482

463

454

456

-5

Copper, filtered, (ig/L

6.9

5.8

6.1

5.4

5.8

-16

Copper, unfiltered-recoverable, (ig/L

30

23

24

25

23

-23

Zinc, unfiltered-recoverable, (ig/L

39

24

24

22

19

-51

Arsenic, filtered, (ig/L

11

11

13

11

11

0

Arsenic, unfiltered-recoverable, (ig/L

16

14

15

14

14

-13

Suspended sediment, mg/L

18

15

14

15

12

-33



Clark Fork at Goldcreek, Montana 1

[sampling site 16,fig. 1, table 1)



Specific conductance, (iS/cm

425

418

406

398

398

-6

Copper, filtered, (ig/L

4.8

3.8

4.3

3.8

3.9

-19

Copper, unfiltered-recoverable, (ig/L

19

19

15

14

15

-21

Zinc, unfiltered-recoverable, (ig/L

27

20

13

15

13

-52

Arsenic, filtered, (ig/L

9.4

8.2

8.8

8.6

8.2

-13

Arsenic, unfiltered-recoverable, (ig/L

12

10

10

10

9.7

-19

Suspended sediment, mg/L

15

17

8.3

13

11

-27


-------
30 Water-Quality Trends and Constituent-Transport Analysis for Selected Sampling Sites

Table 6. Summary of flow-adjusted trend results for selected sampling sites and constituents, water years 1996-2015.—Continued

[Water year is the 12-month period from October 1 through September 30 and is designated by the year in which it ends. Gray shading indicates a statistically
significant (p-value less than 0.01) trend for the trend period before the shaded value. /?-value, statistical probability level; jaS/cm, microsiemen per centimeter at
25 degrees Celsius; ju.g/L, microgram per liter; mg/L, milligram per liter]

Fitted trend values

Percent change from

Constituent or property, flow-adjusted
units of measurement

Start of
water year
1996
(start of
period 1)

Start of
water year
2001
(start of
period 2)

Start of
water year
2006
(start of
period 3)

Start of
water year
2011
(start of
period 4)

End of
water year
2015
(end of
period 4)

start of
period 1
through end of
period 41

Clark Fork near Drummond, Montana

(sampling site 18,fig. 1, table 1)





Specific conductance, (iS/cm

461

459

449

434

461

0

Copper, filtered, (ig/L

3.9

3.9

4.3

3.3

3.7

-5

Copper, unfiltered-recoverable, (ig/L

17

15

14

13

12

-29

Zinc, unfiltered-recoverable, (ig/L

36

19

15

17

13

-64

Arsenic, filtered, (ig/L

9.6

9.0

9.4

8.4

8.6

-10

Arsenic, unfiltered-recoverable, (ig/L

12

10

11

10

10

-17

Suspended sediment, mg/L

21

16

13

16

13

-38

Clark ForkatTurah Bridge near Bonner, Montana (sampling site 20,fig. 1, table 1)

Specific conductance, (iS/cm

347



324

334

327

-6

Copper, filtered, (ig/L

3.3

2.5

2.8

2.6

2.1

-36

Copper, unfiltered-recoverable, (ig/L

10

9.0

8.3

8.2

7.9

-21

Zinc, unfiltered-recoverable, (ig/L

21

13

9.2

14

9.7

-54

Arsenic, filtered, (ig/L

5.4

5.1

5.4

5.5

4.7

-13

Arsenic, unfiltered-recoverable, (ig/L

6.8

6.1

6.1

6.6

5.6

-18

Suspended sediment, mg/L

13

12

8.8

12

9.5



Shading represents qualitative observations on overall trend magnitudes (percent change from start of water year 1996 to end of water year 2015) as follows:
no shading—minor (the absolute value was less than about 20 percent); green shading—small (the absolute value was in the range of about 20—40 percent; tan
shading—moderate (the absolute value was in the range of about 40-60 percent; and purple shading—large (the absolute value was greater than about 60 per-
cent).


-------
Water-Quality Trends and Constituent-Transport Analysis Results 31

Table 7. Summary of flow-adjusted trend results for Clark Fork above Missoula, Montana (sampling site 22), for selected constituents,
water years 1996-2015.

[Water year is the 12-month period from October 1 through September 30 and is designated by the year in which it ends. Gray shading indicates a statistically
significant (p-value less than 0.01) trend for the trend period before the shaded value. /?-value, statistical probability level; jaS/cm, microsiemen per centimeter at
25 degrees Celsius; ju.g/L, microgram per liter; mg/L, milligram per liter]

Fitted trend values

Constituent or property,
flow-adjusted units
of measurement

Start of
water year
1996
(start of
period 1)

Start of
water year
2001
(start of
period 2)

Start of
water year
2006
(start of
period 3A)

March 28,2008
(start of
period 3B)

Start of
water year
2011
(start of
period 4)

End of
water year
2015
(end of
period 4)

Percent change

from start of
period 1 through
end of period 41

ClarkForkabove Missoula, Montana (sampling site 22,fig. 1,table 1)

Specific conductance, (iS/cm

277

275

270

273

283

265

-4

Copper, filtered, (ig/L

2.3

1.7

2.1

2.4

1.9

1.4

-39

Copper, unfiltered-recoverable, (ig/L

6.4

4.9

6.9

15

6.3

3.0



Zinc, unfiltered-recoverable, (ig/L

14

7.2

10

30

10

5.0

-64

Arsenic, filtered, (ig/L

3.3

2.8

3.2

3.6

3.4

2.6

-21

Arsenic, unfiltered-recoverable, (ig/L

4.2

3.3

3.9

4.8

4.0

3.0

-29

Suspended sediment, mg/L

7.7

7.4

9.2

25

9.9

6.0

-22

Shading represents qualitative observations on overall trend magnitudes (percent change from start of water year 1996 to end of water year 2015) as follows:
no shading—minor (the absolute value was less than about 20 percent); green shading—small (the absolute value was in the range of about 20—40 percent; tan
shading—moderate (the absolute value was in the range of about 40-60 percent; and purple shading—large (the absolute value was greater than about 60 per-
cent).


-------
32

Water-Quality Trends and Constituent-Transport Analysis for Selected Sampling Sites

Period

.

: a

I I I |

. Unfiltered-recoverable

i i i 1

copper

I I I |

1 1 1 1

' 15









15

11

^		

11

11



	^

		

			 8,1 ~



9.3

7.9

7.0





i i i i

i i i 1

i i i 1

5.0

i

B. Unfiltered-recoverable arsenic

22

22.

23

23

19

EXPLANATION

[Water year is defined as the
12-month period from October 1
through September 30 and is
designated by the year in which
it ends, p-value, statistical
probability level]

— Reach inflow—Silver Bow
Creek at Warm Springs
(sampling site 8, fig. 1,
table 1)

	Reach outflow—Clark Fork

near Galen (sampling site
11, fig. 1, table 1)

15 Fitted trend value at start or
end of period

9.3 Bold values indicate statistical
significance (p-value less
than 0.01) for period before
value presented in bold

C. Suspended sediment

5.3

- 5.2

3.

5.8'

1995

2000	2005

Wateryear (October-September)

2010

2015

Figure 5. Flow-adjusted fitted trends for selected constituents for sampling sites in reach 4, extending from Silver Bow Creek at Warm
Springs, Montana (sampling site 8), to Clark Fork near Galen, Montana (sampling site 11), water years 1996-2015.


-------
Water-Quality Trends and Constituent-Transport Analysis Results 33

Period

100

E 10

1

100

: a

i i i |

. Unfiltered-recoverable

i i i I

copper





0 u

' 15

23

	>

24

25

7

23 '



	11

				

11

\

11

		





i i i I

i i i 1

i i i 1

i i i 1

S E

2? E 10
S <=

1

100

c: os

05 .ti

E —

05 °-

V) V)

-o E

05 TO

J1
S-E 10

V) —

5 .E

c=

¦23 2

V) '-t-J
= !°

5 £
I

6. Unfilteretl-recoverable arsenic

V

14

1995

2000	2005

Water year (October-September)

2010

EXPLANATION

[Water year is defined as the
12-month period from October 1
through September 30 and is
designated by the year in which
it ends, p-value, statistical
probability level]

—	Reach inflow—Clark Fork

nearGalen (sampling site
11,fig. 1, table 1)

—	Reach outflow—Clark Fork at

Deer Lodge (sampling site
14,fig. 1, table 1)

30 Fitted trend value at start or
end of period

11 Bold values indicate statistical
significance (p-value less
than 0.01) for period before
value presented in bold

_

-

. 18

i i i |

.Suspended sediment
	..	15

i i i 1

14

i i i |

15

\

1 1 1 1 -

12

" /

- 5.2

		

	_	 4.7

5,1

\

"		







3.8

2015

Figure 6. Flow-adjusted fitted trends for selected constituents for sampling sites in reach 5, extending from Clark Fork near Galen,
Montana (sampling site 11), to Clark Fork at Deer Lodge, Montana (sampling site 14), water years 1996-2015.


-------
34 Water-Quality Trends and Constituent-Transport Analysis for Selected Sampling Sites

Period

100

E 10

100

S> E

o 'g 10

: a

. Unfiltered-recoverable

copper





30

	_____ 23

24

25
	^

23 '

/

iT7





			^



15

14

15













:

.Unfiltered-recoverable arsenic





- 16

\

14.

15

14

14v

1-/











10'

/

10

o

	\

- ^—1—1—1—1	1	1	1	

CT>

EXPLANATION

[Water year is defined as the
12-month period from October 1
through September 30 and is
designated by the year in which
it ends, p-value, statistical
probability level]

	Reach inflow—Clark Fork at

Deer Lodge (sampling site
14, fig. 1, table 1)

— Reach outflow—Clark Fork at
Goldcreek (sampling site 16,
fig. 1, table 1)

19 Fitted trend value at start or end
of period

8.3 Bold values indicate statistical
significance (p-value less
than 0.01) for period before
value presented in bold

1995

2000	2005

Water year (October-September)

2010

2015

Figure 7. Flow-adjusted fitted trends for selected constituents for sampling sites in reach 6, extending from Clark Fork at Deer Lodge,
Montana (sampling site 14), to Clark Fork at Goldcreek, Montana (sampling site 16), water years 1996-2015.


-------
Water-Quality Trends and Constituent-Transport Analysis Results 35

Period

100

T

AUnfiltered-recoverable copper

10

12 1

100

EXPLANATION

[Water year is defined as the
12-month period from October 1
through September 30 and is
designated by the year in which
it ends, p-value, statistical
probability level]

	Reach inflow—Clark Fork at

Goldcreek (sampling site
16,fig. 1,table 1)

	Reach outflow—Clark Fork

near Drummond (sampling
site 18, fig. 1, table 1)

17 Fitted trend value at start or
end of period

8.3 Bold values indicate

statistical significance
(p-value less than 0.01) for
period before value
presented in bold

S.Unfiltered-recoverable arsenic

10

\

T

10.

9.7

_l	I	I	L

_l	I	I	L_

100

~i	1	1	r

~i	1	1	r

C. Suspended sediment



05 .±±

E —

15 E

05 °-

C/3 V)

-o E

05

.=>
05 =

S-'e

5 .E

c=

£3 o

V)

3 ro

CD

10

1995

2000	2005

Wateryear (October-September)

2010

2015

Figure 8. Flow-adjusted fitted trends for selected constituents for sampling sites in reach 7, extending from Clark Fork at Goldcreek,
Montana (sampling site 16), to Clark Fork near Drummond, Montana (sampling site 18), water years 1996-2015.


-------
36 Water-Quality Trends and Constituent-Transport Analysis for Selected Sampling Sites

100

Period

1 2 3
1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	r

E 10

1

100

AUnfiltered-recoverable copper

9.0

8.3

8.2

7.9

_
:

i i i l i i i

.Unfiltered-recoverable arsenic

1 1

i i i |

1 1 1 1 -

n

\

in

		



11

\

10.

10N











6'8

/

6.1

/

6.1

6.6

7

5.6

> E

o £

O 05
CD O

-TD .2

o e 10

1

100

_

- c

21

i i i |

.Suspended sediment

		 16

I I I |

		 13

I I I |

16

\

1 1 1 1

	 13

_ 13

12

i i i 1

/

8.8

i i i 1

		

	—		 12

i i i 1

		*

9.5 ;

i i i 1

c 0)

05 .±±

E —

15 S

05 °"

V> V)

-o E

cu ro

,E>

05 =

S-'e

5 .E
~o

-M 2
00 '4-1
Z5

10

1995

2000

2005	2010

Water year (October-September)

2015

EXPLANATION

[Water year is defined as the
12-month period from October 1
through September 30 and is
designated by the year in which
it ends, p-value, statistical
probability level]

	Reach inflow—Clark Fork near

Drummond (sampling site
18, fig. 1, table 1)

Reach outflow—Clark Fork at
Turah Bridge (sampling site
20, fig. 1, table 1)

10 Fitted trend value at start or end
of period

10 Bold values indicate statistical
significance (p-value less
than 0.01) for period before
value presented in bold

Figure 9. Flow-adjusted fitted trends for selected constituents for sampling sites in reach 8, extending from Clark Fork near Drummond,
Montana (sampling site 18),to Clark Fork atTurah Bridge near Bonner, Montana (sampling site 20), wateryears 1996-2015.


-------
Water-Quality Trends and Constituent-Transport Analysis Results 37

Period

100

_

: a

I I I |

.Unfiltered-recoverable

I I I |

copper

i i ! i |

1
1
1
1
1
1

15 |

' ' ' l .

10

	 9.0

8.3

/ I \ 82

7.9 "

~



			

i i i 1

>



- 6.4

4.9

i i i 1

"eo

CD

3.0

i

100

S.Unfiltered-recoverable arsenic

6.8

6.1

6.1

4.2

3.3

3.9

EXPLANATION

[Water year is defined as the
12-month period from October 1
through September 30 and is
designated by the year in which
it ends, p-value, statistical
probability level]

	Reach inflow—Clark Fork near

Turah Bridge (sampling site
20, fig. 1, table 1)

— Reach outflow—Clark Fork
above Missoula (sampling
site 22, fig. 1, table 1)

6 4 Fitted trend value at start or
end of period

25 Bold values indicate statistical
significance (p-value less
than 0.01) for period before
value presented in bold

1995

2000	2005

Water year (October-September)

2010

2015

Figure 10. Flow-adjusted fitted trends for selected constituents for sampling sites in reach 9, extending from Clark Fork at Turah Bridge
near Bonner, Montana (sampling site 20), to Clark Fork above Missoula, Montana (sampling site 22), water years 1996-2015.


-------
38 Water-Quality Trends and Constituent-Transport Analysis for Selected Sampling Sites

Copper

Trend results indicate that FACs of unfiltered-recoverable
copper decreased at tlie sampling sites from the start of period
1 through the end of period 4 (tables 6 and 7); the decreases
ranged from large for one sampling site (Silver Bow Creek at
Warm Springs [sampling site 8]) to moderate for two sampling
sites (Clark Fork near Galen [sampling site 11] and Clark Fork
above Missoula [sampling site 22]) to small for four sampling
sites (Clark Fork at Deer Lodge [sampling site 14], Clark Fork
at Goldcrcck [sampling site 16], Clark Fork near Druinmond
[sampling site 18], and Clark Fork at Turah Bridge [sampling
site 20]). For period 4 (water years 2011-15), the most notable
changes indicated for the Milltown Reservoir/Clark Fork River
Superfund Site in the upper Clark Fork Basin were statistically
significant decreases in FACs of unfiltered-recoverable copper
for sampling sites 8 and 22. For all other sampling sites, the
period 4 changes in FACs of unfiltered-recoverable copper
were not statistically significant.

Arsenic

Trend results indicate that FACs of unfiltered-recoverable
arsenic decreased at the sampling sites from the start of
period 1 through the end of period 4 (tables 6 and 7); the
decreases ranged from minor for six sampling sites (sampling
sites 8-20) to small for one sampling site (sampling site 22).
For period 4 (water years 2011-15), the most notable changes
indicated for the Milltown Reservoir/Clark Fork River
Superfund Site in the upper Clark Fork Basin were statisti-
cally significant decreases in FACs of unfiltered-recoverable
arsenic for sampling site 8 and near statistically significant
decreases for sampling site 22; the /rvalue (0.012) for the
period 4 decrease for sampling site 22 is not statistically sig-
nificant but is only slightly larger than the selected alpha level
(0.01 in this report). For all other sampling sites, the period 4
changes in FACs of unfiltered-recoverable arsenic were not
statistically significant.

Suspended Sediment

Trend results indicate that FACs of suspended sedi-
ment decreased at the sampling sites from the start of period
1 through the end of period 4 (tables 6 and 7); the decreases
ranged from moderate for one sampling site (sampling site 8)
to small for six sampling sites (sampling sites 11-22). For
period 4 (water years 2011-15), the changes in FACs of
suspended sediment were not statistically significant for any
sampling sites.

Overview of Water-Quality Trend Results

The most notable changes in water quality in period 4
were indicated for Silver Bow Creek at Warm Springs (sam-
pling site 8; reach 4 inflow) and Clark Fork above Missoula

(sampling 22; reach 9 outflow). Trend results for sampling
site 8 indicated more substantial changes than most oilier sam-
pling sites; the decreases in specific conductance, unfiltered-
recoverable copper, unfiltered-recoverable zinc, and unfiltered-
recoverable arsenic were statistically significant (fig. 5 and
3-1; tables 6 and 3-1). The most extensive remediation
activities in the upper Clark Fork Basin have been conducted
in the Silver Bow Creek Basin upstream from the reach 4
inflow (sampling site 8). Sando and others (2014) noted that
among the most notable changes indicated in the upper Clark
Fork Basin during water years 1996-2010 were moderate to
large decreases in FACs and loads of copper and suspended
sediment in Silver Bow Creek upstream from Warm Springs.
The period 4 (water years 2011-15) statistically significant-
decreases in FACs of unfiltered-recoverable copper and zinc
provide indication that FACs of metallic contaminants contin-
ued to substantially decline at sampling site 8.

The removal of the former Milltown Dam, which was
located between Clark Fork at Turah Bridge (sampling site 20;
reach 9 inflow) and Clark Fork above Missoula (sampling
site 22; reach 9 outflow), in 2008 was an important reme-
diation activity in the upper Clark Fork Basin and strongly
affected water-quality trends and transport characteristics
within reach 9. As such, detailed discussion of trends is pre-
sented for reach 9. During periods 1 and 2, the former Mill-
town Dam was in place, and large amounts of contaminated
sediments were retained in the former Milltown Reservoir in
reach 9; however, the contaminated sediments largely were
unavailable for mobilization and transport because of back-
water effects of the former Milltown Dam (Sando and Lamb-
ing, 2011). Remediation activities preparing for the removal of
the former Milltown Dam started in period 2 but were focused
early in period 3 and included physical removal of large
amounts of contaminated sediments; however, substantial
amounts of contaminated sediments still remained in the Clark
Fork channel and flood plain in reach 9. With the removal of
the former Milltown Dam in 2008, the remaining contami-
nated sediments in reach 9 became more available for mobi-
lization and transport than before the dam removal. Because
of the substantial effect of the intentional breach of Milltown
Dam on March 28, 2008, for sampling site 22, period 3 was
subdivided into period 3 A (October 1, 2005-March 27, 2008)
and period 3B (March 28, 2008-September 30, 2010).

A statistically significant increase in FACs of unfiltered-
recoverable copper is indicated for period 3 A for sampling
site 22 (117 percent, from 6.9 to 15 jxg/L; table 7). The
temporary increase in FACs is associated with activities that
prepared for the removal of the Milltown Dam, including
construction of roads and facilities, reservoir level drawdowns,
and physical removal of large amounts of contaminated
sediments, which likely increased mobilization of sediments
enriched in trace elements (Sando and Lambing, 2011). After
the intentional breach, statistically significant decreases were
indicated for unfiltered-recoverable copper for period 3B
(-58 percent, from 15 to 6.3 jxg/L) and period 4 (-52 percent,
from 6.3 to 3.0 ng/L). For unfiltered-recoverable arsenic, an


-------
Water-Quality Trends and Constituent-Transport Analysis Results 33

increase in FACs is indicated for period 3A (23 percent, from
3.9 to 4.8 |ig/L). After the intentional breach, a decrease is
indicated for unfiltered-recoverable arsenic for period 3B
(-17 percent, from 4.8 to 4.0 (ig/L) and a near statistically
significant decrease is indicated for period 4 (-25 percent,
from 4.0 to 3.0 jig/L; /rvalue of 0.012). For suspended
sediment, a statistically significant increase is indicated for
period 3 A (172 percent, from 9.2 to 25 mg/L). After the
intentional breach, a statistically significant decrease for
suspended sediment is indicated for period 3B (-60 percent,
from 25 to 9.9 mg/L), and a decrease is indicated for period 4
(-39 percent, from 9.9 to 6.0 mg/L). For period 4 (water years
2011-15), trend results for the reach 9 outflow (sampling
site 22) indicate more substantial changes than most other
sampling sites; decreases in unfiltered-recoverable copper,
unfiltered-recoverable zinc, and filtered arsenic were statisti-
cally significant. The p-value (0.012) for the period 4 decrease
in FACs of unfiltered-recoverable arsenic for sampling site 22
is not statistically significant but is only slightly larger than the
selected alpha level (0.01 in this report).

The somewhat high streamflow conditions of period 4
promoted mobilization of trace-element contaminants from
the former Milltown Reservoir, thus decreasing within-reach
source materials and resulting in lower FACs. The substan-
tial decreases in FACs of unfiltered-recoverable copper for
period 3B continued in period 4. Comparison of the period 4
fitted trends for unfiltered-recoverable copper betw een the
reach 9 inflow (sampling site 20) and the reach 9 outflow
(sampling site 22) indicates large deviation from the start of
to the end of period 4 (fig. 10.4) and provides evidence of
continued effects of the removal of the former Milltown Dam.
Deviations in fitted trends between the period 4 reach inflow
and reach outflow also are apparent for unfiltered-recoverable
arsenic (fig. lOfi) and suspended sediment (fig. 10C); however,
the deviations are not as strong for those constituents as for
unfiltered-recoverable copper.

Constituent-Transport Analysis Results

Estimated normalized loads are presented in the frame-
work of a transport analysis to assess the temporal trends in
FACs in the context of sources and transport. Drainage area
and streamflow information relevant to the transport analysis
are presented in table 8. Balance calculations for the trans-
port analysis (that is, differences between reach inflows and
reach outflows) arc presented in tables 4-1 through 4-6 for
reaches 4-9, respectively, in appendix 4. The transport bal-
ance calculations indicate within-reach changes in estimated
normalized loads and allow assessment of temporal changes
in relative contributions from upstream source areas to loads
transported past each reach outflow.

Hydrologic characteristics of the source areas (geo-
metric mean streamflow; table 8) and balance results for
the transport analysis arc illustrated by using pie charts that
show source-area information and load contributions to reach
outflow. Pie charts illustrating temporal patterns in estimated

normalized loads for all data-summary reaches are presented
in figures 11—13 for unfiltered-recoverable copper, unfiltered-
recoverable arsenic, and suspended sediment, respectively.
The pie charts provide a side-bv-side graphical summary
for evaluating spatial and temporal variability in constituent
transport relative to streamflow contributions in the Milltown
Reservoir/Clark Fork River Superfund Site in the upper Clark
Fork Basin. The estimated normalized loads (hereinafter
referred to as "loads") do not represent actual magnitudes
of total mass transport, but rather provide information on
relative temporal changes in constituent transport character-
istics in the upper Clark Fork Basin quantified with respect to
near-median conditions.

In figures 11-13, geometric mean strcamflows (water
years 1996-2015) for each reach are shown across the top
of each figure, with the size (area) of each pie chart being
proportional to the geometric mean streamflow for Clark
Fork above Missoula (sampling site 22; reach 9 outflow). Pie
charts that illustrate the constituent-transport analysis results
for each reach for periods 1-4 are shown below the pie charts
representing geometric mean streamflows. Pic charts illus-
trating loads are sized proportionally to the period 1 reach 9
outflow load. The period 1 reach 9 outflow load was selected
as an index for sizing the pie charts because it represents the
total load transported from the Milltown Reservoir/Clark Fork
River Superfund Site somewhat near the start of remedia-
tion activities. As such, the period 1 reach 9 outflow load is a
useful index in evaluating effects of remediation in the upper
Clark Fork Basin.

Figure 11 presents pie charts representing loads for
unfiltered-recoverable copper and serves as an example
for explaining the presentation of the constituent-transport
analysis results. The size (area) of each loads pic chart rep-
resents the total outflow from the reach, with colored areas
indicating relative contributions from each of the two source
areas; that is, (1) the reach inflow and (2) the intervening
drainage between the reach inflow and outflow (or within-
reach sources). The left-hand column of the load pie charts
presents results for reach 4 for periods 1-4. The period 1
load transported past the reach 4 outflow (sampling site 11)
is 3.7 kilograms per day (kg/d), which is 13 percent of the
period 1 load transported past the reach 9 outflow (29 kg/d
at sampling site 22 shown in right-hand column); thus, the
size of the period 1 reach 4 pie chart is 13 percent of the size
of the period 1 reach 9 pic chart. The blue-colored part of
the period 1 reach 4 pie chart represents the load (1.9 kg/d)
transported past the reach 4 inflow (sampling site 8). The
orange-colored part of the period 1 reach 4 pie chart represents
the total within-reach change in load (that is, net mobilization
from all within-reach sources including groundwater inflow,
tributaries, the main-stem channel, and flood plain). The total
within-reach change in load (1.8 kg/d) was calculated by
subtracting the reach inflow (1.9 kg/d) from the reach out-
flow (3.7 kg/d). In figure 11, results for reach 9 are not shown
for period 3 because of effects of the removal of the former


-------
40 Water-Quality Trends and Constituent-Transport Analysis for Selected Sampling Sites

Table 8. Drainage area and streamflow information relevant to the transport analysis for data-summary reaches in the Milltown
Reservoir/Clark Fork River Superfund Site in the upper Clark Fork Basin, Montana, water years 1996-2015.

[Water year is the 12-month period from October 1 through September 30 and is designated by the year in which it ends, ff/s, cubic foot per second]

Abbreviated sampling site name (table 1)
and number or summation category

Drainage area,
in square miles

Geometric mean
streamflow,
water years

1996-2015,
in ff/s

Reach 4

[extending about 2 river miles from Silver Bow Creek at Warm Springs (sampling site 8, fig. 1, table 1)
to Clark Fork near Galen (sampling site 11, fig. 1, table 1)]

Inflow

Silver Bow Creek at Warm Springs (sampling site 8)

Outflow

Clark Fork near Galen (sampling site 11)

Within-reach change—outflow (sampling site 11) minus inflow (sampling site 8)

(contributions from all within-reach sources, includiim "roundwater inflow and tributaries)

473

651

178

64

118

:>4

Reach 5

[extending about 21 river miles from Clark Fork near Galen (sampling site 11, fig. 1, table 1)
to Clark Fork at Deer Lodge (sampling site 14, fig. 1, table 1)]

651

Inflow

Clark Fork near Galen (sampling site 11)

Outflow

Clark Fotk al Deer Lodge (sampling site 14)

Within-reach change—outflow (sampling site 14) minus inflow (sampling site 11)

(contributions from all within-reach sources, includiim "roundwater inflow and tributaries)

344

118

208

90

Reach 6

[extending about 26 river miles from Clark Fork at Deer Lodge (sampling site 14, fig. 1, table 1)
to Clark Fork at Goldcreek (sampling site 16, fig. 1, table 1)]

Inflow

Clark Fork at Deer Lodge (sampling site 14)

Outflow

Clark Fork at Goldcreek (sampling site 16)

Within-reach change—outflow (sampling site 16) minus inflow (sampling site 14)

(contributions from all within-reach sources, including groundwater inflow and tributaries)

995

1,704

709

208

406

198

Milltown Dam and difficulties in presenting those results in
conjunction with results for other reaches.

Constituent-transport analysis results are described for
copper, arsenic, and suspended sediment in the following
subsections. Observations are made comparing the relative
proportions of within-reach contributions of constituent loads
and within-reach contributions of streamflow. Those propor-
tional comparisons indicate the importance of a given reach
as a source of constituent loading to Silver Bow Creek or the

Clark Fork. If the contribution of a constituent from within a
reach is proportionally much larger than the contribution of

streamflow from within a reach, the given reach is indicated to
be an important disproportionate source of constituent loading.
Conversely, if the contribution of a constituent from within a
reach is proportionally smaller than or similar to the contribu-
tion of streamflow from within a reach, the given reach is not
indicated to be an important disproportionate source of constit-
uent loading and generally acts as a flow-through reach.


-------
Water-Quality Trends and Constituent-Transport Analysis Results 41

Table 8. Drainage area and streamflow information relevant to the transport analysis for data-summary reaches in the Milltown
Reservoir/Clark Fork River Superfund Site in the upper Clark Fork Basin, Montana, water years 1996-2015.—Continued

[Water year is the 12-month period from October 1 through September 30 and is designated by the year in which it ends. ft3/s, cubic foot per second]

Abbreviated sampling site name (table 1)
and number or summation category

Drainage area,
in square miles

Geometric mean
streamflow,
water years

1996-2015,
in ff/s

Reach 7

[extending about 31 river miles from Clark Fork at Goldcreek (sampling site 16,
to Clark Fork near Drummond (sampling site 18, fig. 1, table 1)]

fig. 1, table 1)



Inflow

Clark Fork at Goldcreek (sampling site 16)

1,704

406

Outflow

Clark Fork near Drummond (sampling site 18)

2,501

589

Within-reach change—outflow (sampling site 18) minus billow (sampling site 16)

(contributions from all within-reach sources, including groundwater inflow and tributaries)

797

183

Reach 8

[extending about 34 river miles from Clark Fork near Drummond (sampling site 18, fig. 1, table 1)
to Clark Fork at Turah Bridge (sampling site 20, fig. 1, table 1)]

Inflow

Clark Fork near Drummond (sampling site 18)

2,501

589

Outflow

Clark Fork at Turah Bridge (sampling site 20)

3,641

1,060

Within-reach change—outflow (sampling site 2.0) minus inflow (sampling site 18)

(contributions from all within-reach sources, including groundwater inllow and tributaries)

1.140

470

Reach 9

[extending about 9 river miles from Clark Fork at Turah Bridge (sampling site 20, fig. 1, table 1)
to Clark Fork above Missoula (sampling site 22, fig. 1, table 1)]

Inflow

Clark Fork at Turah Bridge (sampling site 20)

3,641

1,060

Outflow

Clark Fork above Missoula (sampling site 22)

5,999

2,100

Within-reach change—outflow (sampling site 22.) minus inllow (sampling site 20)

(contributions from all within-reach sources, including groundwater inllow and tributaries)

2.358

1,04!)


-------
Reach 4

118

©

Reach 5

208

&

118
90

Geometric mean streamflow, in cubic feet per second (water years 1996-2015)

406

Reach 6

Reach 7

589

1,060

Reach 8
©

Reach 9

2,100

a>
CD
¦

O
c

Q>

Estimated normalized unfiltered-recoverable copper load,1 in kilograms per day (kg/d)

Wateryears /TiP
1996-2000 o7U-
(period 1)	v 1

Wateryears
2001-5 3.1
(period 2)

©

Wateryears
2006-10
(period 3)

Wateryears
2011-15
(period 4)

3.2

.2.0 >

27

0.94

12

13

12

m 29

Reach 9 pie chartfor
period 3 notshown
because of effects of
removal of theformer
Milltown Dam and
difficulties in
presentation in
conjunction with
results for other
reaches.

23

Reach inflow = 21 kg/d
Within-reach change = 32 kg/d
Reach outlfow = 54 kg/d

Estimated normalized load calculated by
multiplying the mean annualfitted trend
concentration (determined by using the
time-series model)forthe indicated period
times the geometric mean streamflowfor
wateryears 1996-2015 and a units conversion
factor. Calculation of the estimated normalized
constituent load is described in detail in the
section of this report"Estimation of Normalized
Constituent Loads." Loads are reported to two
significantfigures; however, beforefinal
rounding, calculations used three significant
figures when necessary. As a result, some of
the load values have minor rounding artifacts.

Circular pie chart
represents geometric
mean streamflow
at reach outflow

Size (area) of circularpie
charts is proportional
to geometric mean
streamflowforClark Fork
above Missoula, Montana
(sampling site 22)

EXPLANATION

[streamflow pie charts]

Geometric mean
streamflow, in cubic
feetpersecond

Ail within-reach sources

Value-

Geometric
mean streamflow,
in cubicfeet per second
at reach outflow

«/>
a>

o
o

a>
3

OO

¦o
o
3.

V)

-------
Geometric mean streamflow, in cubic feet per second (water years 1996-2015)

Reach 4

118

©

Water years / 34
1996-2000 4.2
(period 1)

Wateryears / r> c
2001-5 4.2
(period 2)

0.78

0.70

Wateryears / 36
2006-10 3.9
(period 3)	u,d1

Wateryears 00
2011-15
(period 4)

0.46

Reach 5

Reach 6

Reach 7

Reach 8

Reach 9

208

&¦

118
90

406

589

1,060

2,100

Estimated normalized unfiltered-recoverable arsenic load,1 in kilograms per day (kg/d)

10

10

9.9

19

18

22

18

Estimated normalized load calculated by
multiplying the mean annualfitted trend
concentration (determined by using the
time-series model)forthe indicated period
times the geometric mean streamflowfor
wateryears 1996-2015 and a units conversion
factor. Calculation of the estimated normalized
constituent load is described in detail in the
section of this report"Estimation of Normalized
Constituent Loads." Loads are reported to two
significantfigures; however, beforefinal
rounding, calculations used three significant
figures when necessary. As a result, some of
the load values have minor rounding artifacts.

Circular pie chart
represents geometric
mean streamflow
at reach outflow

Size (area) of circularpie
charts is proportional
to geometric mean
streamflowfor Clark Fork
above Missoula, Montana
(sampling site 22)

EXPLANATION

[streamflow pie charts]

Geometric mean
streamflow, in cubic
feetpersecond

All within-reach sources

Reach inflow

Value—

Geometric
mean streamflow,
in cubicfeet per second
at reach outflow

Circular pie charts
representloads
transported past
reach outflow

Size (area) of circular pie charts
is proportionalto period 1
estimated normalized load for
Clark Fork above Missoula, Montana
(sampling site 22)

EXPLANATION

[estimated normalized load pie charts]

Reach inflow

Value —Load
transported
past reach outflow'
in kilograms per day

Estimated
normalized load,
in kilograms per day

Total within-reach change in
load (netmobilizationfrom all
within-reach sources)

a>
CD
¦

O
c

a>

a>

o
o

a»
=

(A

¦o

o
3.

V)
'

CD

CO

Figure 12. Pie charts representing geometric mean streamflow and estimated normalized unfiltered-recoverable arsenic loads contributed from reach inflow
and within-reach sources for data-summary reaches for selected periods.

CO


-------
Reach 4

118

&

Wateryears
1996-2000 1,600 U-=f_
(period 1)	^

920
670

Wateryears

2001"5 1.500 0=

(period 2)

850
670

Wateryears	/—»—•

2006-10 1,400 N-3_

(period 3)

570
860

Wateryears	^-v-460

?011.-j5 1.300 ©C
(period 4)	^^^8

"820

Reach 5

Geometric mean streamflow, in cubic feet per second (water years 1996-2015)

Reach 6	Reach 7	Reach 8

208

&

118
90

406

589

1,060

Estimated normalized suspended-sediment load,1 in kilograms per day (kg/d)



1,600

1,300 ( L—X	16,000

\ 6,700/

26,000

7,200

1,500	( 7,200

12,000

21,000

33,000

26,000

7,200

\5,800/

10,000

21,000

27,000

6,800

1,300

12,000

21,000

28,000

Reach 9

2,100

39,000

42,000

Reach 9 pie chartfor Reach inflow= 27,000 kg/d
period 3 notshown

because of effects of Within-reach change = 56,000 kg/d
removal of theformer

Milltown Dam and Reach outlfow = 83,000 kg/d

40,000

Estimated normalized load calculated by
multiplying the mean annualfitted trend
concentration (determined by using the
time-series model)forthe indicated period
times the geometric mean streamflowfor
wateryears 1996-2015 and a units conversion
factor. Calculation of the estimated normalized
constituent load is described in detail in the
section of this report"Estimation of Normalized
Constituent Loads." Loads are reported to two
significantfigures; however, beforefinal
rounding, calculations used three significant
figures when necessary. As a result, some of
the load values have minor rounding artifacts.

Circular pie chart
represents geometric
mean streamflow
at reach outflow

Size (area) of circularpie
charts is proportional
to geometric mean
streamflowforClark Fork
above Missoula, Montana
(sampling site 22)

EXPLANATION

[streamflow pie charts]

/Geometric mean
streamflow, in cubic
feetpersecond

All within-reach sources

Value-

Geometric
mean streamflow,
in cubicfeet per second
at reach outflow

a>
CD
¦

O
c

a>

«/>
a>

o
o

a»
3

OO

¦o
o
3.


-------
Water-Quality Trends and Constituent-Transport Analysis Results 45

Copper

The transport-analysis results indicate that outflow loads
of unfi 1 tered-recoverable copper decreased from the center of
period 1 through the center of period 4 for all reaches (fig. 11).
The largest decrease was for the reach 4 outflow load (about
-27 percent, from 3.7 to 2.7 kg/d). The decrease in the reach
4 outflow load (sampling site 11) largely was because of a
substantial decrease (-50 percent, from 1.9 to 0.94 kg/d) in the
reach 4 inflow load (sampling site 8), with little change indi-
cated for witliin-reach sources. The smallest decrease was for
the reach 5 outflow load (about -8 percent from 13 to 12 kg/d).
Decreases in outflow loads for the other reaches (reaches 6-9)
ranged from about -16 to -25 percent.

Contributions of unfiltered-recoverable copper from reach
4 sources were proportionally similar to or slightly larger
than streamllow contributions from within reach 4 (fig. 11,
tables 8 and 4-1) for all periods, and thus reach 4 is somewhat
indicated to be a disproportionate source of copper loading.
However, the period 4 net mobilization from sources within
reach 4(1.8 kg/d) was only about 8 percent of the period 4
reach 9 outflow load (Clark Fork above Missoula, sampling
site 22; 23 kg/d). Contributions of unfi 1 tered-recoverable cop-
per from reach 5 sources were proportionally much larger than
streamllow contributions from within reach 5 for all periods;
the period 4 net mobilization from sources within reach 5
(9.4 kg/d) accounted for a substantial part (about 41 percent)
of the period 4 reach 9 outflow load. Thus, reach 5 is indicated
to be an important disproportionate source of copper loading.
Contributions of unfiltcrcd-recovcrable copper from sources
within the other reaches (reaches 6-9) were proportionally
smaller than the within-reach streamllow contributions.

The removal of the former Mill town Dam in 2008 war-
rants more detailed discussion of transport analysis results
for reach 9. The segregation of period 3 into periods 3 A and
3B for the reach 9 outflow (sampling site 22) is not directly
incorporated into the transport analysis for reach 9; thus, the
transport-analysis balance calculations for period 3 reflect
the net changes in transport characteristics before and after
the removal of the former Milltown Dam. For unfiltered-
recoverable copper (fig. 11), the reach 9 outflow load (sam-
pling site 22) decreased by about 21 percent from the center
of period 1 (29 kg/d) to the center of period 4 (23 kg/d). Net
mobilization from sources within reach 9 increased between
periods 1 and 2 and also between periods 2 and 3 (fig. 11).
Net mobilization from sources within reach 9 substantially
decreased between periods 3 and 4. Net mobilization from
sources within reach 9 were proportionally larger than
streamllow contributions from within reach 9 for period 3 but
were proportionally smaller than strcamflow contributions
for the other periods. Net mobilization from sources within
reach 9 were smaller for period 4 (2.2 kg/d) than for period 1
(3.7 kg/d).

Arsenic

The transport-analysis results indicate that outflow loads
of unfi 1 tered-recoverable arsenic decreased from the center of
period 1 through the center of period 4 for all reaches (fig. 12).
Decreases in outflow loads for the reaches ranged from about
-5 to -12 percent. Temporal decreases in unfiltcrcd-recovcrable
arsenic were smaller than copper and suspended sediment,
which probably reflects the dispersion and solubility character-
istics of arsenic.

At the upstream end of the Milltown Reservoir/Clark
Fork River Superfund site, the reach 4 inflow load is a
disproportionate source of arsenic loading, with the inflow
load being proportionally larger than the strcamflow (fig. 12,
tables 8 and 4-1). Contributions of unfiltered-recoverable
arsenic from reach 4 sources were proportionally smaller
than streamllow contributions from within reach 4 for all
periods. Downstream from reach 4, contributions of unfiltered-
recoverable arsenic from sources within reaches 5 and 7 were
proportionally similar to witliin-reach streamllow contribu-
tions. Contributions of unfiltered-recoverable arsenic from
sources within the other reaches (reaches 6, 8, and 9) were
proportionally smaller than the within-reach streamllow
contributions.

For unfiltered-recoverable arsenic (fig. 12), the reach 9
outflow load (sampling site 22) decreased by about 5 percent
from the center of period 1(19 kg/d) to the center of period 4
(18 kg/d). Net mobilization from sources within reach 9
increased between periods 2 and 3 (fig. 12). Net mobilization
from sources within reach 9 substantially decreased between
periods 3 and 4. Contributions of unfiltered-recoverable
arsenic from reach 9 sources were proportionally smaller than
streamllow contributions from within reach 9 for all periods.
Net mobilization from sources within reach 9 were slightly
smaller for period 4 (2.1 kg/d) than for period 1 (2.5 kg/d).

Suspended Sediment

The transport-analysis results indicate that outflow loads
of suspended sediment decreased from the center of period
1 through the center of period 4 for reaches 4-8 but slightly-
increased for reach 9 (fig. 13). Decreases in outflow loads for
reaches 6-8 ranged from about -15 to -25 percent.

Contributions of suspended sediment from reach 4
sources were proportionally similar to or slightly larger than
streamllow contributions from within reach 4 (fig. 13, tables 8
and 4-1) for all periods, and thus, reach 4 is somewhat indi-
cated to be a disproportionate source of suspended-sediment
loading. However, the period 4 net mobilization from sources
within reach 4 (820 kg/d) was only about 2 percent of the
period 4 reach 9 outflow load (Clark Fork above Missoula,
sampling site 22; 40,000 kg/d). Contributions of suspended
sediment from reach 5 sources were proportionally much
larger than strcamflow contributions from within reach 5;
the period 4 net mobilization from sources within reach 5
(5,500 kg/d) accounted for about 14 percent of the period 4


-------
46 Water-Quality Trends and Constituent-Transport Analysis for Selected Sampling Sites

reach 9 outflow load. Thus, reach 5 is indicated to be a dispro-
portionate source of suspended-sediment loading. Downstream
from reach 5, contributions of sediment from sources within
reach 7 were proportionally similar to within-reach stream-
flow contributions; the period 4 net mobilization from sources
within reach 7 (9,100 kg/d) accounted for about 23 percent of
the period 4 reach 9 outflow load. Contributions of suspended
sediment from sources within the other reaches (reaches 6,
8, and 9) were proportionally smaller than the within-reach
streamflow contributions.

For suspended sediment (fig. 13), the reach 9 outflow
load (sampling site 22) increased by about 3 percent from
the center of period 1 (39,000 kg/d) to the center of period 4
(40,000 kg/d). Net mobilization from sources within reach 9
increased between periods 1 and 2 and also between periods 2
and 3 (fig. 13). Net mobilization from sources within reach 9
substantially decreased between periods 3 and 4. Net mobili-
zation from sources within reach 9 was proportionally larger
than streamflow contributions from within reach 9 for period

3	but was proportionally smaller than streamflow contribu-
tions for the other periods. Net mobilization from sources
within reach 9 were larger for period 4 (12,000 kg/d) than
for period 1 (6,000 kg/d). The increase in net mobilization
of suspended sediment from sources within reach 9 between
periods 1 and 4 is in contrast to decreases in net mobilization
of unfiltered-recoverable copper and arsenic between periods 1
and 4. A possible explanation for this pattern might relate to
flood-plain disturbance and placement of uncontaminated fill
in the flood plain associated with remediation activities. The
artificially installed uncontaminated fill might be more avail-
able for mobilization than sediment within the former Mill-
town Reservoir during period 1.

Overview of Constituent-Transport Analysis
Results

At the upstream end of the Milltown Reservoir/Clark
Fork River Superfund site, the reach 4 inflow had substan-
tial decreases from the center of period 1 to the center of
period 4 in unfiltered-recoverable copper and suspended-
sediment loads (about -50 percent for both constituents), but
the reach 4 inflow accounts for small parts of the streamflow
(about 3 percent), unfiltered-recoverable copper load (about

4	percent), and suspended-sediment load (about 1 percent) of
the reach 9 outflow in period 4 (figs. 11 and 13). The reach 4
inflow is a disproportionate source of unfiltered-recoverable
arsenic and accounts for about 18 percent of the reach 9
outflow load in period 4 (fig. 12). Some downstream reaches
(including reaches 5 and 7) have within-reach contributions of
unfiltered-recoverable arsenic that are proportionally similar
to streamflow contributions and also substantially contribute
to the reach 9 outflow load. For all reaches, temporal changes
for unfiltered-recoverable arsenic loads are smaller than for
unfiltered-recoverable copper and suspended-sediment loads.

Reach 5 is a large source of unfiltered-recoverable copper
and suspended sediment, which strongly affects downstream
transport of those constituents (figs. 11 and 13). Mobilization
of unfiltered-recoverable copper and suspended sediment from
flood-plain tailings and the streambed of the Clark Fork and its
tributaries within reach 5 results in a contribution of those con-
stituents from within reach 5 that is proportionally much larger
than the contribution of streamflow from within reach 5. In
reach 5, unfiltered-recoverable copper loads in the Clark Fork
increased by a factor of about 4 and suspended-sediment loads
increased by a factor of about 5, whereas streamflow increased
by a factor of slightly less than 2 (fig. 11). For period 4 (water
years 2011-15), unfiltered-recoverable copper and suspended-
sediment loads sourced from within reach 5 accounted for
about 41 and 14 percent, respectively, of the loads at Clark
Fork above Missoula (sampling site 22), whereas streamflow
sourced from within the reach accounted for about 4 percent
of the streamflow at sampling site 22. During water years
1996-2015, decreases in unfiltered-recoverable copper and
suspended-sediment loads (fig. 11 and 13) for the reach 5
outflow and for sources within reach 5 generally were propor-
tionally smaller than for most other reaches.

For the reaches downstream from reach 5 (reaches 6-8),
contributions of copper loads sourced from within the reaches
were proportionally smaller than contributions of streamflow
sourced from within the reaches (fig. 11); thus, the lower
reaches contributed proportionally much less than reach 5
to unfiltered-recoverable copper loading in the Clark Fork.
Although substantial decreases in unfiltered-recoverable
copper and suspended-sediment loads were indicated for the
reach 4 inflow (sampling site 8), those substantial decreases
were not translated to the downstream reaches (reaches 5-8).
The effect of reach 5 as a large source of unfiltered-
recoverable copper and suspended sediment, in combination
with little temporal change in those constituents for the reach 5
outflow, contributes to this pattern.

For unfiltered-recoverable copper, unfiltered-recoverable
arsenic, and suspended sediment, contributions from within
reach 8 generally increased between periods 2 and 4; this
pattern is in contrast to patterns for most other reaches. A pos-
sible explanation for this pattern might relate to effects of the
removal of the former Milltown Dam during period 3. Before
the removal of the former Milltown Dam, backwater effects of
the dam during high-flow conditions might have extended far
enough upstream to affect the hydraulic gradient at the reach 8
outflow (sampling site 20) and also affect the transport of
materials from reach 8. After the removal of the former Mill-
town Dam, the hydraulic gradient at sampling site 20 might
have steepened and promoted transport of materials from
reach 8 during high streamflow conditions.

With the removal of the former Milltown Dam in 2008,
substantial amounts of contaminated sediments that remained
in the Clark Fork channel and flood plain in reach 9 became
more available for mobilization and transport than before
the dam removal. Net mobilization of unfiltered-recoverable


-------
Summary and Conclusions 47

copper, unfiltered-rccovenible arsenic, and suspended sedi-
ment from sources within reach 9 substantially decreased
between periods 3 and 4. Net mobilization of unfiltered-
recoverable copper and arsenic from sources within reach 9
is smaller for period 4 than for period 1 when the former
Milltown Dam was in place, providing evidence that con-
taminant source materials have been substantially reduced in
reach 9. However, net mobilization of suspended sediment
from sources within reach 9 were slightly larger for period 4
than for period 1. A possible explanation for this pattern might
relate to flood-plain disturbance and placement of uncon-
taminated fill in the flood plain associated with remediation
activities. The artificially installed uncontaminated fill might
be more available for mobilization than sediment within the
former Milltown Reservoir during period 1.

Summary and Conclusions

This report characterizes temporal trends in flow-adjusted
concentrations (filtered and unfiltered) of mining-related
contaminants and assesses those trends in the context of
source areas and transport of those contaminants through the
Milltown Reservoir/Clark Fork River Supcrfund Site in the
upper Clark Fork Basin in Montana. The Milltown Reservoir/
Clark Fork River Superfund Site extends about 123 river miles
from the outlet of Warm Springs Ponds on Silver Bow Creek
to the outlet of the former Milltown Reservoir near Missoula.
Trend analysis was done on specific conductance, selected
trace elements (arsenic, copper, and zinc), and suspended sedi-
ment by using a joint time-series model (TSM) for concentra-
tion and streamfiow for seven sampling sites for water years
'1996-2015. The most upstream site included in trend analysis
is Silver Bow Creek at Warm Springs, Montana (sampling
site 8), and the most downstream site is Clark Fork above Mis-
soula, Montana (sampling site 22), which is just downstream
from the former Milltown Dam.

During the extended history of mining in the upper Clark
Fork Basin in Montana, large amounts of waste materials
enriched with metallic contaminants (cadmium, copper, lead,
and zinc) and the metalloid trace element arsenic were gener-
ated from mining operations near Butte, and the milling and
smelting operations near Anaconda. Extensive deposition of
mining wastes in the Silver Bow Creek and Clark Fork chan-
nels and flood plains had substantial effects on water quality.
Federal Superfund remediation activities in the upper Clark
Fork Basin began in 1983 and have included substantial reme-
diation near Butte and removal of the former Milltown Dam.

Water-quality data collection by the U.S. Geological
Survey (USGS) in the upper Clark Fork Basin began dur-
ing 1985-88 with the establishment of a small long-term
monitoring program that has expanded through time and
continued through present (2016). A previous study analyzed
the monitoring data and characterized flow-adjusted trends in
mining-related contaminants for 22 sampling sites in the upper

Clark Fork Basin for water years 1996-2010 (water year is
the 12-month period from October 1 through September 30
and is designated by the year in which it ends). An update of
flow-adjusted water-quality trends for the monitoring data was
needed for seven sampling sites to provide timely information
for the 2016 5-year review for the Milltown Reservoir/Clark
Fork River Superfund Site.

The TSM was used to detect trends in flow-adjusted con-
centrations (FACs). The intent of flow-adjustment is to iden-
tify- and remove streamfiow-related variability in concentration
and thereby enhance the capability to detect trends indepen-
dent from effects of climatic variability. To provide temporal
resolution of changes in water quality, trend analysis was con-
ducted on four sequential 5-year periods: period 1 (water years
1996-2000), period 2 (water years 2001-5). period 3 (water
years 2006-10), and period 4 (water years 2011-15). Because
of the substantial effect of the intentional breach of Milltown
Dam on March 28. 2008, for Clark Fork above Missoula (sam-
pling site 22). period 3 was subdivided into period 3A (Octo-
ber 1, 2005 March 27, 2008) and period 3B (March 28, 2008-
September 30, 2010). The TSM was applied as consistently as
possible among sampling sites and is considered to be a useful
tool for simplifying the environmental complexity in the upper
Clark Fork Basin to provide a large-scale evaluation of general
temporal changes in constituent transport independent from
streamfiow variability.

In conjunction with the trend analysis, estimated normal-
ized constituent loads were calculated and presented in the
framework of a constituent-transport analysis to assess the
temporal trends in FACs in the context of sources and trans-
port. The transport analysis allows assessment of temporal
changes in relative contributions from upstream source areas
to loads transported past each reach outflow.

Trend results are presented for all constituents investi-
gated; however, emphasis is placed on copper, arsenic, and
suspended sediment. Trend results were considered statisti-
cally significant when the statistical probability level (p-value)
was less than 0.01.

Trend results indicate that FACs of unfiltered-recoverable
copper decreased at the sampling sites from the start of
period 1 through the end of period 4; the decreases ranged
from large for one sampling site (Silver Bow Creek at Warm
Springs [sampling site 8]) to moderate for two sampling sites
(Clark Fork near Galen, Montana [sampling site 11] and Clark
Fork above Missoula [sampling site 22]) to small for four
sampling sites (Clark Fork at Deer Lodge. Montana [sampling
site 14], Clark Fork at Goldcreek, Montana [sampling site 16],
Clark Fork near Drummond, Montana [sampling site 18], and
Clark Fork at Turah Bridge near Bonner, Montana [sampling
site 20]). Forperiod 4 (water years 2011-15). the most notable
changes indicated for the Milltown Reservoir/Clark Fork
River Superfund Site in the upper Clark Fork Basin were sta-
tistically significant decreases in FACs and loads of unfiltered-
recoverable copper for sampling sites 8 and 22. For all other
sampling sites, the period 4 changes in FACs of unfiltered-
recoverable copper were not statistically significant.


-------
48 Water-Quality Trends and Constituent-Transport Analysis for Selected Sampling Sites

Trend results indicate that FACs of unfiltered-recoverable
arsenic decreased at the sampling sites from the start of
period 1 through the end of period 4; the decreases ranged
from minor (sampling sites 8-20) to small (sampling site 22).
For period 4 (water years 2011-15), the most notable changes
indicated for the Milltown Reservoir/Clark Fork River Super-
fund Site in the upper Clark Fork Basin were statistically sig-
nificant decreases in FACs and loads of unfiltered-recoverable
arsenic for sampling site 8 and near statistically significant
decreases (p-value of 0.012) for sampling site 22. For all other
sampling sites, the period 4 changes in FACs of unfiltered-
recoverable arsenic were not statistically significant.

Trend results indicate that FACs of suspended sediment
decreased at the sampling sites from the start of period 1
through the end of period 4; the decreases ranged from
moderate (sampling site 8) to small (sampling sites 11-22).
For period 4 (watervears 2011-15), the changes in FACs of
suspended sediment were not statistically significant for any
sampling sites.

The reach of the Clark Fork from Galen to Deer Lodge
is a large source of metallic contaminants and suspended
sediment, which strongly affects downstream transport of
those constituents. Mobilization of unfiltered-recoverable
copper and suspended sediment from flood-plain tailings and
the streambed of the Clark Fork and its tributaries within the
reach results in a contribution of those constituents that is
proportionally much larger than the contribution of streamflow
from within the reach. Within the reach, unfiltered-recoverable
copper loads increased by a factor of about 4 and suspended-
sediment loads increased by a factor of about 5, whereas
streamflow increased by a factor of slightly less than 2. For
period 4 (water years 2011-15), unfiltered-recoverable cop-
per and suspended-sediment loads sourced from within the
reach accounted for about 41 and 14 percent, respectively, of
the loads at Clark Fork above Missoula (sampling site 22).
whereas streamflow sourced from within the reach accounted
for about 4 percent of the streamflow at sampling site 22.
During water years 1996-2015, decreases in FACs and loads
of unfiltered-recoverable copper and suspended sediment for
the reach generally were proportionally smaller than those for
most other reaches.

Unfiltered-recoverable copper loads sourced w ithin the
reaches of the Clark Fork between Deer Lodge and Turah
Bridge near Bonner were proportionally smaller than con-
tributions of streamflow sourced from within the reaches;
these reaches contributed proportionally much less to copper
loading in the Clark Fork than the reach between Galen and
Deer Lodge. Although substantial decreases in FACs and
loads of unfiltered-recoverable copper and suspended sedi-
ment were indicated for Silver Bow Creek at Warm Springs
(sampling site 8), those substantial decreases were not
translated to downstream reaches between Deer Lodge and
Turah Bridge near Bonner. The effect of the reach of the Clark
Fork from Galen to Deer Lodge as a large source of copper

and suspended sediment, in combination with little temporal
change in those constituents for the reach, contributes to this
pattern.

With the removal of the former Milltown Dam in 2008,
substantial amounts of contaminated sediments that remained
in the Clark Fork channel and flood plain in reach 9 became
more available for mobilization and transport than before
the dam removal. After the removal of the former Milltown
Dam, the Clark Fork above Missoula (sampling site 22)
had statistically significant decreases in FACs of unfiltered-
recoverable copper in period 3B (March 28, 2008, through
water year 2010) that continued in period 4 (water years
2011-15). Also, decreases in FACs of unfiltered-recoverable
arsenic and suspended sediment were indicated for period 4
at this site. The decrease in FACs of unfiltered-recoverable
copper for sampling site 22 during period 4 was proportion-
ally much larger than the decrease for the Clark Fork at Turah
Bridge near Bonner (sampling site 20). Net mobilization of
unfiltered-recoverable copper, unfiltered-recoverable arsenic,
and suspended sediment from sources within reach 9 substan-
tially decreased between periods 3 and 4. Net mobilization of
unfiltered-recoverable copper and arsenic from sources within
reach 9 were smaller for period 4 than for period 1 when the
former Milltown Dam was in place, providing evidence that
contaminant source materials have been substantially reduced
in reach 9. However, net mobilization of suspended sediment
from sources w ithin reach 9 were slightly larger for period 4
than for period 1. A possible explanation for this pattern might
relate to flood-plain disturbance and placement of uncoii-
taminated fill in the flood plain associated w ith remediation
activities. The artificially installed uncontaminated fill might
be more available for mobilization than sediment within the
former Milltown Reservoir during period 1.

References

Andrews, E.D., 1987, Longitudinal dispersion of trace metals
in the Clark Fork River, Montana, in Averett, R.C., and
McKnight, D.M., eds., The chemical quality of water and
the liydrologic cycle: Chelsea, Mich., Lewis Publishers,
Inc., p. 1-13.

CDM, 2005, Second five year review report for Silver Bow
Creek/Butte Area Superfund Site: CDM, Helena, Montana,
217 p.

Chanat, J.G., Rice, K.C., and Hornberger, G.M, 2002, Con-
sistency of patterns in concentration-discharge plots: Water
Resources Research, v. 38, 10 p.

Chatham, J.R., 2012, Chemical cycling and nutrient loading
at Warm Springs Ponds (MT): LaPalma, Calif., Atlantic
Richfield Company, 96 p.


-------
References 49

Cleveland, W.S., 1985, The elements of graphing data: Mon-
terey, Calif., Wadsworth Books, 323 p.

Cleveland, W.S., and McGill, Robert, 1984, The many faces
of a scatterplot: Journal of the American Statistical Associa-
tion, v. 79, p. 807-822.

Cohn, T.A., 1988, Adjusted maximum likelihood estimation of
the moments of lognormal populations from type 1 censored
samples: U.S. Geological Survey Open-File Report 88-350,
34 p.

Colby, B.R., 1956, Relationship of sediment discharge to
streamflow: U.S. Geological Survey Open-File Report
56-27, 170 p.

Dodge, K.A., Hornberger, M.I., and Dyke, J.L., 2015, Water-
quality, bed-sediment, and biological data (October 2013
through September 2014) and statistical summaries of data
for streams in the Clark Fork Basin, Montana: U.S. Geolog-
ical Survey Open-File Report 2015-1223, 125 p., accessed
February 1, 2016, at https://pubs.er.usgs.gov/publication/
O&20151223.

Fishman, M.J., 1993, Methods of analysis by the U.S. Geo-
logical Survey National Water Quality Laboratory—Deter-
mination of inorganic and organic constituents in water and
fluvial sediments: U.S. Geological Survey Open-File Report
93-125, 217 p.

Freeman, H.C., 1900, A brief history of Butte, Montana—The
world's greatest mining company: Chicago, 111., Henry O.
Shepard Company, 128 p.

Gammons, C.H., Metesh, J.J., and Duaime, T.E., 2006, An
overview of the mining history and geology of Butte,
Montana: Mine Water and the Environment, v. 25, p. 70-75,
accessed November 19, 2012, at http://dx.doi.org/10.1007/
S10230-006-0113-7.

Garbarino, J.R., Kanagy, L.K., and Cree, M.E., 2006, Determi-
nation of elements in natural-water, biota, sediment, and soil
samples using collision/reaction cell inductively coupled
plasma-mass spectrometry: U.S. Geological Survey Tech-
niques and Methods, book 5, chap. IB, 88 p.

Garbarino, J.R., and Struzeski, T.M., 1998, Methods of analy-
sis by the U.S. Geological Survey National Water Quality
Laboratory—Determination of elements in whole-water
digests using inductively coupled plasma-optical emis-
sion spectrometry and inductively coupled plasma-mass
spectrometry: U.S. Geological Survey Open-File Report
98-165, 101 p.

Helsel, D.R., 2005, Nondetects and data analysis—Statistics
for censored environmental data: New York, John Wiley
and Sons, 250 p.

Helsel, D.R., and Hirsch, R.M., 2002, Statistical methods in
water resources: Techniques of Water-Resources Investiga-
tions of the United States Geological Survey, book 4, chap.
A3, 510 p.

Lambing, J.H., 1998, Estimated 1996-97 and long-term
average annual loads for suspended sediment and selected
trace metals in streamflow of the upper Clark Fork Basin
from Warm Springs to Missoula, Montana: U.S. Geological
Survey Water-Resources Investigations Report 98-4137, 35
p., accessed November 19, 2012, at http://pubs.er.usgs.gov/
publication/wri984137.

Montana Department of Environmental Quality, 2012, DEQ-7
Montana numeric water quality standards: Helena, Mon-
tana, Water Quality Planning Bureau, Water Quality Stan-
dards Section, 76 p.

Montana Department of Environmental Quality, 2016, Clark
Fork River Operable Unit: Montana Department of Envi-
ronmental Quality Web page, accessed May 24, 2016, at

https://deq.mt.gov/Land/fedsuperfund/cfr.

Nimick, D.A., Gammons, C.H., Cleasby, T.E., Madison, J.R,
Skaar, Don, and Brick, C.M., 2003, Diel cycles in dissolved
metal concentrations in streams—Occurrence and possible
causes: Water Resources Research, v. 39, 17 p.

Nimick, D.A., Gammons, C.H., and Parker, S.R., 2011, Diel
biogeochemical processes and their effect on the aqueous
chemistry of streams—A review: Chemical Geology, v. 283,
p. 3-17.

Sando, S.K., and Lambing, J.H., 2011, Estimated loads of
suspended sediment and selected trace elements transported
through the Clark Fork Basin, Montana, in selected periods
before and after the breach of Milltown Dam (water years
1985-2009): U.S. Geological Survey Scientific Investiga-
tions Report 2011-5030, 64 p., accessed November 19,
2012, at http://pubs.usgs.gov/sir/2011/5030.

Sando, S.K., Vecchia, A.V., Lorenz, D.L., and Barnhart, E.P,
2014, Water-quality trends for selected sampling sites in the
upper Clark Fork Basin, Montana, water years 1996-2010:
U.S. Geological Survey Scientific Investigations Report
2013-5217, 162 p., with appendixes.

Smith, J.D., Lambing, J.H., Nimick, D.A., Parrett, Charles,
Ramey, Michael, and Schafer, William, 1998, Geomorphol-
ogy, flood-plain tailings, and metal transport in the upper
Clark Fork Valley, Montana: U.S. Geological Survey Water-
Resources Investigations Report 98-4170, 56 p., accessed
November 19, 2012, at http://pubs.er.usgs.gov/publication/
wri984170.

Stumm, Werner, and Morgan, J.J., 1970, Aquatic chemistry—
An introduction emphasizing chemical equilibria in natural
water: New York, John Wiley and Sons, 583 p.


-------
50 Water-Quality Trends and Constituent-Transport Analysis for Selected Sampling Sites

Tasker, G.D., 1978, Relation between standard errors in log
units and standard errors in percent: U.S. Geological Survey
Water Resources Division Bulletin, January-March 1978,
p. 86-87.

Taylor, J.K., 1987, Quality assurance of chemical measure-
ments: Chelsea, Mich., Lewis Publishers, 328 p.

U.S. Environmental Protection Agency, 2000, Five-year
review report, Silver Bow Creek/Butte Area Superfund Site
with emphasis on the Warm Springs Ponds performance
review: Helena, Montana, U.S. Environmental Protection
Agency, Region 8, 266 p.

U.S. Environmental Protection Agency, 2004, Milltown Res-
ervoir Sediments Operable Unit of the Milltown Reservoir/
Clark Fork River Superfund Site—Record of decision, part
2: U.S. Environmental Protection Agency Decision Sum-
mary, 141 p.

U.S. Environmental Protection Agency, 2005, Second five-
year review report, Silver Bow Creek/Butte Area Superfund
Site: Helena, Montana, U.S. Environmental Protection
Agency, Region 8, 217 p.

U.S. Environmental Protection Agency, 2010, Five-year
review report, Anaconda Smelter National Priority List Site,
Deer Lodge County, Montana: Helena, Montana, U.S. Envi-
ronmental Protection Agency, Region 8, 104 p.

U.S. Geological Survey, 2015, National Water Information
System (NWISWeb): U.S. Geological Survey database,
accessed November 1, 2015, at http://waterdata.usgs.gov/
nwis.

U.S. Geological Survey, variously dated, National field
manual for the collection of water-quality data: U.S. Geo-
logical Survey Techniques of Water-Resources Investiga-
tions, book 9, chaps. A1-A9, 2 v. [variously paged]. [Also
available at http://pubs.water.usgs.gov/twri9A. Chapters
originally were published from 1997-1999; updates and
revisions are ongoing and are summarized at http://water.
usgs.gov/owq/FieldMamial/mastereriata.html.]

Vecchia, A.V., 2003, Water-quality trend analysis and sam-
pling design for streams in North Dakota: U.S. Geological
Survey Water-Resources Investigations Report 03-4094,
79 p.

Vecchia, A.V., 2005, Water-quality trend analysis and sam-
pling design for streams in the Red River of the North
Basin, Minnesota, North Dakota, and South Dakota,
1970-2001: U.S. Geological Survey Scientific Investiga-
tions Report 2005-5224, 60 p.


-------
Appendixes


-------
52 Water-Quality Trends and Constituent-Transport Analysis for Selected Sampling Sites

—Summary Information Relating to Quality-Control Data

Appendix 1

Summary information is presented relating to quality-
control data. Results for quality-control equipment blank and
replicate samples collected during water years 1993-2015
are summarized in table 1-1. Spike recoveries for laboratory-
spiked deionized-water blank samples collected during water
years '1993-2015 are presented in table '1-2. Spike recoveries
for laboratory-spiked stream-water blank samples collected
during water years 1993-2015 are presented in table 1-3. For
reference, aquatic-life standards (based on median hardness
for water years 2011—15, Montana Department of Environ-
mental Quality, 2012) arc presented in table 1-4.

Evaluation of long-term spike-recovery data is particu-
larly relevant to the long-term trend analysis. Spike-recov-
eries during water years '1993-2015 for laboratorv -spiked
deionized-water blank samples (table 1-2 and fig. 1-1)
and laboratory-spiked stream-water samples (table 1-3 and
fig. 1-2) indicate generally consistent recoveries overtime.

typically varying within plus or minus 10 percent of 100 per-
cent recovery. However, before about water year 2000, spike
recoveries for unfiltcrcd-recovcrable copper in spiked stream-
water samples generally were near 100 percent (mean annual
spike recover^' for water years 1993-99 of 99.1 percent),
whereas after about water year 2000, spike recoveries mostly
were less than 100 percent (mean annual spike recovery
for water years 2000-15 of 94.3 percent). Changes in spike
recoveries in about water year 2000 probably were related
to a change in about water year 2000 by the U.S. Geological
Survey National Water Quality Laboratory from analysis of
most metallic elements by graphite furnace atomic absorption
spectrophotometry (Fishman, 1993) to inductively coupled
plasma-mass spectrometry (Garbarino and Struzesld, 1998;
Garbarino and others, 2006). The potential effects of temporal
changes in spike recoveries on trend results were evaluated in
exploratory analyses, as described in appendix 2.


-------
Table 1-1. Summary information relating to quality-control samples (field equipment blank and replicate samples) collected at sampling sites in the upper Clark Fork Basin,
Montana, water years 1993-2015.

f Water year is the 12-month period from October 1 through September 30 and is designated by the year in which it ends, LRL, laboratory reporting level; SRL, study reporting level; RSD, relative standard
deviation; US/cm, microsiemen per centimeter at 25 degrees Celsius; NA, not applicable; ug L, microgram per liter; mg/L, milligram per liter]

Summary information for field blank samples

Constituent or property,
units of measurement

Number of field
blank samples
Number of with detected
field blank concentrations
samples greater than the
LRL at the time
of analysis

Percentage
of field blank
samples with

detected
concentrations
greater than the
LRL at the time
of analysis

Maximum
detected

concentration
for field blank
samples

Median
concentration
in field blank
samples with

detected
concentrations
greater than the
LRL at the time of
analysis

SRL used in
application

of the
time-series
model

Percentage of
detections in
blank samples at
concentrations
greater than the
SRL used in the
application of
the time-series
model

Summary information
for field replicate
samples

Number

of field RSD,1
replicate in percent
pairs

Specific conductance, p.S/cm

NA

NA

NA

NA

NA

NA

NA

162

0.1

Cadmium, filtered, jig/L

193

5

2.6

0.337

0.071

NA

NA

179

13.4

Cadmium, unfiltered-recoverable, p,g/L

189

1

0.5

0.010

0.010

NA

NA

180

4.5

Copper, filtered,2 ug/L

192

15

7.8

3.6

0.50

1.0

1.0

182

12.4

Copper, unfiltered-recoverable,2 mg/L

189

11

5.8

3.0

1.0

1.0

2.1

180

9.0

Iron, filtered, (ig/L

189

4

2.1

5.9

4.8

NA

NA

171

9.8

Iron, unfiltered-recoverable, jig/L

185

10

5.4

35.6

7.0

NA

NA

178

s s

Lead, filtered, |ig/L

193

6

3.1

0.600

0.101

NA

NA

178

11.0

Lead, unfiltered-recoverable, ug/L

189

10

5.3

0.16

0.05

NA

NA

180

16.3

Manganese, filtered, |ig/L

188

22

11.7

0.62

0.36

NA

NA

183

5.7

Manganese, unfiltered-recoverable, ug/L

185

10

5.4

0.3

0.2

NA

NA

180

5.8

Zinc, filtered, (.ig/L

191

39

20.4

6.2

0.9

NA

NA

181

9.6

Zinc, unfiltered-recoverable,2 ug/L

187

20

10.7

3.4

1.4

2.0

2.7

181

9.0

Arsenic, filtered,2 jig/L

193

1

0.5

0.1

0.1

1.0

0.0

182

5.4

Arsenic, unfiltered-recoverable,2 (ig/L

189

3

1.6

0.1

0.1

1.0

0.0

181

6.8

Suspended sediment,2 mg/L

NA

NA

NA

NA

NA

1

NA

170

9.1

WiSD is calculated according to the following equation (Taylor, 1987):

RSD = -£tx100,

X

where

RSD is the relative standard deviation;

S is the standard deviation; and

X is the mean concentration for all replicate analyses.
2 Property or constituent was analyzed for temporal trends.


-------
Table 1-2. Summary information relating to quality-control samples (laboratory-spiked deionized-water blank samples) collected at sampling sites in the upper Clark Fork
Basin, Montana, water years 1993-2015.

[Water year is the 12-month period from October 1 through September 30 and is designated by the year in which it ends. F, filtered; UFR, unfiltered-recoverable]

Water

Cadmium,

Cadmium,

Copper,

Copper,

Iron,

Iron,

Lead,

Lead,

Manganese,

Manganese,

Zinc,

Zinc,

Arsenic,

Arsenic,

year

F

UFR

F

UFR

F

UFR

F

UFR

F

UFR

F

UFR

F

UFR







Mean spike recovery, in percent (values

in parentheses indicate 95 percent confidence intervals)







1993

93.4

97

99.5

101.7

94

103.3

105.8

100.5

96.9

95.6

106.5

96.3

94

102.6



(85.9, 101)

(93.5, 101)

(95.9, 103)

(94.4, 109)

(90.0, 98.0)

(92.4, 114)

(99.5, 112)

(95.2, 106)

(96.3, 97.5)

(82.2, 109)

(99.7, 113)

(94.1, 98.5)

(89.6, 98.4)

(95.8, 109)

1994

97.5

98.8

101.1

99.7

100

94.6

100.5

99.1

95.7

101.5

106.5

102.6

100.6

109.3



(89.1, 106)

(90.6, 107)

(98.4, 104)

(94.3, 105)

(93.0, 107)

(84.2, 105)

(98.5, 102)

(94.3, 104)

(90.8, 100)

(96.2, 107)

(95.8, 117)

(91.5, 114)

(95.6, 106)

(104, 114)

1995

100

101.3

102.7

97.6

102.2

93.8

102.3

100.8

96.5

98.5

102.3

101.5

103.9

106.8



(97.3, 103)

(97.5, 105)

(101, 105)

(92.3, 103)

(97.8, 107)

(87.9, 99.7)

(97.7, 107)

(96.6, 105)

(92.0, 101)

(93.1, 104)

(97.1, 108)

(97.1, 106)

(99.1, 109)

(103, 110)

1996

95.3

82.3

99.2

99.6

89.8

90.8

100.5

97.4

89.2

96.5

96.1

87.8

89.7

104.1



(92.2, 98.4)

(79.7, 84.9)

(91.4, 107)

(93.5, 106)

(76.0, 104)

(70.9, 111)

(93.3, 108)

(80.2, 115)

(77.9, 100)

(91.6, 101)

(84.3, 108)

(82.8, 92.8)

(77.1, 102)

(101, 107)

1997

98.5

85.7

101.1

106.4

94.7

96.1

101

101.1

90.3

99.3

97.9

92.7

93.9

106.1



(92.1, 105)

(77.7, 93.7)

(86.2, 116)

(82.0, 131)

(78.5, 111)

(80.2, 112)

(93.4, 109)

(88.9, 113)

(82.7, 97.9)

(95.8, 103)

(78.1, 118)

(86.4, 99.0)

(87.8, 100)

(104, 108)

1998

104

97.4

100.4

103.4

101.8

95.7

100.2

104.8

102.8

99

95.2

101.3

91.5

105.4



(93.8, 114)

(87.0, 108)

(93.4, 107)

(98.8, 108)

(90.7, 113)

(89.9, 102)

(91.8, 109)

(88.8, 121)

(94.4, 111)

(92.1, 106)

(85.9, 104)

(86.9, 116)

(87.3, 95.7)

(99.2, 112)

1999

100.9

103.4

107.5

105

97.7

96.5

97.4

96.2

96

95.9

96.9

93.3

108.9

102.9



(92.6, 109)

(99.9, 107)

(99.5, 116)

(102, 108)

(94.3, 101)

(90.0, 103)

(87.9, 107)

(85.2, 107)

(91.8, 100)

(86.3, 106)

(92.9, 101)

(88.9, 97.7)

(95.4, 122)

(97.8, 108)

2000

103.8

105

104

100.3

97.4

100.6

98.3

102.6

100.8

103.2

107.8

102.6

101.6

101.4



(97.3, 110)

(96.0, 114)

(96.0, 112)

(92.4, 108)

(92.3, 102)

(89.2, 112)

(88.9, 108)

(97.3, 108)

(93.3, 108)

(96.8, 110)

(95.8, 120)

(90.0, 115)

(95.3, 108)

(95.1, 108)

2001

102.9

107.9

105.2

96.8

101.3

98.3

97.3

96.4

101.9

103.7

102

99.1

99.2

97.7



(98.9, 107)

(101, 115)

(98.6, 112)

(93.7, 99.9)

(95.5, 107)

(86.7, 110)

(91.9, 103)

(93.7, 99.1)

(79.0, 125)

(89.9, 118)

(87.9, 116)

(82.7, 116)

(92.3, 106)

(86.6, 109)

2002

101.1

97.6

99.4

98.8

95.1

102.3

98.5

96.9

98.5

96.5

103.9

98.3

105.1

97.9



(98.8, 103)

(96.3, 98.9)

(95.0, 104)

(96.7, 101)

(89.3, 101)

(93.0, 112)

(89.9, 107)

(90.5, 103)

(95.4, 102)

(88.8, 104)

(94.4, 113)

(91.8, 105)

(95.8, 114)

(93.0, 103)

2003

98.6

97.5

100.4

97.6

101.6

93.1

97.2

96

95.8

96.6

101.4

99.1

87.9

96.6



(92.6, 105)

(94.1, 101)

(93.0, 108)

(93.2, 102)

(96.4, 107)

(87.4, 8.8)

(92.3, 102)

(93.9, 98.1)

(90.7, 101)

(79.7, 114)

(89.8, 113)

(93.2, 105)

(71.3, 104)

(78.5, 115)

2004

97.4

100

98.9

99.6

101

96.1

96

98.9

99.1

98.6

102

100

101

102



(95.6, 99.2)

(98.6, 101)

(92.7, 105)

(95.4, 104)

(96.3, 106)

(88.8, 103)

(91.9, 100)

(97.3, 100)

(92.3, 106)

(90.6, 107)

(91.7, 112)

(96.3, 104)

(75, 127)

(93.6, 110)

2005

102

97.5

102

97.6

97.6

100

101

104

93.8

102

102

96.1

97.4

101



(97.3, 106)

(88.1, 107)

(97.4, 107)

(88.4, 107)

(90.5, 105)

(95.2, 105)

(95.5, 106)

(99.4, 108)

(82.2, 105)

(86.4, 117)

(88.3, 116)

(83.5, 109)

(95.5,99.3)

(90.7, 111)

2006

100

98.9

102

98.7

106

103

99

98

97

105

105

94.9

95.2

98.5



(92.6, 107)

(94.1, 104)

(97.7, 107)

(93.8, 104)

(101, 112)

(95.4, 111)

(89.3, 109)

(91.2, 105)

(90.7, 103)

(95.3, 115)

(95.4, 115)

(90.1, 100)

(89.2, 101)

(94.7, 102)

2007

107

103

105

98.4

99.9

104

99.6

103

107

107

107

103

105

102



(103,112)

(94.4, 111)

(99.2, 111)

(86.9, 110)

(92.1, 108)

(98.5, 110)

(93.9, 105)

(100, 106)

(99.9, 114)

(97.0, 116)

(102, 113)

(96.5, 110)

(96.6, 114)

(95.2, 109)

2008

102

101

105

97.9

103

101

101

101

102

102

99.8

103

103

102



(88.2, 116)

(91.9, 110)

(88, 121)

(87.2, 109)

(95.9, 110)

(96.5, 106)

(89, 112)

(98, 105)

(92.9, 111)

(92.5, 112)

(87.9, 112)

(96, 111)

(89.2, 117)

(93.9, 110)

2009

102

97.2

102

96

102

104

102

98.4

105

99.7

111

93.3

101

97



(97.4, 107)

(93.6, 101)

(92.0, 113)

(94.0, 97.0)

(91.4, 112)

(78.8, 130)

(96.0, 107)

(96.1, 101)

(103, 106)

(94.6, 105)

(104, 118)

(88.5, 98.1)

(92.3, 110)

(94.9, 99.1)

2010

106

100

97.2

98.6

108

102

102

102

103

105

113

101

105

102



(94.9, 117)

(88.4, 112)

(84.9, 109)

(84.0, 113)

(101, 115)

(95.8, 108)

(91.5, 113)

(91.0, 113)

(95.2, 111)

(97.2, 112)

(94.7, 132)

(89.6, 113)

(96.7, 113)

(89.7, 114)


-------
Table 1-2. Summary information relating to quality-control samples (laboratory-spiked deionized-water blank samples) collected at sampling sites in the upper Clark Fork
Basin, Montana, water years 1993-2015.—Continued

[Water year is the 12-month period from October 1 through September 30 and is designated by the year in which it ends. F, filtered; UFR, unfiltered-recoverable]

Water

Cadmium,

Cadmium,

Copper,

Copper,

Iron,

Iron,

Lead,

Lead,

Manganese,

Manganese,

Zinc,

Zinc,

Arsenic,

Arsenic,

year

F

UFR

F

UFR

F

UFR

F

UFR

F

UFR

F

UFR

F

UFR

Mean spike recovery, in percent (values in parentheses indicate 95 percent confidence intervals)—Continued

2011

105

95.7

96.2

93.9

ill

107

106

99.8

101

98.9

108

96.1

105

94.7



(97.9, 111)

(92.4, 99)

(89.4, 103)

(91.6, 96.2)

(89.3, 132)

(98.2, 117)

(98.8, 113)

(98.4, 101)

(97.0, 104)

(97.8, 100)

(94.3, 122)

(92.2, 100)

(102, 109)

(90.2, 99.3)

2012

102

101

98.4

100

105

106

102

103

105

101

103

100

98.1

101



(93.2, 112)

(95.1, 108)

(93.1, 104)

(92.5, 107)

(102, 108)

(96.2, 117)

(96.8, 106)

(98.4, 107)

(101, 110)

(95.4, 106)

(96.5, 109)

(94.9, 106)

(90.4, 106)

(94.3, 108)

2013

96.3

96.6

92.4

96.3

103

105

97.5

99.9

98.1

98.5

98.6

95.2

98

99.3



(92.4, 100)

(92.9, 100)

(87, 97.9)

(92.6, 100)

(95.5, 111)

(98.2, 112)

(92.3, 103)

(97.1, 103)

(92.3, 104)

(94.8, 102)

(90.9, 106)

(91.7, 98.7)

(93.1, 103)

(96, 103)

2014

99.4

101

98.1

100

103

103

102

103

99.2

100

110

101

94.7

102



(95.1, 104)

(99.0, 104)

(91.0, 105)

(98.8, 102)

(95.8, 111)

(99.7, 106)

(100, 104)

(100, 107)

(91.6, 107)

(97.7, 103)

(103, 117)

(97.1, 104)

(87.6, 102)

(99.0, 105)


-------
Table 1-3. Summary information relating to quality-control samples (laboratory-spiked stream-water samples) collected at sampling sites in the upper Clark Fork Basin,
Montana, water years 1993-2015.

[Water year is the 12-month period from October 1 through September 30 and is designated by the year in which it ends. F, filtered; UFR, unfiltered-recoverable]

Water

Cadmium,

Cadmium,

Copper,

Copper,

Iron,

Iron,

Lead,

Lead,

Manganese,

Manganese,

Zinc,

Zinc,

Arsenic,

Arsenic,

year

F

UFR

F

UFR

F

UFR

F

UFR

F

UFR

F

UFR

F

UFR







Mean spike recovery, in percent (values

in parentheses indicate 95 percent confidence intervals)







1993

97.1

98.1

97.4

97.2

94.6

102.2

104.7

96

95.7

100.2

105.7

95.7

95.2

99.9



(92.3, 102)

(95.2, 101)

(95.8, 99.0)

(92.3, 102)

(86.7, 103)

(94.4, 110)

(98.5, 111)

(93.0, 99.0)

(92.1,99.3)

(96.4, 104)

(93.4, 118)

(92.2, 99.2)

(92.0, 98.3)

(96.5, 103)

1994

101.3

97.9

96.6

98.4

98.2

99.3

103

99.3

98.1

100.4

97.5

106

97.3

106.9



(97.5, 105)

(94.4, 101)

(93.3, 99.8)

(91.1, 106)

(94.8, 102)

(90.6, 108)

(101, 105)

(95.6, 103)

(95.4, 101)

(95.4, 105)

(92.4, 102)

(95.4, 117)

(90.4, 104)

(101, 113)

1995

101.3

102.9

99.8

98

99.5

101.4

102.9

100

97.4

103.8

104.7

101.1

103.8

102.2



(96.7, 106)

(98.0, 108)

(96.2, 103)

(92.7, 103)

(96.1, 103)

(96.2, 107)

(98.6, 107)

(96.7, 103)

(92.9, 102)

(99.0, 109)

(101, 108)

(99.1, 103)

(94.6, 113)

(97.1, 107)

1996

100.2

88.4

101.1

100.3

93.8

101.5

105.1

105.6

90.3

99.5

103.2

99.3

105.9

102.8



(91.5, 109)

(57.8, 119)

(91.9, 110)

(92.3, 108)

(73.3, 114)

(88.5, 114)

(90.4, 120)

(98.4, 113)

(79.1, 102)

(92.9, 106)

(90.2, 116)

(74.8, 124)

(94.4, 117)

(96.0, 110)

1997

98.1

84.3

97.3

100.5

99.3

97.5

100.8

102.1

93

99.8

97

92.7

93.3

107.1



(83.5, 113)

(75.0, 93.6)

(88.3, 106)

(71.9, 129)

(81.0, 118)

(78.2, 117)

(91.6, 110)

(99.1, 105)

(84.0, 102)

(94.5, 105)

(89.9, 104)

(74.4, 111)

(73.5, 113)

(99.9, 114)

1998

104.4

99.5

97.2

99.1

97.5

101.8

102.2

105

99.5

101.5

99.5

98.8

90.1

104



(97.3, 112)

(92.7, 106)

(90.6, 104)

(88.4, 110)

(82.8, 112)

(90.2, 113)

(94.3, 110)

(92.9, 117)

(85.8, 113)

(98.0, 105)

(89.1, 110)

(85.6, 112)

(85.5, 94.7)

(95.8, 112)

1999

102.6

103

102.7

100.5

97.2

99.9

100.2

101.1

99.8

98.8

98.6

96.2

105.2

103.6



(92.4, 113)

(100, 106)

(89.1, 116)

(97.5, 104)

(93.5, 101)

(90.6, 109)

(94.0, 106)

(93.7, 108)

(92.8, 107)

(89.3, 108)

(95.7, 102)

(91.1, 101)

(97.5, 113)

(96.4, 111)

2000

104.2

98.1

101.6

94.6

96.5

98

101.4

105.3

97.3

101.7

101.5

97.8

102.5

98.9



(100, 108)

(88.9, 107)

(97.3, 106)

(87.7, 102)

(88.0, 105)

(88.3, 108)

(97.3, 106)

(103, 108)

(83.3, 111)

(91.4, 112)

(90.9, 112)

(91.1, 104)

(97.5, 108)

(87.8, 110)

2001

103.2

105.8

106.8

91.8

95.8

101.6

99.7

97.3

100

100.9

100.8

96.9

102.8

100.1



(100, 106)

(95.9, 116)

(104, 110)

(87.7, 95.9)

(91.4, 100)

(92.1, 111)

(95.2, 104)

(95.3, 99.3)

(84.4, 116)

(90.3, 112)

(85.7, 116)

(75.9, 118)

(95.1, 110)

(96.7, 104)

2002

106

102

97.3

96.9

92.6

107.1

101.4

98.9

98.3

94.3

101.3

95.8

105.8

99.9



(97.5, 114)

(98.6, 101)

(91.2, 103)

(92.9, 101)

(83.3, 102)

(103, 111)

(91.9, 111)

(92.2, 106)

(92.5, 104)

(88.4, 100)

(92.6, 110)

(89.9, 102)

(97.1, 114)

(86.0, 114)

2003

100.5

99

95.8

91.6

106.4

96.7

96

96.8

93.9

99.3

98.4

93

94.6

108.6



(91.4, 110)

(94.4, 104)

(88.9, 103)

(89.7, 93.5)

(100, 113)

(91.6, 102)

(90.2, 102)

(93.7, 99.9)

(78.8, 109)

(86.2, 112)

(93.6, 103)

(87.5, 98.5)

(80.2, 109)

(100, 117)

2004

101

101

95.4

93.8

104

111

98.7

100

103

96

100

94.4

97.3

112



(94.2, 108)

(100, 103)

(93.8, 97)

(89.5, 98.1)

(99.5, 108)

(91.2, 130)

(93, 104)

(98.6, 102)

(89.8, 117)

(91.8, 100)

(95.3, 105)

(91,97.8)

(86.9, 108)

(106, 118)

2005

97.8

98.2

93.6

93

102

99.3

102

103

88.3

97.5

94.3

91.6

103

104



(62.7, 133)

(88.5, 108)

(57.9, 129)

(84.8, 101)

(95.9, 108)

(95.6, 103)

(96.1, 109)

(99.7, 106)

(78.3, 98.3)

(87.3, 108)

(60.8, 128)

(80.8, 102)

(98.3, 107)

(101, 108)

2006

104

99.6

101

94.8

105

102

102

100

94.9

106

108

91.2

96.5

99.1



(99.0, 108)

(94.7, 104)

(96.7, 104)

(91.0, 98.6)

(102, 109)

(93.6, 110)

(94.2, 111)

(92.9, 106)

(88.2, 102)

(97.9, 113)

(93.3, 123)

(87.8, 94.6)

(89.0, 104)

(94.9, 103)

2007

108

98

100

96.3

107

103

109

104

106

101

104

98

106

102



(102, 114)

(92.2, 104)

(89.8, 110)

(91.8, 101)

(103, 111)

(94.7, 112)

(103, 115)

(102, 107)

(100, 113)

(96.1, 106)

(95.7, 113)

(89.2, 107)

(100, 113)

(98.2, 106)

2008

101

97

98.9

92.8

105

99.4

100

103

98.9

98.4

106

95.7

100

101



(91, 112)

(93.6, 100)

(92, 106)

(86.4, 99.1)

(94.1, 117)

(92, 107)

(91.3, 109)

(99.5, 106)

(90.3, 108)

(92.5, 104)

(88.1, 124)

(93.1, 98.2)

(90.2, 110)

(98.5, 104)

2009

106

94.7

96.2

91.4

107

102

100

100

97

92.8

114

89.8

106

100



(101, 112)

(89.5, 99.8)

(91.2, 101)

(87.8, 95.0)

(89.7, 124)

(86.9, 118)

(97.0, 103)

(98.8, 101)

(88.0, 106)

(81.7, 104)

(104, 124)

(80.4, 99.2)

(97.7, 114)

(89.6, 111)

2010

110

98.2

93.8

96.5

105

111

101

104

104

98.7

109

94

106

102



(87.6, 132)

(87.1, 109)

(83.6, 104)

(84.4, 108)

(91.7, 119)

(103, 118)

(87.7, 115)

(91.5, 116)

(93.3, 114)

(86.4, 111)

(101, 118)

(81.3, 107)

(96.0, 116)

(90.1, 113)


-------
Table 1-3. Summary information relating to quality-control samples (laboratory-spiked stream-water samples) collected at sampling sites in the upper Clark Fork Basin,
Montana, water years 1993-2015.—Continued

[Water year is the 12-month period from October 1 through September 30 and is designated by the year in which it ends. F, filtered; UFR, unfiltered-recoverable]

Water

Cadmium,

Cadmium,

Copper,

Copper,

Iron,

Iron,

Lead,

Lead,

Manganese,

Manganese,

Zinc,

Zinc,

Arsenic,

Arsenic,

year

F

UFR

F

UFR

F

UFR

F

UFR

F

UFR

F

UFR

F

UFR

Mean spike recovery, in percent (values in parentheses indicate 95 percent confidence intervals)—Continued

2011

104

93.9

96.6

88.3

108

101

104

96.5

98.2

91.3

102

86.7

106

94.7



(99.2, 109)

(91.5, 96.3)

(79.9, 113)

(85.4, 91.2)

(92.0, 124)

(85.2, 117)

(98.8, 110)

(94.5, 98.4)

(92.2, 104)

(88.3, 94.2)

(90.2, 114)

(80.7, 92.7)

(101, 111)

(90.5, 99.0)

2012

107

98.8

94

93.9

108

100

102

101

101

95.5

102

89.8

104

97.5



(104, 110)

(91.9, 106)

(90.9, 97)

(87.2, 101)

(102, 114)

(98.6, 102)

(97.9, 107)

(96.3, 105)

(97.7, 104)

(88, 103)

(95.2, 109)

(82.4, 97.2)

(101, 106)

(91.8, 103)

2013

94.8

91.3

90.9

90

102

101

101

96.7

97.2

93

99.5

84.1

99.5

94.9



(90.4, 99.3)

(87, 95.7)

(86, 95.8)

(87.5, 92.4)

(94.8, 110)

(92.6, 110)

(92.8, 108)

(92.3, 101)

(95.4, 99)

(84.9, 101)

(92, 107)

(79.5, 88.7)

(91.2, 108)

(91, 98.8)

2014

103

95.5

96.6

93.8

97.6

101

100

99.7

97.1

94.8

101

88.9

92.4

97.7



(95.6, 110)

(92.0, 99.0)

(90.1, 103)

(89.8, 97.8)

(92.7, 103)

(92.7, 109)

(96.7, 103)

(94.9, 104)

(90.4, 104)

(89.3, 100)

(94.2, 108)

(82.7, 94.6)

(82.7, 102)

(93.5, 102)

2015

104

106

97.4

97.8

93.5

104

103

106

102

101

93.8

98.1

96.8

104



(97.6, 111)

(96.6, 115)

(92.3, 102)

(92.9, 103)

(83.2, 104)

(101, 106)

(101, 105)

(96.0, 115)

(98.3, 105)

(92.0, 110)

(86.2, 101)

(88.9, 107)

(86.5, 107)

(87.3,121)


-------
58 Water-Quality Trends and Constituent-Transport Analysis for Selected Sampling Sites

Table 1-4, Aquatic-life standards (based on median hardness for water years 2011-15) for selected sampling sites in the Milltown
Reservoir/Clark Fork River Superfund Site in the upper Clark Fork Basin, Montana,

[Water year is the 12-month period from October 1 through September 30 and is designated by the year in which it ends. CaCO , calcium carbonate]

Sampling
site

number

(fig- 1,

table 1)

Aquatic-life standards (Montana Department of Environmental Quality, 2012),
in micrograms per liter

Abbreviated sampling site name
(table 1)

Cadmium

Copper

Lead

Zinc

Median
hardness for

water years	

2011-15, in

milligrams per Acute Chronic Acute Chronic Acute Chronic Acute Chronic
liter as CaCO,

8

Silver Bow Creek at Warm Springs

170

3.66

0.401

23.1

14.7

160

6.25

188

188

11

Clark Fork near Galen

164

3.53

0.390

22.3

14.2

153

5.97

182

182

14

Clark Fork at Deer Lodge

200

4.32

0.452

26.9

16.9

197

7.69

216

216

15

Clark Fork near Garrison

202

4.36

0.456

27.2

17.0

199.8

7.79

217

217

16

Clark Fork at Goldcreek

165

3.54

0.391

22.4

14.3

154

6.00

183

183

18

Clark Fork near Drummond

190

4.09

0.435

25.6

16.1

184

7.18

206

206

20

Clark Fork at Turah Bridge

132

2.82

0.331

18.1

11.8

116

4.51

151

151

22

Clark Fork above Missoula

109

2.33

0.288

15.2

10.0

91

3.55

129

129


-------
Appendixes 59

Water year (October-September)

EXPLANATION

95 percent confidence interval about the mean spike recovery
Mean spike recovery

Figure 1-1. Spike recoveries for laboratory-spiked deionized-water blank samples, water years 1993-2015. A, copper,
filtered; 6, copper, unfiltered-recoverable; C, arsenic, filtered; D, arsenic, unfiltered-recoverable.


-------
Water-Quality Trends and Constituent-Transport Analysis for Selected Sampling Sites

Water year (October-September)

EXPLANATION

	 95 percent confidence interval about the mean spike recovery

—— Mean spike recoveiy

Figure 1-2. Spike recoveries for laboratory-spiked stream-water samples, water years 1993-2015. A, copper, filtered;
6, copper, unfiltered-recoverable; C, arsenic, filtered; D, arsenic, unfiltered-recoverable.


-------
Appendixes 61

Appendix 2—Summary of the Time-Series Model as Applied in this Study

This appendix presents somewhat detailed information on
theoretical and computational aspects of the time-series model
(TSM). Also, specific aspects of the application of the TSM in
this study are described.

Theoretical and Computational Information

The theory and parameter estimation for the TSM arc
described in detail in Vecchia (2005). In the TSM, log-trans-
formed concentration data are partitioned into several compo-
nents according to equation 1:

log(C) = Mc +ANN + SEASC + TREND + HFVC (1)

where

log
C

.
-------
62 Water-Quality Trends and Constituent-Transport Analysis for Selected Sampling Sites

period and the geometric mean concentration at the start of the
period and is determined by the equation

%tSFAC = 100(l07 -l),	(2)

where

%AFAC is the percentage change in the geometric

mean of the flow-adjusted concentration,
and

y	is the slope coefficient of the trend for the

specified trend-analysis period in log-
transformed units.

Log-transformed concentrations that have ANNC and
SFASC removed are referred to in this report as "flow-adjusted
concentrations." By using equation 1, the flow-adjusted con-
centration is defined as

FAC = log(C) -ANNC - SEASC = Mc + TREND + HFVC (3)

where FAC is the flow-adjusted value, as the base-10 loga-
rithm of the original units of measurement. The FA Cs defined
by equation 3 are analogous to FACs defined in other publica-
tions as the residuals from a regression model that relates con-
centration to concurrent daily streamflow (Helsel and Hirsch,
2002); however, the TSM approach generally is more effective
than a regression-based approach for removing streamflow-
related variability (Vccchia, 2005). Time-series plots show-
ing FACs along with the fitted trend (Mc + TREND) illustrate
long-term changes in geometric mean concentration that might
indicate changes in effects of mining wastes on water-quality
in the selected watersheds.

The key to making TSM a powerful trend-analysis tool is
that the entire time series of daily streamflow data are used in
the model, not just streamflow for the days when concentra-
tion samples arc available. The model uses a thrcc-pcr-month.
or approximately 10-day, sampling frequency. Each month
is divided into three intervals—days 1-10, days 11-20, and
day 21 through the end of the month. If a water-quality sample
is available for a particular interval, it is paired with daily
streamflow for the same day of the water-quality sample. If no
water-quality sample is available, the concentration value for
the interval is missing, and streamflow for the middle of the
interval (day 5, 15, or 25) is used. If more than one concen-
tration sample is available for the interval, the value nearest
to the midpoint of the interval is used. The log-transformed
streamflow time series (consisting of three values per
month) is divided into an annual anomaly, seasonal anomaly,
and high-frequency variability according to the following
equation:

log (6) = Mg + ANNq + SEASe + HFVe	(4)

where

O is daily mean streamflow, in cubic feet per
second;

M is the mean of the log-transformed streamflow
for the entire trend-analysis period, as the
base-10 logarithm of cubic feet per second;

ANNq is the annual streamflow anomaly, computed
as the 1-year lagged moving average of
l°g(0 ~~Mq (dimensionless);

SE/lSg is the seasonal streamflow anomaly, computed
as the 3-month lagged moving average of
log(0 -M -ANN (dimensionless); and

HFVg is the high-frequency streamflow variability,
computed as log(O) - Mg - ANNg - SFASg
(dimensionless).

The water-quality time-series model (equation 1) is
directly tied to the streamflow time-series model because the
streamflow anomalies (ANN and SEASgfmm equation 4)
arc used as predictor variables for concentration (equation 1).
For example, ANNC is assumed to equal a constant coefficient
(estimated from the TSM) times ANNg. The different scales of
streamflow variability often affect concentration in different
ways. The relation between IIFV". and HFVn can be particu-

V;

larly complicated, changing depending on the time of year and
the degree of serial correlation in the concentration data and
cross-correlation between concentration and streamflow.

Specific Aspects of the Application of the Time-
Series Model in this Study

The TSM residuals for each sampling-site and constitu-
ent combination were examined graphically to verify- the
model assumptions that the residuals had constant variance,
were serially uncorrelated, and were approximately normally
distributed. Because of the application of the TSM to the large
number of sampling-site and constituent combinations and
practical considerations to keep the trend periods comparable
among sampling sites and constituents, some minor deviations
of the residuals from model assumptions were tolerated. Such
deviations included small changes in residual variance through
time and short-term (about 1-2 years) unresolved trending in
the residuals. In cases where unresolved residual trends were
considered to be large enough to possibly affect the magni-
tudes and significance levels of reported fitted trends, more
complicated trend models were tested, and in all cases the
more complicated models did not substantially affect the over-
all descriptions of the trends and also did not change the gen-
eral findings and conclusions of this report. Thus, the reported
TSM results were judged to provide acceptable fits representa-
tive of linearity through nearly all of the range in FACs for


-------
Appendixes 63

a given sampling-site and constituent combination. Standard
errors of estimates (SEEs) for the TSM analyses are presented
in table 2-1. In this report, SEEs are expressed in percent and
were converted from log units by using procedures described
by Tasker (1978). Mean SEEs for all trace elements combined
range from 20.8 to 50.7 percent. Mean SEEs for unfiltered-
recoverable copper and arsenic concentrations are 48.3 and
27.3 percent, respectively. Mean SEE for suspended-sediment
concentration (65.2 percent) is substantially higher than mean
SEEs for trace elements. The SEEs indicate reasonably accu-
rate definition of concentration and streamflow relations for
the purpose of trend analysis; however, a higher mean SEE for
suspended sediment than mean SEEs for trace elements indi-
cates lower confidence in results. For each sampling-site and
constituent combination, the fit of the TSM can be assessed
by examination of the fitted trends in relation to FACs that are
shown in figures 3-1 through 3-7 in appendix 3. The distri-
bution of FACs about the fitted trend lines shows the extent
to which the residuals might exhibit nonconstant variance or
unresolved trends.

Application of the TSM in this study generally followed
the methods applied by Sando and others (2014) who reported
water-quality trends for 22 sampling sites in the upper Clark
Fork Basin for water years 1996-2010. However, two factors
might contribute to differences between Sando and others
(2014) and this study: (1) this study included additional data
collected after the study period of Sando and others (2014).
and (2) this study included preliminary dummy trend peri-
ods that were inserted prior to period 1. The additional data
after the study period of Sando and others (2014) represent
an increase of about 25 percent and provide improvement in
definition of concentration and streamflow relations used in
determining FACs. Also, during exploratoiy analysis for this
study, close scrutiny of the fitted trends reported by Sando
and others (2014) indicated that in some cases the fitted
trend values at the start of period 1 (1996) were not precisely
centered at the median FAC at the start of period 1. In this
study, dummy trend periods were inserted before period 1
to more precisely center the 1996 fitted trend values at the
median FAC. The combination of the two factors (inclusion
of additional data and insertion of preliminary dummy trends)
sometimes resulted in generally minor differences in the fitted
trend lines betw een this report and Sando and others (2014).
The trend results of this report supersede the trend results of
Sando and others (2014).

Exploratory analyses were conducted to investigate two
ancillary factors that might affect trend results, including
potential effects of (1) temporal changes in spike recover-
ies (as discussed in appendix 1) and (2) diel cycling of trace
elements. The potential effects of temporal changes in spike
recoveries (as discussed in appendix 1) on trend results were
evaluated by using two approaches: (1) exploratoiy trend
analysis with inclusion of a step trend in the trend model and
(2) exploratory trend analysis on constituent concentrations
adjusted based on annual mean spike recoveries. For the
exploratoiy step-trend approach, a step trend for the period

water years 1996-99 was included in the TSM model for
each sampling-site and constituent combination, in addition
to including trends for periods 1-4. Inclusion of a step trend
allowed evaluation of whether there was a distinct change
in data structure between pre-2000 and post-2000 data that
might have affected trend results. Results of the exploratory
step-trend analysis indicated that among all sampling-site and
constituent combinations, statistically significant step trends
were infrequently detected (less than 20 percent of analyses).
In all cases of statistically significant step trends, the differ-
ence in the percent change from the start of period 1 to the
end of period 4 between the exploratory analysis including
the step trend and the reported analysis without the step trend
was less than 5 percent. Thus, it was concluded that temporal
changes in spike recoveries did not have a substantial effect on
the overall trend results and the study objectives of evaluat-
ing relative spatial and temporal changes in FACs in the upper
Clark Fork Basin as a whole. For the exploratoiy spike-
recovery adjustment approach, constituent concentrations for
each year were adjusted by multiplying the concentrations
times the annual mean spike recovery for laboratory-spiked
stream-water samples; then exploratory trend analysis was
done. Results of the exploratory spike-recovery adjustment
analysis were similar to the results for the exploratory step-
trend approach and resulted in the same general conclusion
that temporal differences in spike recoveries had minor effects
on trend results.

An important consideration in trend analysis for trace
elements is potential effects of diel cycling in trace-element
concentrations. Complex biogeochemical processes affected
by the daily solar photocycle produce regular and dynamic
changes in many physical and chemical characteristics of
streams (Nimick and others, 2011). In some streams (including
some of the sampling sites in this study), the biogeochemical
processes can result in diel variability in trace-element concen-
trations (Nimick and others, 2003).

Diel cycling in trace-element concentrations has the
potential to affect trend results if (1) there is strong diel
cycling for a given sampling-site and constituent combination
and (2) there is a systematic temporal bias in the dataset with
respect to the time of day of sampling. During exploratory
analysis, potential effects of diel cycling on the trend results
were quantitatively evaluated by including decimal day (time
of sampling) as an ancillary' variable in the trend models. The
decimal day variable indicates the strength of diel cycling for
a given sampling-site and constituent combination and also
allows evaluation of the effect of temporal variability in time
of sampling on the trend results. Although some sampling-
site and constituent combinations had statistically significant
diel cycling, in no case did the inclusion of the decimal day
variable in trend models provide substantially different trend
results from the reported results. Thus, potential effects on
trend results of diel cycling of trace elements were determined
to be minor; however, it should be noted that samples were
collected during daylight hours and diel variations in the night
cannot be evaluated.


-------
Water-Quality Trends and Constituent-Transport Analysis for Selected Sampling Sites

Table 2-1. Statistical summaries of standard errors of estimates for the trend models.

[SEE, standard error of estimate]

Constituent or property

Number of sites for which



SEE, in percent



trend results are reported

Minimum

Mean

Maximum

Specific conductance

7

8,2

11.0

13.1

Copper, filtered

7

24,6

31.6

37.4

Copper, mifiltered-recoverable

7

38.3

48.3

60.7

Zinc, imfiltered-recoverable

7

41.0

50.7

65.7

Arsenic, filtered

7

15.2

20.8

26.7

Arsenic, imfiltered-recoverable

7

21.8

27.3

34.0

Suspended sediment

7

57.4

65.2

80.5


-------
Appendixes 65

Appendix 3—Trend-Analysis Results

For all constituents investigated, detailed results for trend
magnitudes, computed as the total percent changes in FAC
geometric means from the beginning to the end of eacli 5-year
period, are presented in tables 3-1 (for most sampling sites)
and 3-2 (for Clark Fork above Missoula, Montana [sampling
site 22]). Detailed trend results are graphically presented in
figures 3-1 through 3-7. The detailed graphical presentations
in appendix 3 present fitted trends for all constituents and
allow evaluation of the fitted trends for a given sampling site
in conjunction with FACs.


-------
Table 3-1. Flow-adjusted trend results for selected water-quality constituents and properties for selected sampling sites in the Milltown Reservoir/Clark Fork River Superfund
Site in the upper Clark Fork Basin, Montana, water years 1996-2015.

[Water year is the 12-month period from October 1 through September 30 and is designated by the year in which it ends. Values in parentheses indicate p-values for associated percentage change. Gray shading
indicates statistical significance at p-value less than 0.01. /?-value, statistical probability level; SEE, standard error of estimate; <, less than; NR, not reported]

Constituent or property

Number
of

samples

Total percentage	Total percentage	Total percentage	Total percentage	value for

change for water change for water	change for water	change for water	..	. SEE,

years 1996-2000	years 2001-5	years 2006-10	years 2011-15	overa in percent

(period 1)	(period 2)	(period 3)	(period 4)	ana ysis

Percentage of values
affected by recensoring
at study reporting level
used in the application of
the time-series model2

Silver Bow Creek at Warm Springs, Montana (sampling site 8, fig. 1, table

Specific conductance

186

-1(0.645)

-3 (0.226)

2 (0.380)

-13 (<0.001)

<0.001

10.5

0.0

Copper, filtered

186

-48 (<0.001)

-12 (0.187)

-8 (0.427)

-24(0.023)

<0.001

32.9

0.0

Copper, unfiltered-recoverable

186

-38 (<0.001)

-14 (0.105)

-12 (0.246)

-2.8(0.005)

<0.001

38.3

0.0

Zinc, unfiltered-recoverable

178

-54 (<0.001)

-47 (<0.001)

16(0.112)

-37 (<0.001)

<0.001

45.0

4.5

Arsenic, filtered

186

! (!) 902'!

5 (0,449)

5(0.48!)

-18(0.015)

0.002

24.8

0.0

Arsenic, unfiltered-recoverable

186

-1 (0.907)

5 (0.303)

I (0.894)

-16 (0.004)

0.002

24.5

0.0

Suspended sediment

188

17(0.450)

-27 (0.072)

-4!) (0.010)

15(0.515)

<0.00!

65.9

0.0

Clark Fork near Galen, Montana (sampling site 11, fig. 1, table 1)

Specific conductance

217

I (0.134)

-8 (NR3)

7 (NR3)

-12 (NR3)

0.02.7

12.7

0.0

Copper, filtered

215

-45 (<0.001)

-5 (0.593)

-17(0.085)

4 (0.759)

<0.001

28.4

0.0

Copper, unfiltered-recoverable

213

-31 (<0.001)

7(0.527)

-5 (0.702)

-24 (0.035)

<0.001

44.4

0.0

Zinc, unfiltered-recoverable

205

-56 (<0.001)

-31 (0.003)

30 (0.060)

-39 (0.001)

<0.001

41.0

4.8

Arsenic, filtered

215

-8 (0 3s2 'i

^ v -	** J

12 (0.165)

-21 (0.014)

1 1 (0.303)

<0.001

26.7

0.0

Arsenic, unfiltered-recoverable

215

-3 (0.708)

3 (0.741)

-17(0.082)

13 (0.294)

0.005

29.4

0.0

Suspended sediment

229

12 (0.494)

-19(0.211)

8 (0.678)

-25 (0.168)

0.002

60.0

0.0





Clark Fork at Deer Lodge,

Montana (sampling site 14,fig. 1, table 1)







Specific conductance

264

1 (0.747)

-4 (0.089)

-2 (0.419)

1 (0.860)

<0.00 i

11.2

0.0

Copper, filtered

231

-16 (0.003)

6 (0.400)

-12 (0.087)

8 (0.397)

<0.001

28.9

0.0

Copper, unfiltered-recoverable

229

-22 (0.019)

5 (0.661)

1 (0.963)

-8 (0.595)

<0.001

52.7

0.0

Zinc, unfiltered-recoverable

227

-37 (<0.001)

-2 (0.850)

-7 (0.560)

-13 (0.334)

<0.001

54.0

0.9

Arsenic, filtered

231

-3 (0.501)

i 7 (<0.001 I

-16 (<0.001)

4 (0.540)

0.001

15.6

0.0

Arsenic, unfiltered-recoverable

230

-8 (0.184)

7(0.308)

-12(0.114)

2 (0.828)

0.357

27.0

0.0

Suspended sediment

281

-17(0.121)

-8 (0.555)

8 (0.643)

-17(0.294)

0.00!

80.5

0.0


-------
Table 3-1. Flow-adjusted trend results for selected water-quality constituents and properties for selected sampling sites in the Milltown Reservoir/Clark Fork River Superfund
Site in the upper Clark Fork Basin, Montana, water years 1996-2015.—Continued

[Water year is the 12-month period from October 1 through September 30 and is designated by the year in which it ends. Values in parentheses indicate p-values for associated percentage change. Gray shading
indicates statistical significance at p-value less than 0.01. p-value, statistical probability level; SEE, standard error of estimate; <, less than; NR, not reported]

.	Total percentage

Number	, ,

_ ^ ,	change for water

Constituent or property of	years 1996„2000

S8mpleS	(periodl)

Total percentage	Total percentage	Total percentage

change for water change for water	change for water

years 2001-5	years 2006-10	years 2011-15

(period 2)	(period 3)	(period 4)

Percentage of values
p-value for	affected by recensoring

overall trend . ' at study reporting level
analysis'	used in the application of

the tinie-series model2

Clark Fork at Gold creek, Montana (sampling site 16, fig. 1, table 1)

Specific conductance

186

-2 (0.372)

-3 (0.063)

-2 (0.317)

0 (0.972)

<0.001

9.9

0.0

Copper, filtered

18?

-20 (0.003)

13 (0.046)

-12 (0.077)

3 (0.752)

0.002.

24.6

0.0

Copper, unfiltered-recoverable

185

-5 r(i.AKK)

-18 (0.036)

-6 (0.564)

7(0.569)

0.002

44.0

0.0

Zinc, unfiltered-recoverable

183

-2d (0.0Id)

-37 (<0.001)

24 (0.103)

-14(0.349)

<0.001

43.8

1.7

Arsenic, filtered

186

-13 (NR3)

8 (0.048)

-3 (0.548)

-4 (0.365)

0.026

15.2

0.0

Arsenic, unfiltered-recoverable

186

-17 (NR3)

3 (0.582)

-4 (0.522)

-3 (0.616)

0.086

21.8

0.0

Suspended sediment

187

15 (0.396)

-51 (O.OOl)

54 (0.012)

-17(0.352)

<0.001

58.4

0.0

Clark Fork near Drummond, Montana (sampling site 18, fig. 1, table 1)

Specific conductance

186

0(0.535)

-2(0.018)

-3 (<0.001)

6 (<0.001)

<0.001

11.3

0.0

Copper, filtered

183

0(0.991)

10 (0.037)

-2.4 (<0.001)

13(0.194)

0.013

33.5

0.0

Copper, unfiltered-recoverable

184

-13 (0.219)

-9 (0.369)

-10 (0.408)

-5 (0.730)

0.002

47.6

0.0

Zinc, unfiltered-recoverable

182

-48 (<0.001)

-18 (0.067)

12 (0.437)

-23 (0.147)

<0.001

50.7

2.2

Arsenic, filtered

186

-6 (0.093)

4(0.107)

-11 (NR3)

3 (0.378)

0.907

15.9

0.0

Arsenic, unfiltered-recoverable

186

-15(0.001)

3 (0.398)

-6 (0.171)

0 (0.930)

0.003

23.9

0.0

Suspended sediment

187

-24 (0.134)

-20 (0.174)

29 (0.190)

-23 (0.242)

0.065

65.6

0.0

Clark Fork at Turah Bridge near Bonner, Montana (sampling site 20, fig. 1, table 1)

Specific conductance

259

-5 (<0.001)

-2 (0.378)

3(0.184)

-2 (0.502)

<0.001

13.1

0.0

Copper, filtered

228

-2.3 (<0.001)

9 (0.357)

-6 CO.52.5)

-20 (0.077)

<0.001

35.0

0.0

Copper, unfiltered-recoverable

227

-13 (0.073)

-8 (0.385)

-1 (0.920)

-4 (0.762)

0.002.

50.3

0.0

Zinc, unfiltered-recoverable

219

-36 (<0.001)

-32 (0.005)

52 (0.004)

-31 (0.026)

<0.00!

55.1

5.0

Arsenic, filtered

229

-5 (<0.00 i)

5 (0.002.)

3 (0.435)

-16(0.002)

<0.00 i

21.6

0.0

Arsenic, unfiltered-recoverable

229

-10(0.051)

-1 (0.879)

9 (0.258)

-16(0.052)

0.204

30.3

0.0

Suspended sediment

284

-13 (0.222)

-25 (0.059)

36 (0.067)

-21 (0.246)

0.002

57.4

0.0

'Determination of and distinction between />-value for individual trend period and/>-value for overall trend analysis are discussed in the section of this report "Appendix 2—Summary of the Time-Series Model as Applied in this Study."
Procedures for determining and applying the study reporting level used in the application of the time-series model are discussed in the section of this report "General Description of the Time-Series Model."

3Results not reported because of nonsignificant overall trend analysis (p-value greater than 0.01).


-------
Table 3-2. Flow-adjusted trend results for selected water-quality constituents and properties for Clark Fork above Missoula, Montana (sampling site 22), water years 1996-2015.

[Water year is the 12-month period from October 1 through September 30 and is designated by the year in which it ends. Values in parentheses indicate p-values for associated percentage change. Gray shading
indicates statistical significance at p-value less than 0.01. /?-value, statistical probability level; SEE, standard error of estimate; <, less than; NR, not reported]

Constituent or property	.

samples

Clark Fork above Missoula, Montana (sampling site 22, fig.1, table 1)

Specific conductance

227

0 (0.840)

-2 (0.250)

1 (0.585)

4(0.101)

-7 (NR3)

0.161

8.2

0.0

Copper, filtered

206

-25 (0.006)

25 (0.057)

13 (0.357)

-21 (0.089)

-27(0.032)

0.001

37.4

0.0

Copper, unfiltered-recoverable

205

-23 (0.035)

41 (0.017)

i 20 (<0.001)

-59 (<0.001)

-52 (0.002)

<0.001

60.7

0.0

Zinc, unfiltered-recoverable

186

-49 (<0.001)

43 (0.082)

192. (<0.001)

-65 (<0.001)

-r>2 (0.003)

<0.001

65.7

8.5

Arsenic, filtered

207

-15 (0.005)

14(0.033)

10(0.171)

-3 (0.664)

-24 (<0.00 i)

<0.001

26.1

0.0

Arsenic, unfiltered-recoverable

207

-2.1 (0.006)

16(0.110)

25 (0.036)

-17(0.099)

-25 (0.012)

<0.001

34.0

0.0

Suspended sediment

250

-4 (0.796)

25 (0.242)

168 (<0.001)

-60 (<0.001)

-40 (0.032.)

<0.001

68.7

0.0

determination of and distinction between p-v alue for individual trend period and p-v alue for overall irend analysis arc discussed in the section of this report "Appendix 2—Summary of the Time-Series
Model as Applied in this Study."

Procedures for determining and applying the study reporting level used in the application of the time-series model are discussed in the section of this report "General Description of the Time-Series Model."
3Results not reported because of nonsignificant overall trend analysis (p-value greater than 0.01).

Total
percentage
change for
water years

1996-2000
(period 1)

Total
percentage
change for
water years

2001-5
(period 2)

Total
percentage
change for
October 1,2005-
March 27,2008
(period 3A)

Total
percentage
change for
March 28,2008-
Septeniber 30,
2010
(period 3B)

Total
percentage
change for
water years
2011-15
(period 4)

p-value for
overall trend
analysis'

Percentage of values
affected by
SEE, recensoring at study
in percent reporting level used in
the application of the
time-series model2


-------
Appendixes 69

1,000

Period

12	3	4

100
1,000

100

E 10



Specific conductance (|xS/cm)







%

o

0

° 0





iqSkfi

°Ofl i

; o_





-











-



O /

s°o7

" o 9
o /

o °?



521

514
... I

501

513
... I

1

	.CO —

•53-

1,000

100

:

,,, 11,,, i,,,, i,,,, i

Filtered copper (microgram per liter)

,, ,, 1 E

-

8.9 4.6

4:

3.8

2.9-

r



o

'a _o°0A«





0(5

o





1

1,000

100

—I—I—I—I—I—I—I—I—I—I—I—I—I—I—I—I—I—I—I—I—I—I—I—I—I I

Suspended sediment (milligram per liter)

rDOC o o

0°° " oo o " o

J	i_i	i	i 1m	i	l2i	1_5

[Graph included as a place holder to assist in
comparisons.]

1

1,000

100

:

... 1

, , , , |, , , , |

,,,, i,,, 11 _

:

Unfiltered-recoverable

:

-

copper (microgram per liter)

-

-

15

9.3 79

7 0 5.0

r

8 \
t

... i

0 \ 0 1

O O \ '

|0 ° „ \ o _
.... i.... i

° o 1 ° ° 1
o o ® &

, . . . 1 . . . . 1

EXPLANATION

[Water year is defined as the
12-month period from October 1
through September 30 and is
designated by the year in which
it ends. |jS/cm, microsiemen
per centimeter at 25 degrees
Celsius; p-value, statistical
probability level]

o Flow-adjusted

concentration (FAC)
determined by
using the time-series
model

— Flow-adjusted fitted
trend determined
by using the time-
series model

521 Fitted trend value at

start or end of period

446 Bold values indicate

statistical significance
(p-value less than 0.01)
for period before value
presented in bold

1,000

100











i i . I _

:

Filtered arsenic





:

-

(microgram per liter





-

:



0





—

"



o

°



-

-

Wok



P7

00 "J

oxi



19

19

20

21

17:
... i

1

1,000

100

_

i i i 1 i i i i 1

1

1

1 1 1 1E

:

Unfiltered-recoverable



:

-

zinc (microgram per liter)



-

-



O

o



° o -

-





o

CP

)fl° o

:



o

&

k

o r
/c



35 16

1 1 1 I 1 1 1 1 1

oo

9.8

1 1 1 1

6.1-

¦ i . i

T~\

1

1

1

i i i i 1

I I I | -

:

Unfiltered-recoverable



:

:

arsenic (microgram per liter)



:

i-



o





-

:

o 0





8.^° „



.



_
a

lT oBWflu

ODQ rj-jjAn-,









S>

o

JO

-

7

7

/

7

	

22

¦ i i 1

/

22

¦ i i 1

23

¦ i i 1

23

1 1 1 1 1

19:

i i i I

Water year (October-September)

Figure 3-1. Flow-adjusted fitted trends for selected water-quality constituents and properties for Silver Bow Creek at Warm Springs,
Montana (sampling site 8), wateryears 1996-2015.


-------
70 Water-Quality Trends and Constituent-Transport Analysis for Selected Sampling Sites

1,000

Period

2	3

100
1,000

100

-

Specific conductance (fxS/cm)





-

0

oo c

0

O ^ 0



-



° CfljLoh°

°dS

2o,4°y



-



o o 0 St
o



Wy



-



o /

0 °

/



° 8T-



447

454

415

443

CO
CO

— CO -

1



Filtered copper (microgram per liter)



-

7.6 4.2

4.0

3.3

3.4-

-

I o





1 ~

_

o A

o





!





^ q aim £.

o



C

o









1 1 1 1 1 1 1 1 1

1 1 1 1 1

i i i i 1

, . ,0, 1

[Graph included as a place holder to assist in
comparisons.]

1,000

100

Period

2	3



Suspended sediment (milligram per liter)



r

5.2
\

5.8

4.7

5.1

3.8:

r

oo V

1

° 1

% 01
° „

0

' 0-



oO^

o

, . . 1

, , , 1

0 _ c
°° ,

*f«5!

l,UUU















:

Unfiltered-recoverable



1



-

copper (microgram per liter)



-

100

-

15

11

11

11

8.1_



¦

0 b

°\

O 1
O

0 \
0 A

op y
°J3 0

\

1U





o®8
1 1 1 1 1

° 0
1 1 1 I

0

0

0

w Q

1













1 ,uuu















:

Unfiltered-recoverable



:



-

zinc (microgram per liter)



-

100

r









-



:

°ft) 00 0

0 0

0

0 0
% 0

A°°° '

10

:





0 /

°

0

:

10

r

0 X

0 I

sPo O0 (S

0



10

r

/o



O®

0 f

0 -





12

11

13

, , , 1

10

11:
, , , 1





15

14

, , , 1

15

... 1

12

14;

Wateryear (October-September)

Figure 3-2. Flow-adjusted fitted trends for selected water-quality constituents and properties for Clark Fork near Galen, Montana
(sampling site 11), wateryears 1996-2015.


-------
Appendixes 71

[Graph included as a place holderto assist in
comparisons.]

EXPLANATION

1	1	1	1	1	1	1	1	1	1	1	1	1	1—

Specific conductance (|iS/cm)

t—i—i—|—i—i—i—i—|—i—i—i—i—[~

Unfiltered-recoverable
copper (microgram per liter)

1,000^ I I I I I I I I I I I I I I I I I I I I I I I I I

Filtered copper (microgram per liter)

Period

12	3	4

I I	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1 I

= Suspended sediment (milligram per liter)	=

Period

1	2

Unfiltered-recoverable
zinc (microgram per liter)

: Unfiltered-recoverable
arsenic (microgram per liter)

[Wateryear is defined as the
12-month period from October 1
through September 30 and is
designated by the year in which
it ends. |jS/cm, microsiemen
per centimeter at 25 degrees
Celsius; p-value, statistical
probability level]

o Flow-adjusted

concentration (FAC)
determined by
using the time-series
model

— Flow-adjusted fitted
trend determined
by using the time-
series model

479 Fitted trend value at

start or end of period

5.8 Bold values indicate

statistical significance
(p-value less than 0.01)
for period before value
presented in bold

I,uuu



Filtered arsenic
(microgram per liter







100

—









-



-





.AO





10

:















11

11

13

11

11:

Wateryear (October-September)

Figure 3-3. Flow-adjusted fitted trends for selected water-quality constituents and properties for Clark Fork at Deer Lodge, Montana
(sampling site 14), wateryears 1996-2015.


-------
72 Water-Quality Trends and Constituent-Transport Analysis for Selected Sampling Sites

Period

1,000

100
1,000

100

E 10

12	3	4

1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1-

Specific conductance (jitS/cm)

	I	i	i	i	i	I	i	i	i	i	I	i	i	i	i	I	i	i	i	i	L

~i—i—i—i—|—i—i—i—i—|—i—i—i—i—|—i—i—i—i—|—i—i—i—i—p

Filtered copper (microgram per liter)

4.8	3.8 4.3 3.8 I 3.9.

°o <&o ^ 6®
j	i	i	i	I	i	i	i	i	I	i	i	i	i	I	i	i	i	i	I	i	i	i	i	L

Period

J	2	3	4

1,000 e-!—1—1—1—i—(—1—1—1—i—(—1—1—1—i—1—1—1—1—i—1—1—1—1—r=

Suspended sediment (milligram per liter)



100

1

1,000



—i—i—i—|—

—1—1—1—1—

—I—I—I—|—

—I—I—I—1—

1 I

:

Unfiltered-recoverable



:

-

copper (microgram per liter)



-



19

19

15

14

15



°







h

=

m

cP

o

1 1 1 1

Q 0 \

J

o I

0 O ft



o ^



i i i 1

o o

1 1 1 1 1

ocRid<^

0

1	1 1 1 1

° °o Q

¦ i i 1

EXPLANATION

[Wateryear is defined as the
12-month period from October 1
through September 30 and is
designated by the year in which
it ends. |jS/cm, microsiemen
per centimeter at 25 degrees
Celsius; p-value, statistical
probability level]

o Flow-adjusted

concentration (FAC)
determined by
using the time-series
model

— Flow-adjusted fitted
trend determined
by using the time-
series model

425 Fitted trend value at

start or end of period

8.3 Bold values indicate

statistical significance
(p-value less than 0.01)
for period before value
presented in bold

p i—i—i—i—|—i—i—i—i—|—i—i—i—i—|—i—i—i—i—|—i—i—i—i—[—

Unfiltered-recoverable
zinc (microgram per liter)

[Graph included as a place holder to assist in
comparisons.]

100 r

10 =

1,000

100

n—i—i—|—i—i—i—i—|—i—i—i—i—|—i—i—i—i—|—i—i—i—i—j i

Filtered arsenic
(microgram per liter)

_i	I	I	I	I	I	I	I	I	I	I	I	I	!	I	I	I	I	I	l	I	I	I	

I ,uuu











1 1 1 I :



=

Unfiltered-recoverable



:



-

arsenic (microgram per liter)



"

100

-









-



-

o o

0





"







0

0 _

o_g

y° -8

10

F

W

°°

O™



0







12

10

10

10

9.7;

Water year (October-September)

Figure 3-4. Flow-adjusted fitted trends for selected water-quality constituents and properties for Clark Fork at Goldcreek, Montana
(sampling site 16),wateryears 1996-2015.


-------
Appendixes 73

1,000

100
1,000

S> 100

Period

1	2	3	4

Period



1 1 1 I 1 1 1 1 I 1 1 1 1 I 1 1 1 1 I

Specific conductance (jxS/cm)

1 1 1 I







0



-

Ml



° °

o, -

-

° /

•fw)

o o

° / ° /

O o /



/ o









461

o 459

449 434

1

CO
•53-

1,000 E 1

100 E~

Suspended sediment (milligram per liter)

10 r

16

13

° 1

UP0 ° ° 1

° 1

f, 0 % T
CU® ° ]

° 8

c

o

o

o o

_

1 1 1 1

1 1 1 1 1

1 1 1 1

1 1 1 1 1

1 1

1 1

1,000

_

I I I |

1 1 1 1 1

1 1 1 1

1 1 1

1 1

i i i | .

:

Filtered copper (microgram per liter)







:

Unfiltered-recoverable





:

-















-

copper (microgram per liter)





-



3.9

3.9

4.3

3.3



3.7





17

15

14



13

12

:













100

:

0

o

0 o °\
oo\

O ffi

o °

\

o \



o '

















rr O n ^
O O ^ o

s o-





^ a#!8*!

-



o '

o 0

o





10

-

o











©>°c
o

	

% oj

° C
... 1

o®oo^



VTcPO1

0

1	1



	

o

I I I 1

1 I I I 1

o °

1 1 1 1

0

0

1	1 1

i I

1 1 1 1 1

EXPLANATION

[Wateryear is defined as the
12-month period from October 1
through September 30 and is
designated by the year in which
it ends. (jS/cm, microsiemen
per centimeter at 25 degrees
Celsius; p-value, statistical
probability level]

o Flow-adjusted

concentration (FAC)
determined by
using the time-series
model

— Flow-adjusted fitted
trend determined
by using the time-
series model

461 Fitted trend value at

start or end of period

434 Bold values indicate

statistical significance
(p-value less than 0.01)
for period before value
presented in bold

[Graph included as a place holderto assist in
comparisons.]

1,000

100

F~i—i—i—i—|—i—i—i—i—|—i—i—i—i—|—i—i—i—i—|—i—i—i—r

n

Unfiltered-recoverable
zinc (microgram per liter)
o

j	i	i	i	i	i	i	i	i	i	i	i	i	i	i	i	i	i	i	i	i	i	i	i_

1,000

100

1

1 1 1 l

1 1 1 1

1 1 1 1

1 1 1 1

1 1 1 | .

1,000













=

Filtered arsenic





:



=

Unfiltered-recoverable



:

¦

(microgram per liter





¦



~

arsenic (microgram per liter)



"

=-









-

100

=-









-

-









¦



-

O OD







"



0 <33





qO

0







0

rr^T ifiri0 if?

0 0 0

DjpOjfPjfiai

0 0

<&_° -ft

r









rim iifTS- °T^

10

r

O _ /





OQ
-------
74 Water-Quality Trends and Constituent-Transport Analysis for Selected Sampling Sites

1,000

Period

12	3	4

n—i—i—i—|—i—i—i—i—|—i—i—i—i—|—i—i—i—i—|—i—i—i—i—p

Specific conductance (|xS/cm)

347	330	324 334	327

_l	I	I	I	I	I	I	I	I	I	I	I	I	I	I	I	I	I	I	I	I	I	I	I	

I	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	J-

Filtered copper (microgram per liter)

E 10

1,000

100

Period

12	3	4

1

1,000

100

=-1

i i i I

1 1 1 1 I

—I—I—I—1—

—I—I—I—1—

—I—I—I—[—=

I

Suspended sediment (milligram per liter)



-

13

12

8.8

12

9.5"

E





° 1



~E

1

°

8 °

0 o '
<§>

o '
% 8 c
i ° n Q

lJs

o t
° o C[

F

§> cQ)
0







° 8 o
-------
Appendixes 75

Period

Period

Specific conductance (jitS/cm)

Filtered copper (microgram per liter)

Unfiltered-recoverable
copper (microgram per liter)

Unfiltered-recoverable
zinc (microgram per liter)

[Graph included as a place holderto assist in
comparisons]

Unfiltered-recoverable
arsenic (microgram per liter)

EXPLANATION

[Wateryear is defined as the
12-month period from October 1
through September 30 and is
designated by the year in which
it ends. |jS/cm, microsiemen
per centimeter at 25 degrees
Celsius; p-value, statistical
probability level]

o Flow-adjusted

concentration (FAC)
determined by
using the time-series
model

	 Flow-adjusted fitted

trend determined
by using the time-
series model

277 Fitted trend value at

start or end of period

25 Bold values indicate

statistical significance
(p-value less than 0.01)
for period before value
presented in bold

Water year (October-September)

Filtered arsenic
(microgram per liter)

Figure 3-7. Flow-adjusted fitted trends for selected water-quality constituents and properties for Clark Fork above Missoula, Montana
(sampling site 22), water years 1996-2015.


-------
76 Water-Quality Trends and Constituent-Transport Analysis for Selected Sampling Sites

Appendix 4—Transport-Analysis Balance Calculations for Data-Summary Reaches

Balance calculations for the transport analysis (that is,
differences between reach inflows and reach outflows) are pre-
sented in tables 4-1 through 4-6 for reaches 4-9, respectively,
in appendix 4. The transport balance calculations indicate
within-reach changes in estimated normalized loads and allow
assessment of temporal changes in relative contributions from
upstream source areas to loads transported past each reach
outflow.


-------
Appendixes 77

Table 4-1. Constituent-transport analysis balance calculations for sampling sites in reach 4, extending from Silver Bow Creek at Warm
Springs, Montana (sampling site 8), to Clark Fork near Galen, Montana (sampling site 11), for selected periods, water years 1996-2015.

[Water year is the 12-month period from October 1 through September 30 and is designated by the year in which it ends]

Abbreviated sampling site name (table 1) and number or summation category

Estimated normalized load,1
in kilograms per day

Unfiltered- Unfiltered- _ . .

,,	,, Suspended

recoverable recoverable

sediment

copper arsenic

Water years 1996-2000 (period 1)

Inflow

Silver Bow Creek at Warm Springs (sampling site 8)

Outflow

Clark Fork near Galen (sampling site 11)

1.9

3.7

3.4

4.2

920

1,600

Total within-reach change in load—oiitilow (sampling site 11) minus inflow (sampling siie 8)

(positive values indicate nei mobilization from all within-reach sources including groundwater	i .8	0.78	670

inflow, tributaries, the main-stem channel, and Hood plain)

Water years 2001-5 (period 2)

Inflow

Silver Bow Creek at Warm Springs (sampling site 8)	1.4	3.5	850

Outflow

Clark Fork near Galen (sampling site 11)	3 \	4 2 1500

Total within-reach change in load—outflow (sampling siie I!) minus inflow (sampling siie 8)

(positive values indicate nel mobilization from all w ithin-reach sources including groundwaler	1.8	0.70	670

inflow, tributaries, the main-stem channel, and Hood plain)

Wateryears 2006-10 (period 3)

Inflow

Silver Bow Creek at Warm Springs (sampling site 8)	1.2	3.6	570

Outflow

Clark Fork near Galen (sampling site 11)	-$.2	3.9 1 400

Total within-reach change in load—outflow (sampling site 11) minus inflow (sampling siie 8)

(positive values indicate nei mobilization from all within-reach sources including groundwater	2.0	0.31	860

inflow, tributaries, the main-stem channel, and Hood plain)

Wateryears 2011-15 (period 4)

Inflow

Silver Bow Creek at Warm Springs (sampling site 8)	0.94	3.3	460

Outflow

Clark Fork near Galen (sampling site 11)	27	381 300

Total within-reach change in load—outflow (sampling site 11) minus inllow (sampling site 8)

(positive values indicate net mobilization from all wilhin-reach sources including groundwater	1.8	0.46	82.0

inllow. tributaries, the main-siem channel, and flood plain)

'The estimated normalized load was computed by multiplying the mean annual fitted trend concentration (determined by using the time-series model) for the
indicated period times the geometric mean streamflow for water years 1996-2015 and a units conversion factor according to equation 1 in the section of this
report "Estimation of Normalized Constituent Loads." Loads are reported to two significant figures; however, before final rounding, calculations used three

significant figures when necessary. As a result, some of the load values have minor rounding artifacts.


-------
78 Water-Quality Trends and Constituent-Transport Analysis for Selected Sampling Sites

Table 4-2. Constituent-transport analysis balance calculations for sampling sites in reach 5, extending from Clark Fork near Galen,
Montana (sampling site 11), to Clark Fork at Deer Lodge, Montana (sampling site 14), for selected periods, water years 1996-2015.

[Water year is the 12-month period from October 1 through September 30 and is designated by the year in which it ends]

Abbreviated sampling site name (table 1) and number or summation category

Estimated normalized load,1
in kilograms per day

Unfiltered- Unfiltered- _ . .

,,	,, Suspended

recoverable recoverable

sediment

copper arsenic

Water years 1996-2000 (period 1)

Inflow

Clark Fork near Galen (sampling site 11)
Outflow

Clark Fork at Deer Lodge (sampling site 14)

3.7

13

4.2

7.7

1,600

8,300

Total within-reaeh change in load—oiitilow (sampling site 14) minus inflow (sampling siie 11)
(positive values indicate nei mobilization from within-reaeh sources including groundwater
inflow, tributaries, the main-stem channel, and llood plain)

0.8

6.70!)

Water years 2001-5 (period 2)

Inflow

Clark Fork near Galen (sampling site 11)	3.1	4.2 1,500

Outflow

Clark Fork at Deer Lodge (sampling site 14)	12	7 6 1200

Total within-reaeh change in load—outfiow (sampling siie 14) minus inflow (sampling site I i)

(positive values indicate net mobilization from within-reaeh sources including groundwater	9.0	3.4 5,700

inflow, tributaries, the main-stem channel, and flood plain)

Wateryears 2006-10 (period 3)

Inflow

Clark Fork near Galen (sampling site 11)	3.2 3.9 1,400

Outflow

Clark Fork at Deer Lodge (sampling site 14)	13 747 200

Total within-reaeh change in load—outfiow (sampling site 14) minus inflow (sampling siie 11)

(positive values indicate nei mobilization from within-reaeh sources including groundwater	9.4	3.5 5.80!)

inflow, tributaries, the main-siem channel, and flood plain)

Wateryears 2011-15 (period 4)

Inflow

Clark Fork near Galen (sampling site 11)	2.7	3.8 1,300

Outflow

Clark Fork at Deer Lodge (sampling site 14)	12	7 0 6 800

Total within-reaeh change in load—outflow (sampling site 14) minus inflow (sampling site 11)

(positive values indicate net mobilization from within-reaeh sources including groundwater	9.4	3.3 5.500

inflow, tributaries, the main-siem channel, and flood plain)

'The estimated normalized load was computed by multiplying the mean annual fitted trend concentration (determined by using the time-series model) for the
indicated period times the geometric mean streamflow for water years 1996-2015 and a units conversion factor according to equation 1 in the section of this
report "Estimation of Normalized Constituent Loads." Loads are reported to two significant figures; however, before final rounding, calculations used three
significant figures when necessary. As a result, some of the load values have minor rounding artifacts.


-------
Appendixes 79

Table 4-3. Constituent-transport analysis balance calculations for sampling sites in reach 6, extending from Clark Fork at Deer Lodge,
Montana (sampling site 14), to Clark Fork at Goldcreek, Montana (sampling site 16), for selected periods, water years 1996-2015.

[Water year is the 12-month period from October 1 through September 30 and is designated by the year in which it ends]

Abbreviated sampling site name (table 1) and number or summation category

Estimated normalized load,1
in kilograms per day

Unfiltered- Unfiltered- _ . .

,,	,, Suspended

recoverable recoverable

sediment

copper arsenic

Water years 1996-2000 (period 1)

Inflow

Clark Fork at Deer Lodge (sampling site 14)
Outflow

Clark Fork at Goldcreek (sampling site 16)

13

19

7.7

11

8,300

16,000

Total within-reaeh change in load—oulfiow (sampling site 16) minus inflow (sampling siie 14)

(positive values indicate nei mobilization from all within-reaeh sources including groundwater	5.4	3.5 7.50!)

inflow, tributaries, the main-stem channel, and Hood plain)

Water years 2001-5 (period 2)

Inflow

Clark Fork at Deer Lodge (sampling site 14)	12	7.6 7,200

Outflow

Clark Fork at Goldcreek (sampling site 16)	yj	jq	12 000

Total within-reaeh change in load—outflow (sampling siie 16) minus inflow (sampling site 14)

(positive values indicate net mobilization from all within-reaeh sources including groundwater 4.6	2.6 5.000

inflow, tributaries, the main-stem channel, and flood plain)

Wateryears 2006-10 (period 3)

Inflow

Clark Fork at Deer Lodge (sampling site 14)	13	7.4 7,200

Outflow

Clark Fork at Goldcreek (sampling site 16)	15	jq	10 000

Total within-reaeh change in load—outflow (sampling site 16) minus inflow (sampling siie 14)

(positive values indicate nei mobilization Irom all wilhin-reach sources including groundwater 2.2	2.8 3,20!)

inflow, tributaries, the main-stem channel, and flood plain)

Wateryears 2011-15 (period 4)

Inflow

Clark Fork at Deer Lodge (sampling site 14)	12	7.0 6,800

Outflow

Clark Fork at Goldcreek (sampling site 16)	15	9 9 12 000

Total within-reaeh change in load—outflow (sampling site 16) minus inflow (sampling site 14)

(positive values indicate net mobilization from all within-reaeh sources including groundwater	2.7	2.8 4.900

inflow, tributaries, the main-siem channel, and flood plain)

'The estimated normalized load was computed by multiplying the mean annual fitted trend concentration (determined by using the time-series model) for the
indicated period times the geometric mean streamflow for water years 1996-2015 and a units conversion factor according to equation 1 in the section of this
report "Estimation of Normalized Constituent Loads." Loads are reported to two significant figures; however, before final rounding, calculations used three

significant figures when necessary. As a result, some of the load values have minor rounding artifacts.


-------
80 Water-Quality Trends and Constituent-Transport Analysis for Selected Sampling Sites

Table 4-4. Constituent-transport analysis balance calculations for sampling sites in reach 7, extending from Clark Fork at Goldcreek,
Montana (sampling site 16), to Clark Fork near Drummond, Montana (sampling site 18), for selected periods, water years 1996-2015.

[Water year is the 12-month period from October 1 through September 30 and is designated by the year in which it ends]

Abbreviated sampling site name (table 1) and number or summation category

Estimated normalized load,1
in kilograms per day

Unfiltered- Unfiltered- _ . .

,,	,, Suspended

recoverable recoverable

sediment

copper arsenic

Water years 1996-2000 (period 1)

Inflow

Clark Fork at Goldcreek (sampling site 16)
Outflow

Clark Fork near Drummond (sampling site 18)

19

24

11

16

16,000

26,000

Total within-reach change in load—oulfiow (sampling site 18) minus inflow (sampling siie 16)
(positive values indicate nei mobilization from all wilhin-readi sources including groundwater
inflow, tributaries, ihe main-stem channel, and Hood plain)

4.6

5.2

10.00!)

Water years 2001-5 (period 2)

Inflow

Clark Fork at Goldcreek (sampling site 16)	17	10	12,000

Outflow

Clark Fork near Drummond (sampling site 18)	21	15 21 000

Total within-reach change in load—outflow (sampling siie 18) minus inflow (sampling site 16)

(positive values indicate net mobilization from all within-reach sources including groundwater 4.1	5.0 8.300

inflow, tributaries, the main-stem channel, and Hood plain)

Wateryears 2006-10 (period 3)

Inflow

Clark Fork at Goldcreek (sampling site 16)	15	10	10,000

Outflow

Clark Fork near Drummond (sampling site 18)	19	1521 000

Total within-reach change in load—outliow (sampling site 18) minus inflow (sampling siie 16)

(positive values indicate nei mobilization Irom all within-reach sources including groundwater 4.3	4.8 10.00!)

inflow, tributaries, ihe main-siem channel, and Hood plain)

Wateryears 2011-15 (period 4)

Inflow

Clark Fork at Goldcreek (sampling site 16)	15	9.9 12,000

Outflow

Clark Fork near Drummond (sampling site 18)	jg	14	21 000

Total within-reach change in load—outliow (sampling site 18) minus inllow (sampling site 16)

(positive values indicate net mobilization from all within-reach sources including groundwater	2.9	4.6 9,100

inflow, tributaries, ihe main-siem channel, and Hood plain)

'The estimated normalized load was computed by multiplying the mean annual fitted trend concentration (determined by using the time-series model) for the
indicated period times the geometric mean streamflow for water years 1996-2015 and a units conversion factor according to equation 1 in the section of this
report "Estimation of Normalized Constituent Loads." Loads are reported to two significant figures; however, before final rounding, calculations used three
significant figures when necessary. As a result, some of the load values have minor rounding artifacts.


-------
Appendixes 81

Table 4-5. Constituent-transport analysis balance calculations for sampling sites in reach 8, extending from Clark Fork near
Drummond, Montana (sampling site 18), to Clark Fork atTurah Bridge near Bonner, Montana (sampling site 20), for selected periods,
water years 1996-2015.

[Water year is the 12-month period from October 1 through September 30 and is designated by the year in which it ends]

Abbreviated sampling site name (table 1) and number or summation category

Estimated normalized load,1
in kilograms per day

Unfiltered- Unfiltered- _ . .

,,	,, Suspended

recoverable recoverable

sediment

copper arsenic

Water years 1996-2000 (period 1)

24

25

1.6

Water years 2001-5 (period 2)

21

22

1.5

Wateryears 2006-10 (period 3)

19

21

2.3

16

1.5

26,000

33.000

Inflow

Clark Fork near Drummond (sampling site 18)

Outflow

Clark Fork al Turah Bridge (sampling site 20)

Total within-reach change in load—oulllow (sampling site 20) minus inflow (sampling site 18)
(negative values indicate net accumulation in reach channel: positive values indicate net
mobilization from all withiii-reach sources including groundwater infiow. tributaries, the
main-stan channel, and Hood plain)

0.49 6.30!)

Inflow

Clark Fork near Drummond (sampling site 18)

Outflow

Clark Fork at Turah Bridge (sampling site 20)

Total within-reach change in load—oulllow (sampling site 20) minus inflow (sampling site 18)
(negative values indicate nei accumulation in reach channel: positive values indicate net
mobilization from ail within-reach sources including groundwater inflow, tributaries, the
main-stem channel, and llood plain)

15 21,000

16 26,000

0.58 5.900

Inflow

Clark Fork near Drummond (sampling site 18)

Outflow

Clark Fork at Turah Bridge (sampling site 20)

Total within-reach change in load—oulllow (sampling site 20) minus inflow (sampling site 18)
(negative values indicate net accumulation in reach channel: positive values indicate net
mobilization from all within-reach sources including groundwater infiow. tributaries, the
main-stan channel, and flood plain)

15 21,000

16 27,000

5.800

Wateryears 2011-15 (period 4)

18

21

3.2.

14

16

1.3

Inflow

Clark Fork near Drummond (sampling site 18)

Outflow

Clark Fork at Turah Bridge (sampling site 20)

Total within-reach change in load—oulllow (sampling site 20) minus inflow (sampling site 18)

(negative values indicate net accumulation in reach channel: positive values indicate net
mobilization from all within-reach sources including groundwater inllow. tributaries, the
main-stem channel, and flood plain)

'The estimated normalized load was computed by multiplying !bc mean annual filled Irend eoneeniraiion (determined by using ihe lime-series model) lor I be
indicated period limes Ihe geometric mean slreamllow lor v. aler years 1996 201 ? and a unils conversion I actor according lo equal ion I In ibe seel ion ol'lbis
report "Estimation of Normalized Constituent Loads." Loads are reported to two significant figures; however, before final rounding, calculations used three
significant figures when necessary. As a result, some of the load values have minor rounding artifacts.

21,000

28,000

6.900


-------
82 Water-Quality Trends and Constituent-Transport Analysis for Selected Sampling Sites

Table 4-6. Constituent-transport analysis balance calculations for sampling sites in reach 9, extending from Clark Fork at Turah Bridge
near Bonner, Montana (sampling site 20), to Clark Fork above Missoula, Montana (sampling site 22), for selected periods, water years

1996-2015.

[Water year is the 12-month period from October 1 through September 30 and is designated by the year in which it ends]

Abbreviated sampling site name (table 1) and number or summation category

Estimated normalized load,1
in kilograms per day

Unfiltered- Unfiltered- _ . .

,,	,, Suspended

recoverable recoverable ,.

sediment

copper arsenic

Water years 1996-2000 (period 1)

Inflow

Clark Fork at Turah Bridge (sampling site 20)
Outflow

Clark Fork above Missoula (sampling site 22)

25

2.9

17

33,000

39.000

Total within-reach change in load—oiitfiow (sampling site 22.) minus inllow (sampling siie 20)
(positive values indicate net mobilization from all wilhin-reach sources including groundwater
inflow, tributaries, the main-stem channel, and Hood plain)

3.7

2.5

6.000

Water years 2001-5 (period 2)

Inflow

Clark Fork at Turah Bridge (sampling site 20)	22	16 26,000

Outflow

Clark Fork above Missoula (sampling site 22)	30	jg 42 000

Total within-reaeh change in load—outfiow (sampling siie 22) minus inflow (sampling site 2.0)

(positive values indicate net mobilization from all wilhin-reach sources including groundwater	7.7	2.6 16,000

inflow, tributaries, the main-stem channel, and flood plain)

Wateryears 2006-10 (period 3)

Inflow

Clark Fork at Turah Bridge (sampling site 20)	21	16 27,000

Outflow

Clark Fork above Missoula (sampling site 22)	54	22	83 000

Total within-reach change in load—outfiow (sampling site 2.2.) minus inflow (sampling siie 20)

(positive values indicate net mobilization from all wilhin-reach sources including groundwater 32	5.9 56.000

inflow, tributaries, the main-siem channel, and flood plain)

Wateryears 2011-15 (period 4)

Inflow

Clark Fork at Turah Bridge (sampling site 20)	21	16 28,000

Outflow

Clark Fork above Missoula (sampling site 22)	23	18 40 000

Total within-reach change in load—outflow (sampling site 22) minus inflow (sampling site 20)

(positive values indicate net mobilization from all wilhin-reach sources including groundwater	2.2	2. i 12.000

inflow, tributaries, the main-stem channel, and flood plain)

'The estimated normalized load was computed by multiplying the moan annual fitted trend concentration (determined by using the time-series model) lor the
indicated period limes the geometric mean stream Mow lor v. ater years 1996 201 ? and a units conversion I actor according to equation I In the section ol'this
report "Estimation of Normalized Constituent Loads." Loads are reported to two significant figures; however, before final rounding, calculations used three
significant figures when necessary. As a result, some of the load values have minor rounding artifacts.


-------
Publishing support provided by:

Rolla and Lafayette Publishing Service Centers

For more information concerning this publication, contact:
Director, Wyoming-Montana Water Science Center
U.S. Geological Survey
3162 Bozeman Ave
Helena, MT 59601
(406)457-5900

Or visit the Wyoming-Montana Water Science Center Web site at:
http://wy-mt.water.usgs.gov/


-------
ISSN 2328-0328 (online)

http://dx.doi.Org/10.3133/sir20165100


-------