HEALTH CONSULTATION
Environmental Media Investigation Report for Arrow Stone Park in Deer Lodge, Montana
Clark For Operable Unit of the Milltown Reservoir Sediments
National Priorities List Site
MILLTOWN RESERVOIR SEDIMENTS
MILLTOWN, MISSOULA COUNTY, MONTANA
STATEMENT OF ISSUES AND BACKGROUND
Historic mining activities upriver have resulted in the presence of contaminants, including arsenicand other heavy metals, in and along the Clark Fork River (CFR) between Anaconda and Missoula,Montana. This stretch of the CFR has been listed on the National Priorities List as MilltownReservoir Sediments. The U.S. Environmental Protection Agency (EPA) performed a BaselineHuman Health Risk Assessment (HHRA) on the CFR to characterize the range of potential risks torecreational users of the CFR attributable to these mine wastes [1]. Since the HHRA was released,local agencies in Deer Lodge, Montana developed Arrow Stone Park, a public park along the CFRwith different usage patterns than those evaluated in the HHRA. In a health consultation on ArrowStone Park released in March 2001, ATSDR concluded that the existing data do not adequatelyrepresent current site conditions and recommended further characterization of the arsenic levels insoil at Arrow Stone Park to assist in public health decision-making [2].
In July 2000, ATSDR offered resources to collect surface soil samples and test them for arsenic.Before agreeing to the sampling, Powell County requested that ATSDR and EPA reach anagreement on the need for testing and the appropriate risk-based concentration (RBC) to use formaking decisions based on the results. In July 2001, EPA released a provisional addendum to theHHRA which developed RBCs for exposure of park visitors to arsenic in soil and tailings; thisaddendum has since been finalized [3]. ATSDR concurred with the chronic RBC of 684 parts permillion (ppm) for child recreational users in a letter and draft public health consultation [4], and the sampling was approved by Powell County.
A sampling and analysis plan describing the proposed sampling methodology was developed byATSDR with input from EPA, Montana Department of Environmental Quality (DEQ), MontanaDepartment of Public Health and Human Services (DPHHS), Powell County, and Montana StateUniversity (MSU, under contract to EPA) [5]. The soil sampling was performed on September 24-27, 2001. This report includes the sample findings as well as conclusions reached after presentationof the results to the Powell County Health Board, with personnel from EPA, DEQ, DPHHS, the Cityof Deer Lodge, the National Park Service, public environmental groups, ARCO (the potentiallyresponsible party for the contamination), and local and regional newspapers in attendance, on December 5, 2001.
Physical Setting and Conceptual Site Model
The location of Arrow Stone Park is shown in Figure 1. Deer Lodge, a residential community ofabout 3,500 people, is located in Powell County, Montana. The town lies within Reach A of theClark Fork River Operable Unit of the Milltown Reservoir Sediments NPL Site. Arrow Stone Parkstretches along the Clark Fork River from the city limits to the intersection of Interstate 90 andBusiness 90 (Main Street). Until about twenty years ago, portions of Arrow Stone Park were used asa solid waste dumping area. Within the past five years or so, the community has cleaned up the siteand improved several areas with park facilities. ARCO has taken some voluntary actions in areas ofArrow Stone Park to reduce the impact of contaminants, but no formal remediation has occurred.
Arrow Stone Park is split into four subareas, as depicted in Figure 2, and totals approximately 100acres. Subarea 1 was part of a voluntary demonstration project by ARCO receiving some soil tillingand addition of amendments [6]. Picnic tables, benches, walking paths, and a raft launch areamenities of this subarea [7]. As of fall 2001, the paths and other amenities were becomingovergrown with vegetation due to an apparent lack of use and maintenance. Subarea 2 is the onlyarea of the park that appears to receive regular and frequent use by the public. It also was the focusof voluntary actions including soil tilling and addition of amendments, and parts of it received cleancover soil [6]. This subarea has paths, picnic tables, benches, a fishing deck, and handicapped-accessible rest room, all of which appeared to be well-maintained when ATSDR last visited the sitein fall 2001 [7]. Subarea 3 is an undeveloped area of Arrow Stone Park. It contains no amenities[7]. Subarea 4 has a path which is mostly overgrown; there are no other amenities [7]. PowellCounty only owns the lease on the path and the northern 20% of the subarea shown in Figure 2(1). Allthe subareas have varying levels and types of vegetation and ground cover.
The EPA HHRA addendum describes the conceptual site model for Arrow Stone Park [3]. Thepopulation of chief concern for repeated exposures at the park is area residents who may visit thepark and be exposed to site soils on a regular basis for a number of years. Possible activities of parkvisitors include walking, jogging, bike riding, and picnicking. The soil sampling described in thisreport focused on characterizing chronic arsenic exposures to children aged 1 to 10 years old andvisiting the park no more than 48 times per year [3]. Since adults are less at risk than children, conclusions made based on this scenario will be protective of adults as well.
DECISION STATEMENT AND INPUTS NEEDED FOR DECISION
The sampling was designed to determine whether the average arsenic concentrations in current andpotential exposure areas exceed EPA's chronic risk-based concentration (RBC) of 684 parts permillion (ppm) for child recreational users of Arrow Stone Park. The data needed wereconcentrations of arsenic in surface soil in the areas that could currently be used and those where development is possible.
Locations Sampled
Surface soil sampling (0 to 2 inches) took place in each of the four subareas of Arrow Stone Park.To expedite sampling, the samples were grouped into nine sampling units as described below.
- Sampling units 1, 2, and 3 included paths, park bench areas, and other current high use areassuch as rest rooms, fishing deck, and raft launch area in subareas 1, 2, and 4, respectively.Children could occasionally play on the ground in these locations. Each subarea may havedifferent contaminant levels due to differing proximity to the alluvial plain (which received higher contaminant deposition).
- Sampling unit 4 included the areas around the picnic tables in the park. Children could play in the soil around the tables, resulting in more concentrated exposures.
- Sampling unit 5 included bare or poorly vegetated areas in the park. Children playing in thepark have a greater chance of being exposed to soil in bare areas. These sample locationswere selected by all members of the day's sampling team using the criteria of accessibility and presence of contamination-tolerant vegetation.
- Sampling units 6, 7, 8, and 9 included random samples taken in each of the subareas in thepark. The random samples were collected to give a more complete picture of arsenic levelsin the park, since previous sampling had been biased towards bare or obviouslycontaminated "slickens" areas.
Because only surface samples were collected, the sample results are limited temporally. Erosion orflood events could wash away the surface soil or deposit new soil. In this event, the soil surfacewould need to be recharacterized to determine current arsenic concentration.
Sample Collection, Handling and Storage
ATSDR personnel collected surface soil (0 to 2 inches) samples from the locations described above.Personnel from EPA and its contractors MSU and CH2M Hill provided field oversight and assistedin the sampling effort. Sample locations were cited using GPS and photographed. Samples werecollected and composited in a manner consistent with and equivalent to EPA's Standard OperatingProcedure (SOP) SS-2 and SS-6, as detailed in Standard Operating Procedures for Clark ForkRiver Superfund Site Investigations [8]. Decontamination of sampling equipment was consistentwith SOP G-8 [8]. Samples were shipped under chain of custody to an analytical laboratory(DataChem in Salt Lake City, Utah) for the determination of arsenic concentrations. Details of thesampling activities were documented in a field log book and the chain of custody records.
Laboratory Analysis of Samples
DataChem determined the arsenic concentration of each sample using standard digestion (SW 3051)and detection (SW 6010) methods and protocols. The laboratory agreed to retain samples for sixmonths in the event that re-analysis is requested. No party indicated a desire to analyze splitsamples.
QA/QC and Data Validation
Duplicate samples were collected for 5 percent of soil samples. Field standards for arsenic in soil(National Institute of Standards and Technology Standard Reference Material (SRM) 2710 andSRM 2711) were added to the sample stream each day as a quality check. Field and equipmentdecontamination blanks were also collected each day of sampling. The contract laboratory followedEPA-required analytical QA/QC procedures, including laboratory blanks and spikes.
MSU, working under contract from EPA Region 8, Montana Office, validated the laboratory data asspecified in EPA guidance documents for data evaluation [9]. Validation included checking thedata set for all required laboratory deliverables, checking laboratory reported QC results against theraw data, identifying any out-of-control analytical situations, verifying results for the field standards,assigning appropriate data qualifiers, and making a professional assessment of the data set. The datavalidation was reported in a memorandum [9]. None of the data collected were rejected. About halfthe samples were judged to be of enforcement quality. Some of the samples were judged to be ofscreening quality due to laboratory matrix spike recoveries being outside the acceptable controlwindow. The data were judged to be of sufficient accuracy (bias and precision) to support theirintended use [9].
Data Evaluation
Upon receipt of the validated data, ATSDR met with EPA, DEQ, and DPHHS to decide the mosteffective way to present the data. It was agreed to group the data into exposure units correspondingto the subareas of Arrow Stone Park, as described above and shown in Figure 2. Each subarea had adifferent number of samples, since sampling was concentrated in the most heavily used areas of thepark.
Average values (means and/or medians) and statistical confidence intervals for the soil arsenic datafor each subarea were calculated by MSU, under contract from EPA Region 8, Montana Office[10,11]. The calculated 95% upper confidence limit (UCL) is the arsenic value which soil sampleshave a 95% probability of being below. The data set corresponding to subarea 2 was found to benon-normally distributed,. Therefore, the median was used as an estimate of the averageconcentration, and a different kind of statistical analysis than the other subareas was used todetermine the rank equivalent 95% UCL [10]. In EPA's risk assessment process, both averageconcentrations (mean and/or median) and 95% UCLs are typically compared to the RBC, with the 95% UCL being more conservative.(2)
The results of the soil sampling are depicted graphically in Figure 3. This figure shows that the arsenic levels were generally below the chronic RBC. The results are described in more detail by subarea (exposure unit) below.
Subareas 1 and 2
Subareas 1 and 2 are the developed sections of the park; subarea 2 is the most heavily used. The data and summary statistics for subareas 1 and 2 are shown in Tables 1 and 2. For each of these subareas, the 95% UCL (as well as the maximum value) is much lower than the chronic RBC of 684 ppm. Therefore, these subareas are considered safe for the chronic exposures described in the risk assessment addendum [3].
| Sample ID | Type of sample | Arsenic concentration in parts per million |
|---|---|---|
| 5SS-BA12 | Bare area | 82 |
| 3SS-BN1 | Bench | 30 |
| 3SS-BN2 | Bench | 140 |
| 3SS-BN3 | Bench | 99 |
| 6SS-PA1 | Path | 130 |
| 6SS-PA2 | Path | 340 |
| 4SS-PT3 | Picnic Table | 140 |
| 4SS-PT4 | Picnic Table | 260 |
| 4SS-PT5 | Picnic Table | 520 |
| 5SS-R10 | Random | 140 |
| 5SS-R11 | Random | 38 |
| 5SS-R12 | Random | 370 |
| 5SS-R13 | Random | 350 |
| 5SS-R14 | Random | 130 |
| 5SS-R15 | Random | 270 |
| 5SS-R16 | Random | 21 |
| 5SS-R9 | Random | 150 |
| 4SS-S1 | Raft launch | 37 |
| Mean | 180 | |
| 95% Upper Confidence Limit | 250 | |
| Sample ID | Type of sample | Arsenic concentration in parts per million |
|---|---|---|
| 5SS-BA1 | Bare area | 96 |
| 5SS-BA10 | Bare area | 360 |
| 5SS-BA11 | Bare area | 330 |
| 5SS-BA2 | Bare area | 190 |
| 5SS-BA3 | Bare area | 120 |
| 5SS-BA4 | Bare area | 150 |
| 5SS-BA5 | Bare area | 56 |
| 5SS-BA6 | Bare area | 490 |
| 5SS-BA7 | Bare area | 81 |
| 5SS-BA8 | Bare area | 54 |
| 5SS-BA9 | Bare area | 540 |
| 3SS-BN10 | Bench | 10 |
| 3SS-BN11 | Bench | 10.5 |
| 3SS-BN12 | Bench | 23 |
| 3SS-BN4 | Bench | 19 |
| 3SS-BN5 | Bench | 9.5 |
| 3SS-BN6 | Bench | 29 |
| 3SS-BN7 | Bench | 21 |
| 3SS-BN8 | Bench | 9.5 |
| 6SS-PA10 | Path | 9 |
| 6SS-PA3 | Path | 33 |
| 6SS-PA4 | Path | 52 |
| 6SS-PA5 | Path | 20 |
| 6SS-PA6 | Path | 28 |
| 6SS-PA7 | Path | 11 |
| 6SS-PA8 | Path | 10.5 |
| 6SS-PA9 | Path | 19 |
| 4SS-PT1 | Picnic Table | 39 |
| 4SS-PT2 | Picnic Table | 19 |
| 5SS-R1 | Random | 8.5 |
| 5SS-R2 | Random | 10 |
| 5SS-R3 | Random | 9 |
| 5SS-R4 | Random | 9.5 |
| 5SS-R5 | Random | 31 |
| 5SS-R6 | Random | 9.5 |
| 5SS-R7 | Random | 9 |
| 5SS-R8 | Random | 11 |
| 4SS-S2 | fishing deck | 10 |
| 4SS-S3 | restroom | 45 |
| Median | 21 | |
| Rank Equivalent 95% Upper Confidence Limit | 45 | |
Subarea 3
The data and summary statistics for subarea 3 are shown in Table 3. For this subarea, the 95% UCL is lower than the chronic RBC of 684 ppm. Based on current use and these random sampling results, subarea 3 is considered safe for the chronic exposure scenarios described in the risk assessment addendum. It should be noted, however, that any development in this subarea would result in more targeted exposures, and it would be prudent to perform more characterization should development occur.
| Sample ID | Type of sample | Arsenic concentration in parts per million |
|---|---|---|
| 5SS-UA1 | Random | 270 |
| 5SS-UA2 | Random | 40 |
| 5SS-UA3 | Random | 9.5 |
| 5SS-UA4 | Random | 82 |
| 5SS-UA5 | Random | 67 |
| 5SS-UA6 | Random | 71 |
| 5SS-UA7 | Random | 110 |
| 5SS-UA8 | Random | 140 |
| Mean | 99 | |
| 95% Upper Confidence Limit | 166 | |
Subarea 4
The data and summary statistics for subarea 4 are shown in Table 4. For this subarea, the 95% UCL is higher than the RBC of 684 ppm. This result is partially due to the small number of samples collected, which added uncertainty. Only one sample was actually above 684 ppm arsenic, and this sample was collected from an area of very thick, tall grass, making exposure to soil unlikely. Because of this, and the fact that this section of the park is rarely used, the arsenic concentration in subarea 4 is not considered a problem. Powell County, ARCO, and EPA officials stated that this subarea would not be developed until it is cleaned up, and that the cleanup will be addressed in the record of decision (ROD) for Milltown Reservoir Sediments.(3)
| Sample ID | Type of sample | Arsenic concentration in parts per million |
|---|---|---|
| 6SS-PA12 | Path | 530 |
| 6SS-PA13 | Path | 130 |
| 6SS-PA14 | Path | 31 |
| 5SS-R17 | Random | 800 |
| 5SS-R18 | Random | 500 |
| Mean | 398 | |
| 95% Upper Confidence Limit | 789 | |
This sampling effort characterized chronic exposures at Arrow Stone Park. The sampling was not designed to allow decisions to be made regarding risk associated with acute arsenic exposures to children who may exhibit pica behavior (eating soil). This issue was discussed at the Powell County Health Board meeting on December 5, 2001. Because of the large area of the park and the infrequency of pica behavior, it was considered impractical to perform soil testing to assess risk from pica exposure. Instead, the health board has initiated an effort to create and distribute pamphlets to educate parents about the risks associated with arsenic-contaminated dirt and basic hygiene practices that can minimize risk to their children.
- The two developed subareas (1 and 2) in Arrow Stone Park are safe, assuming chronic exposures as described in EPA's HHRA addendum (children 1 to 10 years old who visit the park 48 times a year for many years).
- There is no concern at present for the undeveloped subareas (3 and 4) of the park, since arsenic levels and/or use are low.
- This sampling effort was designed to characterize risk of chronic exposures to arsenic in soil at Arrow Stone Park. The data cannot be used to determine risk from acute arsenic exposures to children who may eat contaminated soil.
- The initiative by the Powell County Health Board to create and distribute educational pamphlets about the risks associated with arsenic-contaminated dirt is a practical way to address acute exposures resulting from pica behavior.
- If subareas 3 and/or 4 are developed, ATSDR recommends further arsenic sampling and/or cleanup.
- ATSDR recommends that Powell County proceed with its effort to educate parents about the risks to children from eating soil.
Authors of Report
Jill J. Dyken, Ph.D., P.E.
Environmental Health Scientist
Superfund Site Assessment Branch
Division of Health Assessment and Consultation
John R. Crellin, Ph.D.
Senior Environmental Epidemiologist
Superfund Site Assessment Branch
Division of Health Assessment and Consultation
Regional Representatives
Dan Strausbaugh
Regional Representative
ATSDR Region 8 Montana Office
Regional Operations
Glenn Tucker, Ph.D.
Senior Regional Representative
ATSDR Region 8
Regional Operations
- Roy F. Weston, Inc. Baseline Human Health Risk Assessment for the Clark Fork RiverOperable Unit of the Milltown Reservoir Sediments National Priority List Site. Prepared for the U.S. Environmental Protection Agency, Region VIII (Montana Office), by Roy F. Weston, Inc., 1998 January.
- Dyken, Jill J., Crellin, John R. Public Health Consultation for Arrow Stone Park, DeerLodge, Powell County, Montana. Atlanta, GA: U.S. Department of Health and HumanServices, Agency for Toxic Substances and Disease Registry, 2001 March 27.
- Syracuse Research Corporation. Baseline Human Health Risk Assessment for the ClarkFork River Operable Unit of the Milltown Reservoir Sediments National Priority List Site.Addendum 1. Risk-based concentrations for exposure of recreational visitors atArrowstone Park to arsenic in soil and tailings. Prepared for the U.S. EnvironmentalProtection Agency, Region VIII, by Syracuse Research Corporation, 2001 October.
- Dyken, Jill J., Crellin, John R. Public Health Consultation Concerning Evaluation ofEPA's Addendum to the Baseline Human Health Risk Assessment: Risk-basedConcentrations for Exposure of Recreational Visitors at Arrowstone Park to Arsenic inSoil and Tailings (draft). Atlanta, GA: U.S. Department of Health and Human Services,Agency for Toxic Substances and Disease Registry, 2001 August 3.
- Dyken, Jill J., Crellin, John R. Environmental Media Investigation Work Plan for ArrowStone Park in Deer Lodge, Montana. Atlanta, GA: U.S. Department of Health and HumanServices, Agency for Toxic Substances and Disease Registry, 2001 September 18.
- ARCO Environmental Remediation LLC. South Deer Lodge Entryway ImprovementProject: Arrow Stone Park Data Summary Report. Anaconda, MT, 2000 March 17.
- Hanson, Ron. Facilities and Use of Patterns of Arrow Stone Park. Deer Lodge, MT, Powell County, 2000 June 13.
- ARCO. Standard Operating Procedures for Clark Fork River Superfund SiteInvestigations. Anaconda, MT: ARCO, 1992 September.
- Neuman, Dennis. Memorandum to Scott Brown (EPA) concerning validation of Arrowstonepark soil arsenic data. Bozeman, MT: Montana State University, 2001 October 21.
- Neuman, Dennis. Memorandum to Jill Dyken (ATSDR) concerning statistical confidenceintervals for Arrowstone park data. Bozeman, MT: Montana State University, 2001November 6.
- Neuman, Dennis. Memorandum to Jill Dyken (ATSDR) concerning statistical confidenceintervals for Arrow Stone park data set 4. Bozeman, MT: Montana State University, 2001November 19.
Figure 1. Location of Arrow Stone Park near Deer Lodge, Montana
Figure 2. Features of Arrow Stone Park, Deer Lodge, Montana
Figure 3. Results of Soil Sampling for Arsenic in Arrow Stone Park
1. According to a conversation with Ron Hanson, Powell County planner, on September 24, 2001.
2. According to conversations with Scott Brown (EPA) on December 4 and 5, 2001.
3. Based on statements by Ron Hanson (Powell County), Barry Duff (ARCO), Scott Brown (EPA), and Bob Fox (EPA) on December 5, 2001.
