Loading...
HomeMy WebLinkAboutSmuggler Superfund - Addendum to Remedial Investigation-Feasibility Studyr ADDENDUM REMEDIAL INVESTIGATION/FEASIBILITY .STUDY SMUGGLER MOUNTAIN, CO Document No.; 149-WP1-RT-CMYS-1 INTRODUCTION AND BACKGROUND This addendum is prepared by the REM II Team as a portion of our technical oversight support for the Smuggler Mountain Site. It will supplement and clarify the RI/FS (Hart 1986) prepared consistent with a consent order. It also will reference the National Contingency Plan where appropriate. The addendum is presented in sections to address (1) additional RI data and analysis, by discipline, and (2) additional FS analysis, by discipline, especially concerning alternatives analysis and applicable or relevant and appropriate requirements. An overall assessment of cost-effectiveness, the comparative cost estimates, and other alternatives analysis criteria is also included. As background, the REM Ii Team was initially assigned the RI/FS for this Sita. Following negotiations with the PRPs, EPA approved PRP plans to conduct the RI/Fs, and a final RI/F'S report was submitted in early 1986. The REM II Team was assigned a Technical Oversight role. This Addendum is another oversight effort of the REM iI Team. REMEDIAL INVESTIGATION {RI) Soil/Tailings EPA has through the endangerment assessment process established a 1,000 ppm Pb action level in soil and tailings to define the site boundaries of Smuggler Mountain. Various schemes have been proposed by tiie EPA, the RE;1 ii Team, and others to define the site, each with its advantages and disadvantages. However, for purposes of the health risk at the Smuggler site, a final endangerment assessment (Clement 1986) has bi�en prepared !' f which selects the statistically based CDM/Geostat Systems map as that which, from a conservative standpoint considering public health effects, is the best approach for site definition. EPA has determined that this map, using all REM II, Hunter Creek, and other data will govern site remediation. This base map is presented as Plate 7 in the RI/FS report. Its use and modification for site remediation will be discussed later in this document. A full tabulation of data present from the X -MET analysis of surface samples, lab analysis of surface and test pit samples, and Hunter Creek. Condominiums data (Engineering -Science 1985) is contained in REM II files. All of these data were used to compile the REM II mapping of the site, using the kriging technique and based on geostatistical techniques as explained below. It should be noted here that the EPA Region VIII Quality Assurance Officer has determined that the X -MET data is acceptable, defensible and technically sound for use in evaluating lead in soil (Hillman 1986). The use of geostati sti cs is well-established in the determination of subsurface mineral deposits, where subsurface borings can help to define areas of varying concentrations to aid in mine planning. For hazardous waste sites, the technique has been used to determine the surface (two-dimensional) and subsurface (three-dimensional) extent of contamination in soil and in ground water. In the Smuggler case, approximately 200 samples (mostly surface) were used to define surface contamination of lead in soil, mine wastes, and mill tailings. The geostatistical technique used was kriging. Kriging is defined as a best unbiased linear estimation method. At the Smuggler site it was enhanced with (1) geostatistical sampling techniques at the threshold (1,000 ppm) line, and (2) use of a broken field sampling pattern in areas where estimates were high. The result is a site definition using statistically generated isolines of lead concentrations. The technique does act to smooth values across areas, so a "hot spot" of several thousand ppm Pb will serve to show a broader area of high lead concentrations than may actually exist. Nonetheless, the mapping does show areas where high -2- lead concentrations pose an "imminent and substantial endangerment to the public health in welfare in the environment" (Section 106, CERCLA) because of the actual or threatened release of a hazardous substance from the site (Clement 1986). This mapping is generally consistent with Plate 7. Discrepancies will be addressed below. The background discussion above leads us to the major topic of this addendum: the definition of the site boundaries as indicated in the RI/FS report. (Its adequacy for remediation, as stated above, will be addressed in the FS section of this addendum). Determination of such adequacy followed several steps. First, the site maps showing detailed mapping of mine tailings, man-made fill, fill, native soil, and newly excavated material were reviewed. Several questionable areas concerning the mapping and the contacts between mapped areas were noted and have been recorded in REM II files. Second, a review was performed of portions of RI/FS Plate 7 which may be in question in terms of contaminated areas. Third, the areas which should be mapped as contaminated based on X -MET data, in addition to those areas already mapped as mine tailings or man-made fill in Plate 7, are (see Figures 1 and 2, rolled and attached as overlays to Plate 7): o Five locations in the Bunter Creek Condominium vicinity, previously mapped as PFILL (Probably Fill; i.e., clean fill dirt) and NS (Native Soil) exhibit Pb values in the 1,000 - 14,000 range. These are downslope from the former Cowenhoven Tunnel portal (now buried) as shown on old maps and photos (Lincoln Devore 1983; Unknown, ca. 1910), would be expected to have scattered pockets of mine wastes, and may contain general areas of surface contamination within 6" of the surface masked by sod. o Two locations in the current Centennial property, which may be isolated cases, show Pb values in t e 3,500 ppm range. The first is just across Spruce Street from the area already mapped as MMF (Man Made Fill, i.e., contaminated fill). This latter MMF area is more immediately downslope of the Cowenhoven, while the Centennial property and the first location above is mostly above or to the side (across slope) from the Cowenhoven portal. The second location is currently paved. -3- Figure 1 (Attached large map w/overlays) -4- r o One location adjacent to the Mollie Gibson Mine drainage near its intersection with Park Ave., shows the highest reading of all samples (21,694 ppm Pb). Field observations indicate that this was a sandy material which appeared to have been dredged from the drainage bottom. The precise location is currently mapped by the PRPs as FILL (i.e., clean fill). o Finally, two areas to the southwest of the area currently mapped as FILL (uncontaminated) an native soil) actually s ow values in the 3,000 - 9,000 ppm range. These areas are on the very edge of the current study area, in the yards of single family homes. It should be noted that one area on the northern edge of the Centennial property has been mapped as MMF (Man Made Fill) in the RI/FS, yet exhibits X -MET values in the 200 - 500 ppm Pb range. This could be remapped as uncontaminated based upon the data. However, the RI/FS report discussion of confidence in the visual mapping scheme and a conservative view regarding public health risk would suggest that this should remain in question as a contaminated area. Figure 1, showing the areas noted above, was compiled from maps of mine workings, surficial geology, and those from RI/FS and REM II mapping efforts. From the RI data presented above, it is clear that the provisions of Section 300.68(d) of the NCP have been carried out regarding an exposure assessment and sampling of the soil/tailings to determine contamination. As one operable unit, a limited emergency action has already taken place on site involving security for a contaminated area north of the tennis courts, including a chain link fence, locked gate, and warning signs. The need for removals (300.68(d)) has been assessed by EPA in various orders pertaining to the site. The next step in the predesign or remedial design stage is to determine how to address remaining contaminated areas on site. To determine such areas, additional investigations may be necessary during conceptual design planning. There is general agreement, from RI data, that approximately the eastern one-half of the site is contaminated, and specific remedial action for one portion has already taken place. However, in the Hunter Creek condominium vicinity, in areas near private residences -5- z on the northwest site boundary, and in mapped and unmapped areas on the southwest and south central site boundaries, the extent of contamination is not precisely known. Additional surface sampling and X -MET analyses are recommended to more precisely determine the 1,000 ppm Pb line and to limit the site in the areas previously highlighted. If contamination extends in spots toward the City of Aspen, remedial measures for these isolated "hot spots" may be needed, especially within 500 meters (about 1/4 mile) of the 1,000 ppm Pb line. The statistical analyses performed in the RI/FS (Spearman Rank Correlation Coefficient) definitively shows that the measured lead concentrations are correlated with the mapped soil type. This correlation holds at both the 99 and 95 percent confidence levels. However, the RI/FS conclusion that the site "could be mapped at the 1,000 ppm isopleth by definition of tailings and manmade fill soil types with an accuracy of 99 percent" is not substantial based on the statistical analyses performed. Essentially, the rank correlation coefficient indicates high levels of correlation between soil type and lead levels (which have been converted to a rank for analyses); however, no conclusion can be made concerning exact concentration isopleths (e.g., the 1,000 ppm line) without further statistical evaluation. Because of the potential for elevated concentrations of uranium and radium at the Smuggler site (Nyberg 1985), selected soil and tailings samples from the RI/FS and REM iI sampling efforts were analyzed for radium -226, gross alpha and uranium. The samples were split from previous samples collected for analyses of lead and were analyzed by the EPA's Region VIII laboratory at the Denver Federal Center. The results of the analyses are provided in Table 1. Also shown on this table are the measured lead concentrations and the field description of the material as recorded by the CDM field geologist. The EPA, under the provisions of the Uranium Mill Tailings Radiation Control Act, has set standards for the maximum allowable radium concentrations in surface soil on open land. These standards are provided in 40 CFR 192.12, Standards for Cleanup of Land and Buildings Contaminated cq Ln U- } f rl N .-1 ►-1 Ql Q N O Pr N +1 N ^I _ d N eh 1 .w G 7 O .-I W Q LCY Lri N N rn a +1 lYJ r 4 Lo O co N r rl 4-3 .�d Ql LL) O ++ 1 r-1 ^I Z +1 rs N 1 lD G 7 O .-I W Q LCY Lri N N rn a CD 2 Lo O +1 w N 4-3 ^4 � n 3 LO Li O Uf r i u, c Ia ++I N O co Ct 1^] w F - 1 c � L+Qa coo CD Ln Ln Li- +I L •? R Ch 01 Lin -4 Ql N ++ 1 r-1 C Z Lr rs n 1 ..a G 7 O .-I W Q b• N N rn a CD 2 Lo O +1 w N 4-3 N C] 3 r LT Ql 1 a G 7 O Ld 2 O N 4-3 � 3 r O Uf r i u, c Ia � r w U 4 r U r + Q CA 4.3 O U) CA UI Q ro . r W LTi 6i U L.3 r 3 y E hi X r r ro = i OJ to—=u r I- U- I.- U- II i1 II !1 1! I LL. w =U - w = with Residual Radioactive Materials from Inactive Uranium Processing Sites. In particular, the concentration of radium -226 in land averaged over any area of 100 square meters shall not exceed the background level by more than: o 5 picocuries of radium -226 per gram of soil averaged over the top 15 cm of soil, and o 15 picocuries of radium -226 per gram of soil averaged over 15 cm layers at depths below 15 cm. As shown on Table 1, the 5 pCi/g level is exceeded in three of the ten samples. However, as stated above, the standard is 5 pCi/g above background. Although background has not been rigorously established, the background value appears to be 1.4 to 1.8 pCi/g (based on samples 94 and 120; see Table 1). Therefore, the high values measured on the site may actually be below or approximately equal to the standard. All of the samples with the elevated radium concentrations contained substantial amounts of tailings and high lead concentrations. From the data in Table 1, it appears that the higher levels of radioactivity are associated with the tailings with lead concentrations greater than 1,000 ppm (location 73 is a slight anomaly). Therefore, remedial actions based on lead concentrations may also alleviate any potential concern from the radium (e.g., the areas may be covered with greater than 15 cm of topsoil). However, this may not be the case in all areas. For example, the three samples with elevated levels are at locations with no remediation currently proposed regarding surface covering of the materials (e.g., the trailer park and "not mapped" area - see Figure 1). These areas are being investigated for additional remedial action. During these evaluations, the concentrations of radium -226 should be considered in relation to current standards, potential health concerns, and normal background levels of radioactivity. Surface Water In order to quantify directly the surface water flows in on-site drainages, flow measurements were made in the Cowenhoven and Mollie Gibson drainages. The flow data were collected on April 25 and 26, 1986, using a Baski portable cutthroat flume. Discharge volumes necessitated the use of an 8 -inch throat for the Mollie Gibson drainage and a 2 -inch throat for the Cowenhoven drainage. For the Cowenhoven, a flow of 14.8 gpm (0.033 cfs) was measured approximately 25 feet downstream from the initial discharge point. At this location, the drainage is loosely lined with plastic, and visual observations indicate that some flow is likely occurring beneath the plastic liner. At the next measuring point, approximately 250 feet down gradient at Lone Pine Road, a flow of 16.9 gpm (0.037 cfs) was recorded, with no flow beneath the liner. Observations indicate that the increase i s due largely to (1) the "surfacing" of flow which was initially beneath the liner, and (2) some seepage from landscaped areas around the Hunter Creek condominiums. Immediately below lone Pine Road, a significant percentage of the flow exits the channel to a broad marshy area, from where it either infiltrates into the Roaring Fork alluvium or evaporates. The flow measurement about 500 feet downstream of the marsh was 8.35 gpm (0.019 cfs), indicating that half of the upstream flow was lost to the marsh. It should be noted that the marsh area exhibits a significant accumulation of tailings and'man-made fill (Figure 1). Additional losses from the Cowenhoven drainage may occur to the east of the Pitkin County Hospital, where it apparently exits from a storm drain system prior to entering Hunter Creek. Data from the Mollie Gibson drainage exhibit significantly higher discharge, with a value of 128.4 gpm (0.285 cfs) recorded at the drainage's emergence south of the Smuggler tennis courts. Flow remains constant for over 500 feet, to below King St. Below this location, however, water is diverted to several small residential ponds along a terrace of the Roaring Fork River. The ponds are apparently used to provide water for sprinkler systems. Additionally, inflow to the Mollie Gibson is derived from a branch of the Salvation Ditch. At the time of measurement, this controlled inflow was approximately 45 gpm (0.1 cfs). At the confluence of the Mollie Gibson with the Roaring Fork River, flow in the latter was measured at 129.4 gpm, indicating a net increase of i gpm from its point of origin. Given the inflow enroute, however, it is apparent that approximately 39 gpm are discharged to the ponds, for subsequent release via seepage, sprinkler (and thence also seepage), or evaporation. No infiltrative losses, however, are exhibited in the drainage reach above King St., where tailings are known to be present. in terms of water chemistry, both drainages exhibit generally low to moderate concentrations of dissolved metals; only manganese (Mn) exhibited high values. Data collected by the EPA Field Investigation Team (FIT) during 1983 showed Mn concentrations ranging from 0.23 mg/L to 0.45 mg/L for the Cowenhoven and Mollie Gibson drainages, respectively. Data collected by the REM 11 team during low flow of March 1985 show dissolved Mn to be 1.4 mg/L for the Cowenhoven drainage, but only 0.1 for the Mollie Gibson. Importantly, the concentrations of cadmium were below detection for both sampling efforts. Lead values were nearly as consistent, with only a trace (0.003 mg/L) detected in the Cowenhoven discharge during March 1985. Zinc values ranged from 0.13 to 0.28 mg/L (Cowenhoven) and 0.38 to 0.73 mg/L (Mollie Gibson) for the two sampling programs. The pH values ranged from 6.4 to 7.8 mg/L (Cowenhoven) and 7.1 to 8.1 mg/L (Mollie Gibson) for the 1983 and 1985 sampling programs, respectively. In the absence of reliable site-specific information for evapotranspirative losses, calculation of an empirical water balance is inappropriate. Given the measured flow rates, existing water quality data, and mapping of areas contaminated by tailings (Figure 1), however, several conclusions can be drawn: 1. There is no evidence that the critical metallic constituents of the tailings, notably cadmium and lead, have contaminated site surface water. 2. The potential for significant infiltration of surface water into the ground water system is limited to a marshy areas adjacent to the Cowenhoven drainage, and a series of small ponds which receive discharge from the Mollie Gibson drainage. 3. The marshy area is dominated by tailings and fill material, confirmed by REM II soil sampling. Conversely, the latter pond area exhibits no visual evidence of tailings or contaminated fill; no sampling has been conducted in this area, however. -10- It is apparent that the potential for contamination of the hydrologic system is dependent on the chemical interaction between water and tailings/fill. From existing data, it is likely that such interaction involving surface water is confined to the marshy area adjacent to the Cowenhoven drainage. The potential geochemical implications, as well as the probable role of ground water, is discussed in the following section. Ground Water Ground water beneath the Smuggler site occurs in both unconsolidated surficial deposits, and within the underlying sedimentary bedrock strata. As stated in the RI/FS report, the bedrock system is decidedly complex, with extensive faulting and fracturing which controls the occurrence and flow of ground water in undisturbed strata. The aquifer(s) are further complicated by underground mine workings, which probably represent the preferred ground water flow paths. In general, however, the bedrock ground water system is of minor importance, due to (1) limited existing ground water use, (2) limited potential for future development, and (3) significantly greater use, and potential for use, associated with the alluvial aquifer of the Roaring Fork River valley. The importance of bedrock ground water, therefore, is restricted to its role, if any, in recharging the alluvial system. Based on existing knowledge of the site, the majority of such recharge is provided by discharges from the Cowenhoven and Mollie Gibson adits. The unsaturated to slightly saturated conditions in evidence from wells and boreholes between and upslope of these adits indicates that any other bedrock contributions are negligible. Considerably more is known regarding the surficiai unconsolidated aquifer. Monitor wells installed by the FIT in February 1985 (4 wells) and by the REM II team in October 1985 (4 wells) clearly show that saturated deposits beneath the site are in direct communication with ground water in the alluvium of the Roaring Fork Valley, in an unconfined aquifer system. Water levels measured from six of the eight wells during November 1985, February 1986, and April 1986 exhibit a piezometric surface which parallels the Roaring Fork River; i.e. flowing in a down -valley, northwesterly direction. The remaining two wells, located upslope from the valley -11- bottom, exhibit unsaturated conditions, indicating that ground water occurrence is limited to beneath the lower slope areas comprising the southwest half of the site. As noted above, some recharge to this aquifer may occur from the Mollie Gibson and Cowenhoven adits; the direct infiltration of incident precipitation and surf i ci al runoff undoubtedly also provides some recharge. Based on the piezometric data, however, it is apparent that the large majority of recharge occurs as underfloor in the alluvial system from up the Roaring Fork Valley. As such, monitor Well 1 is established as an upgradient well, Well 5 as downgradient, and Wells 7, 8, 9 and 10 are situated within the site itself. Ground water chemistry data from these six monitor wells have been collected in November 1985 and February 1986; a third sampling effort was performed during the week of May 12, 1986 (data not yet available). These November and February data are presented in Tables 2 and 3. Of particular importance to the RI/FS is the absence of lead in all samples, and the occurrence of cadmium in Well 7 (both sampling rounds) and Well 5 (November 1985 only; the well was unsaturated in February 1986). A slight increase in cadmium, from 0.007 to 0.010 mg/L, occurred between the two sampling periods. Zinc concentrations are also highest in Well 7 for both sampling periods. Given the absence of cadmium in both the upgradient well (No.1) and the Mollie Gibson and Cowenhoven discharges, the data indicate that the occurrence may be a localized phenomenon, potentially as a result of tailings. Data collected for the RI/FS suggest that the acid-forming potential of the soils is negligible, attributable to an abundance of calcium carbonate in the host rock. These conclusions are valid, but based on a very limited number of samples. There may be significant potential, therefore, for localized pockets of tailings to produce acidic conditions if they are derived from the "core" of the mineralized zone with a low percentage of buffering calcium carbonate. Further monitoring is therefore warranted in all wells. Because of the potential for elevated concentrations of radioactivity, selected ground water samples were analyzed for radium -226, gross alpha and uranium. The samples were split from those collected by REM II personnel and were analyzed by EPA's Region VIII laboratory. The results of the -12- TABLE 2 GROUNDWATER ANALYSES FOR NOVEMBER 1985 � * 11 No. Parameter GW -1 GW -5 GW -7 GW -8 GW -9 GW -10 Arsenic NO ND NO ND ND NO Cadmium NO 0.004 0.007 ND ND ND Calcium 4.59 136 143 20 119 128 Iron NO ND ND ND ND NO Lead ND ND NO ND NO ND Magnesium 14.1 23.8 52.5 5.74 38.5 36.8 Manganese 0.017 ND 0.052 ND NO 0.043 Potassium NO NO 1.92 ND NO ND Sodium 20.5 9.69 6.68 4.16 NO 6.35 Zinc 0.062 0.060 1.00 0.018 0.413 0.053 Notes: Concentrations in mg/L; metals are dissolved. Validation criteria qualifiers pertain to some data; details are included in REM II files. Source: CDM 1986. N 1 O 3 to to O La m r+ qr of to en 01 O LO-4to^O n t1 O 1 antaocoeo.:toon.:dooknr-moo 3 Z Z S'7 Z t'! Z C7 N Ct N [►? r+a C7 .•r N ••4 to (D O 1 to S N O A '••1 O N Q+ r• to O rri z Q N .r ON V 0% A N x .- d � � a c d m r •� W W t+ LL. I C7 N -•� 3 S C7 0 S M LL. LJ Ln �' _.1 LJ O to N r F-• J Q d a Z v to 0 N r w C N N M Ln A• -4 0 •••1 M N A• M t N •••I a QI a L O ro S o C A L.7 1 ro Z b 0 a,�,an,�,a�,a aaaaasa a,aYa,>1 W L L L L L L L i L i L L L L L L L L L a©ccannnonnnoaaaaoo 0 C H N O to to O La m r+ qr of to en O LO-4to^O n t1 O antaocoeo.:toon.:dooknr-moo Z Z S'7 Z t'! Z C7 N Ct N [►? r+a .•r N ••4 to O ro N O A N O O N Q+ r• to O rri Q N .r ON V 0% A N +I o0cC0 0•-1mc;cm,.;0 NLr1 OO Z Z CV Z Z S Z •••I to (%j A• N N .•� C-) .M to C7 -t d O r •� W C7 N -•� to Le] C? N in fn to Ci O S l•f O N O CT d �t to O + 1 �' C] ONOOta000O.+�?ndCT� �d N r O Q a v 0 N r Q CD 4 C N N M Ln A• -4 0 •••1 M N A• M t N •••I O 0 0 C) 1= C' CVA•. -I NNC]LO0 to+100 QI 'C t02 Lo Z .4 CQ O A• L .� N -4 as ul . •L ro o C A ro b a,�,an,�,a�,a aaaaasa a,aYa,>1 L L L L L L L i L i L L L L L L L L L a©ccannnonnnoaaaaoo 0 H 0 ++ N 0 A• C A' Lt) O Nm 04 40 rp LnO LOO Otd O-4 +� +I L 0at000A•00M014LegCH-4A•0 mO w Z Z A• Z w•1 w•1 .-I N •••I Lel Co Leg d N In Q L N w O r w O r� r r +1 r.�rr r r rr+•e r•1�r ro O CrZ C_ 7- Z� Z, C_ ZI� Z ���rr� 7 t0 .•r C71 CI C71 CT O] SZ1 Cr Of a1 D1 CT Of C1 C7I Df b7 L.i C3 Of Cr m O N E@ E 6 E E E E E E E E it E E E c. a E •• o to d i Qy r 1•I ro ro V r 4-A N E 6i E U (M G O V 1r 7 7 to 7 4i7 C N r r r= N r V E E N •�-'o 97 W O 1 46 E i•a r r N C •.• 0 7 VI C N E C r ++ E 7 fly MV O RJ C •r r d iE N a A i ro i••7 aA r a ro L L •r V O w ro r to C O CS Cz O N O 'C! r� rt L7f C C�C��r- V N9 O A r^ L A A i d ro O 0--C)= 3ra ACL CU C��••• Jx VIN 0H_w Loll mM_ wcm T 16 analyses are provided in Table 3. As shown, higher concentrations of uranium and gross alpha exist in Wells 7 and 9. The current MCI. under the Safe Drinking Water Act is 5 pCi/l for the combined total of radium -226 and radium -228. The standard for gross alpha (excluding uranium and radium) is 15 pCi/l. In addition, Colorado has a guidance level of 10 pCi/l (about 0.015 mg/1) for uranium. Comparison of these standards and the values shown in Table 3 indicate that elevated levels of gross alpha and uranium may exist. However, the gross alpha values reported on Table 3 by the laboratory include the radioactivity contributed by all emitters except radon, but including uranium. Therefore, to compare these values with the standard for gross alpha, the contribution of the uranium to the gross alpha radioactivity should be removed. By correcting the reported gross alpha concentration for uranium using the value of 677pCi/mg of uranium, the values do not exceed the standard for gross alpha. However, the uranium values for GW -7, GW -9, and GW -10 do exceed the Colorado guidance level. In addition, substantially higher values for uranium and gross alpha were measured in samples from Wells 7 and 9. This observation is consistent with the wells showing higher TDS and trace metal concentrations. Because radioactivity appears to be only associated with the tailings (see Table 1), the results at Well 7 and 9 may indicate that leaching of tailings is occurring. As previously recommended, these wells should continue to be monitored for radionuclides. Because of the potential of surface waters to recharge ground water, the interaction of the surface water with minerals contained in the tailings and host rock was modeled using the computer program PHREEQE (Parkhurst 1980). PHREEQE is a thermodynamic based program used to model equilibrium water/rock interactions under a variety of pH and oxidation/reduction conditions. In particular, the interaction of waters from the Cowenhoven and Mollie Gibson drainages with minerals assumed to be present at the Smuggler site were modeled at the following conditions: pH: 6.4 to 7.1 s.u. Eh: -200 to +400 my -15- These values represent a range of potential conditions as oxidized surface waters percolate downward, potentially becoming more reducing. The results of the modeling effort indicate that under oxidizing condition (Eh of +400 mv), the carbonate minerals cerrusite and otavite control the solubility of lead and cadmium, respectively. Using surface water analyses representative of the Cowenhoven and Mollie Gibson drainages, cadmium/lead concentrations in the water in equilibrium with the minerals would range from 0.004 to 0.020 mg/l Cd and 0.195 to 0.31.8 mg/l Pb. However, as surface waters percolate downward, they may become reduced. Under reducing conditions ( Eh of -200 mv) , the sulfide minerals galena and greenocki to control the solubility of lead and cadmium, respectively. Under reducing equilibrium conditions, waters of quality similar to the Cowenhoven and Mollie Gibson drainages (except with reducing oxidation/reduction conditions) would contain less than 0.001 mg/1 of both lead and cadmium. Given the small concentrations of cadmium observed in actual well samples, reducing conditions probably dominate the ground water system. However, as shown, under oxidizing conditions, much higher concentrations of lead and cadmium can occur. Because percolating surface waters may oxidize ground water (especially in the Cowenhoven and Mollie Gibson drainages and ponds; see Surface Water discussion), continued monitoring is warranted. Air Quality/Meteorology The use of Grand function, Colorado temperature and precipitation data in the RI/FS to characterize the Smuggler site in Aspen has been questioned by the Colorado Department of Health. Data from the Aspen Water District (precipitation data only, about 25 years of record) and the Aspen Airport (daytime temperature and precipitation data only, about 10 years of record) are more likely to accurately characterize Smuggler site conditions. Further, the sprinkler systems at Centennial (which effectively increase soil moisture and aid dust control) were stili to be installed in May 1986; Hunter Creek had installed sprinkler systems and applied topsoil and sod over many locations, but was also undergoing a major renovation at some buildings, with attendant landscaping changes, in June 1986. -16- FEASIBILITY STUDY (FS) Soil/Tailings As noted in the RI discussion, EPA has determined that Plate 7 from the RI/FS report, as modified by additional areas of contamination from X -MET readings (Figure 1 and 2 overlays) will constitute the site definition map for Smuggler Mountain. This map addresses a manageable area for remediation purposes and identifies the areas of highest contamination in the vicinity, to our knowledge. Selected areas of contaminated mine wastes are reported to exist throughout Aspen and vicinity - primarily under roads, in gardens, and along building foundations. Remedial action for such spots in the vicinity of the site will be addressed during the final design phase. The use of a RCRA 40 CFR 264 multi -layer cap is introduced in the RI/FS (p. 8-14) and included in the RI/FS cost tables, but was dismissed because of the technical constraints (steep slopes) and the lack of need. It is therefore not made a part of the RI/FS recommended remedial action. For this reason, it should be deleted from the costs discussions in the FS (see Cost Estimates for Remedial Alternatives later in this document). However, the NCP requires consideration of all "relevant and appropriate" requirements and so one must continue to consider the multi -layer cap option, or a modification of i,, for remaining exposed contaminated areas in the Smuggler Mountain site vicinity. Actual details of cap construction will be addressed later. Regarding the multi -layer cap and other regulatory and policy issues, one very important aspect of the remediation should be addressed. As required by the NCP and highlighted in several sections throughout that document, applicable or revelant and appropriate (ARA) federal requirements must be considered pursuant to Sections 104 and 106 of CERCLA for response or removal actions (Porter 1985; Section 300.68(e), NCP). Following a review of the alternatives examined by the PRPs, EPA has determined (Staible 1986) that the ARA's below pertain to the site. -17- Applicable Statute or Policy Safe Drinking Water Act Clean Water Act Specific Provisions Maximum concentration levels (MCLs) in PWS; Recommended Maximum Concentration Levels (RMCLs). Non -point source control; water quality standards. Ground Water Protection Strategy Protection and enhancement of Class II aquifers. Relevant and Appropriate Statutes or Policy Specific Provisions RCRA Capping performance requirements, ground water monitoring and corrective action, site access restrictions, run-on - run-off controls, in-place closure of a landfill (waste piles). Off-site Policy All In addition, the following requirements may be relevant and appropriate: State Statutes or Regulations on Water Supply Systems Additions. Colorado Mined Land Reclamation Act and Regulations. RCRA Guide to the Disposal of Chemically Stabilized and Solidified Wastes. RCRA Guidance Manual on Closure and Post Closure Interim Status Standards. These ARA requirements have several implications for site remediation. As these are stated below, the reader should keep in mind that the "facility" for purposes of the remediation is the site as defined by the 1,000 ppm Pb line (Figure 2). A repository for highly contaminated wastes is planned for the Pitkin County park area, to which waste would be transported and consolidated in place on the "facility" site. Further, this remedy will emphasize, as do RCRA and CERCLA, a permanent, stable low -maintenance remedy using best professional judgment, and designed to mitigate existing sources and pathways of contamination, thereby minimizing or eliminating risks to human health and the environment. -13- In a preliminary sense, then, the implications of these ARA requirements are as follows: 1. Monitoring of ground water contamination at the site has taken into account MCLs and RMCLs for contamination of concern (Safe Drinking Water Act). Detection limits are suitable for comparison with these MCLS and RMCLs, as compared and discussed elsewhere in this addendum. 2. Non -point source controls (Clean Water Act) were considered early on in review of site characterization (FIT) data, and in analysis of the REM II low -flow sampling of surface water and sediment. Since surface water contamination does not appear to be a problem at the site, non -point source controls were not emphasized in the remedy. However, surface run-on controls (RCRA Regs., 40 CFR 264.25(c)) have been determined to be relevant and appropriate as a site remedy. Since surface water run-on and leaching of ground water appear to be related at the site, surface water monitoring may be recommended if further ground water monitoring demonstrates increasing contamination. 3. Concerning other provisions of the Clean Water Act, water quality standards for the Roaring Fork River and Hunter Creek have not been exceeded in the site vicinity to our knowledge; therefore technically no action is warranted. 4. Further, to our knowledge neither federal nor state standards regarding aquifer protection have been finally promulgated. Direct remediation of the aquifer appears to be technically unnecessary at this time since ground water contamination does not represent an unreasonable health risk. There are presently no county restrictions on further aquifer use outside the city limits in the site vicinity, aside from the usual quantity and quality requirements (Dunlop 1986). 5. If ground water contamination becomes a problem at a later time, this matter will be addressed under the EPA Ground Water Protection Strategy. However, because of the potential threat to the drinking water supply as shown by recent data, one aspect of the proposed remedy is to place all private well users in the site vicinity on city water. 6. If ground water contamination is detected in further monitoring of wells 7, 9, and 10 (on the facility boundary), consideration will be given to implementing RCRA-recommended ground water monitoring (40 CFR 264.97) for the repository area. 7. Similarly, some RCRA capping requirements for landfills should be considered. An impervious, multi -layer cap well above the water table (as a stable, permanent, low -maintenance remedy) is deemed sufficient. Liners and leachate collection systems are not relevant. -19- 8. Periodic inspection and maintenance is appropriate (40 CFR 264.303), as is post -closure care (40 CFR 264.117). 9. RCRA guidance on disposal of chemically stablized and solidified wastes would be appropriate and relevant only if the plume capture alternative becomes necessary and the attendant water treatment facility on-site produces hazardous sludge to be moved off-site. A final issue concerning the remedial alternative recommended in the RI/FS deserves further discussion. On p. 4-3 of the RI/FS, it notes that grading is planned for the Smuggler Mountain site, especially in the Centennial area, in order to make it suitable for residential use. This grading has the dual purpose of stabilizing the area and minimizing or eliminating the direct contact, air, ground water and surface water pathways of contaminant migration, provided that an appropriate cover is established and maintained. However, the steep slopes mentioned previously under the multi -layer cap discussion in the RI/FS (p. 8-4) must be addressed as well. EPA recommends that stone armor or other measures be considered on top of or under the soil/vegetative cover on steep slopes in contaminated areas in order that erosion be minimized and stability and containment be emphasized. Maintenance of such areas should be included in 0 & M costs. A field trip to assess present site conditions in April 1986 indicates the following regarding completed remediation of contaminated soil/tailings as a result of construction activities: 1. Most construction of the Centennial Condominiums is complete. Some exposed soil and rock is present on the Centennial site; most is covered with buildings, parking lots, sidewalks and topsoil with a grass cover. (Thickness of the topsoil layer in unknown at this time.) 2. The one playground area on the Centennial property has 6-8 inches of sand over soil. 3. Some exposed areas of bare native soil, fill and (perhaps) man-made fill and tailings remain (a) in the Hunter Creek Condominiums vicinity, (b) above the tennis courts, (c) near the trailer park in the proposed County Park area, and (d) scattered throughout private residences in the vicinity. Most are small in extent (from a few hundred square feet to less than i acre). -20- 4. The proposed county Park area and Smuggler Mine No. 2 tailings pile and vicinity show visually the largest remaining area of tailings and apparent contaminated material, totaling several acres in extent. Detailed comments on the field visit are included in the trip report (CDM 1986). Movement of contaminated materials to a repository on-site by the PRPs is yet under negotiation. REM II comments regarding PRP estimated costs for excavation and removal are addressed in a later section of this document. Remedial design for capping and removal efforts will be addressed in a forthcoming Work Plan and Remedial Design Oversight Report prepared by the REM II Team. Surface Water Results from the RI indicate that the existing surface water system, including the Cowenhoven and Mollie Gibson drainages, Hunter Creek, and the Roaring Fork River, has not been contaminated by on-site tailings. The latter two streams exhibit consistently low major ionic and trace metal concentrations. The mine drainages are typified by moderate to high concentrations of selected metals, including zinc, manganese, and iron, but these constituents are not attributable to dissolution of the tailings nor are they considered a significant threat to public health. For these reasons, site remediation must focus on ensuring that surface water conditions are consistent with other protective measures; remediation of the surface water drainages themselves is not warranted. This approach is consistent with the findings of the RI/FS. Critical components of the surface water system which may affect site remediation are as follows: I. Infiltration through tailings and contaminated fill, potentially leaching metals into the ground water system. 2. Erosion of tailings and contaminated fill, either by existing mine drainage channels or overland flow. 3. Instability of the cap and/or other surface remedies due to erosion by surface water. -21- When properly implemented, the remedial measures recommended in the RI/FS are appropriate. The surface sealing (or capping) discussed previously will retard or preclude the infiltration of surface runoff. Areas of known existing percolation will require particular scrutiny. Subject to pre - design studies, restructuring of the Cowenhoven drainage may be necessary to preclude infiltrative losses to the marsh below Lone Pine Road. If this area is floodprone or susceptiie to catchment of overland runoff, consideration must be given to removal or redistribution of the known tailings at this location. At a minimum, surface capping must extend to all areas of known contamination, such as the marsh. (No defined "wetlands" are present at the site.) Additionally, the diversion and use of surface water from the lower reaches of the Mollie Gibson drainage may require additional evaluation. Lining of the Mollie Gibson drainage above this location, as proposed in the RI/FS, does not appear warranted based on the lack of percolation from the surface portion of the ditch. Secondly, the grading and surface water diversions must be engineered so as to preclude or minimize channel instability. Although not clearly stated in the RI/FS, it is assumed that site runoff may be routed towards existing drainages. Because unconcentrated runoff will be controlled, and reduction in infiltration will increase such runoff, the receiving drainages will likely be subjected to greater flow volumes than at present. Quantification of such increase at the design stage is essential to ensure the stability of both diversion ditches and existing channels. Last, site-specific measurement of the cap's in-place characteristics must be accomplished to ensure its resistance to erosion and its adequacy in terms of minimizing infiltration. Numerous in-place compaction/moisture and percolation tests may be required to characterize fully the expected peformance. Theoretical or typical percolation rates discussed in the RI/FS are not adequate to represent cap conditions unless supported by extensive site-specific data. Ground Water As described in the RI section of this addendum, the existing data are not conclusive regarding ground water contamination. initial sampling showed a -22- 1 private well with elevated cadmium levels; during one sampling period the cadmium concentration of 13 ug/L, exceeded the EPA drinking water standard of 10 ug/L, whereas during the second sample round, the cadmium level dropped to 7 ug/L. Similarly, REM II sampling of on-site Well 7 exhibited cadmium levels of 7 ug/L in November 1985 and 10 ug/L in February 1986. It is also important to emphasize that EPA's recent proposed (11/85) Recommended Maximum Contaminant Level (RMCL) for cadmium is 5 ug/L. Because of the (1) lack of a clearly documented ground water contamination problem, (2) insufficient data to attribute the known metal concentrations �r to tailings, and (3) the inapplicability of ground water treatment or I� containment methods to a site where greater than 100 feet of saturated 9�p unconsolidated material is probable, remediation mechanisms which focus on surficial capping to inhibit leachate generation appear to be most appropriate. The assessment that existing site conditions, including soil chemistry and permeability, are presently conducive to reducing the risk of ground water contamination is fundamentally correct. however, limited permeability (1 repacked sample) and acid forming potential (7 test pits) data are not adequate to verify that leachate production is not, or could not be, occurring at certain locations across the site. For this reason, site capping, and maintenance thereof, and continued ground water monitoring on a quarterly basis for five years are necessary components of site remedi ati on. As discussed previously, the emplacement of a low permeability surficial cap will reduce infiltration, thereby inhibiting the potential mobilization and migration. In some areas, however, ground water flow in bedrock strata or up -gradient alluvial deposits may also mobilize metals within the tailings. For example, discharge from the collapsed adits at the Mollie Gibson Mine or Cowenhoven tunnel may include an underflow component in addition to surface drainage. Continued monitoring is essential, therefore, to ensure that migration of metallic contaminants does not occur despite successful surficial remediation. At present, use of the existing six monitor wells where saturation occurs, potentially augmen*ed by -23- W selected domestic wells, will provide an adequate basis for the monitoring program. As the monitoring program proceeds and additional ground water chemistry data are collected, the adequacy of the monitoring will be evaluated and if necessary corrected. Costs Estimates For Remedial Alternatives After a detailed review of the cost estimates compiled for the remedial action alternatives for Smuggler Mountain Site, the following comments are to clarify and substantiate those estimates as presented in RI/FS. Alternative No. 1: Increased Monitoring o Based on previous sampling done by REAM II personnel at the four groundwater wells existing on the site, labor costs should be increased to $3,840 based on (6 MH/sample) x (16 samples) $40/MH = $3,840. This amount also reflects travel to and from the site. o The costs do not reflect the semiannual site inspection included in the alternative description. This would amount to $960 which also would include travel. The total estimated Annual 0 & M cost for Alternative No. 1 would then increase to $13,240, say $13,300. Alternative No. 2: Source Isolation Capital Costs: o It is unclear in the report as to which soil areas are considered to have an excessively high concentration of lead contamination, thereby requiring an impermeable cap. Also, if the alternative of using an impermeable cap is not to be considered for use because of technical difficulties and because it is not deemed necessary, it is unclear why it was retained in the cost estimate. We therefore, have deleted the costs for the impermeable cap from the Alternative No. 2 costs. o The unit costs for excavation and embankment were estimated at $3.90/CY. We feel that a value of $5.90/CY is more appropriate, which includes additional costs relating to the necessity of the workers being in Level C protection. This increases the cost for excavation and embankment to $5.90/CY x 70,000/CY = $413,000. -24- W o Unit costs for revegetation were estimated at $.30/SY; however, we feel $.58/SY is more appropriate. This will result in a total cost of $.58/SY x 125,OOOSY = $72,500. 2—S, t' �,`rs The total Capital Costs of Alternative No. 2 becomes, $1,197,800. Annual 0 & M Costs: o As noted above and in the review of Alternative No. 1, the costs for revegetation and sampling labor should be increased. Rip -rap estimate also was considered low and was increased to $35/CY. This results in an increase in costs to $17,532.50, say $17,600. Alternative No. 3: Replacement of Water Supplies o Although a 2" diameter water supply pipe may not be of sufficient size to provide water service to six residences (an increase in size to 3" is deemed necessary), the unit cost of $30/LF is sufficient to account for the increase in pipe size and for any protection workers might require if trenching occurs through contaminated soil. There is, therefore, no change in the cost estimate for this alternative. Alternative No 4: Plume Capture o During the screening of remedial measures (Section 7.9) ground water barriers were eliminated from further consideration because the bedrock below the site is fractured and does not provide an impermeable barrier. Subsurface collection drains, however, are required to "overlay a uniform layer of impermeable soils or bedrock," as stated in the analyses of remedial action alternatives (Section 8.2.4.). This technical conflict must be resolved before this alternative can implemented. o There are no changes in the cost estimate for this alternative. Alternative No. 5: Source Removal o As in Alternative No. 2, the costs for excavation of contaminated material should be increased from $3.90 to $5.80/CY. This results in a cost of $5.80/CY x 410,000/CY = $2,378,000. o The original estimate for transportation costs assumed 30 CY per truck and a 500 -mile distance resulting in a $40/trip unit cost. Highway weight limitations restrict trucks to approximately 17 CY -25- V per truck based on a 120 pcf soil weight. The report indicated that a possible RCRA approved landfill disposal site would be the Chemical Waste Management site in Aurora, Colorado. As of this report, this site is not open and does not expect to be opened in the near future. The next closest site is the U.S. Pollution Control Site in Knolls, UT. Using a trucking cost of $1.40/mile/truck the total cost per trip becomes $1.40/mile x 200 miles x $280/trip. Assuming 410,000 CY of contaminated material, 410,000/ 17CY/truck = 22,777 trips are required, making the total cost 22,777 trips x $280/trip = $6,377,560. o Revegetation costs should be increased to $.58/SY resulting in a cost of $.58/SY x 125,000 SY = $72,500. o Site grading was originally estimated at $1,000/AC or $.20/SY. This should be increased to $1.20/SY making the total cost for grading $1.20/SY x 145,200 SY = $174,240. o The mobilization costs and engineering services fee were based on 10% and 15% respectively of the raw capital costs. These normally are fairly accurate estimates; however, in this case they appear excessive. Mobilization and engineering fees should be reduced to the appropriate percentage of the raw capital costs without transportation and disposal costs included. This results in a mobilization cost of $408,000 and engineering fee of $613,000. Total costs for Alternative No. 5 are therefore estimated at $62,740,000. Table 4 summarizes cost estimate comparisons between those generated by the PRPs and those calculated by the REM II Team. In conclusion, our analysis and comparison of the cost estimates for each alternative varies from +28Z for Alternative No. 1 to -6.5% for Alternative No. 5. Alternative No. 2 varies +15% for both Capital and Annual 0 & M costs. There was no variance for Alternative Nos. 3 and 4. Each estimate contained in the RI/ES report generally conformed to the guidelines of Section 300.68 (g) (1), NCP. -25- M d V C C O O w N 4J 4- + + tLi N I b Q u a! v N N y L u a e +3 V 91. V i pC 6 O O O w co @ •a` I • + O 4D o r M w Y in •• •� LO w a i rn ICY M1 N O + O # + NI 1 IY C W N C! L 06 Y N 1 1 Ln O L H K1 %a I I O G O t0 ^y e N W 1 1 N C y ID III Y 1 1 V �y L E I 1 V O. 44 O O W W O N --• O o cm C3 n W O �+ 6 Q 1+f n S S N w r••I .+ U U L A Z 4�] ,.O'C 2-4 RY I I C7 M N 1 1.60 v N44 Q O 1 I O N t 1 co i I O O F+.... .�-� I l A w0 G ti 1 1 N O A U _ A C C A b C A to 41 d 4 f1 b _ L M 01 CA N O PN O •+ Nm M +i 41'1 V 01 4m n W N 10 •4 N L1 Q! A Ln O C N W fn 00 O N O l/'f O Q•7 N In W do r a +' N L N O d O O O 3 Lell CD ' CD O Y C H O� UO O1 O n W qr a r L d .-i 40 i2 A lD (, N A V V Y OI L 7 b +-+ E b C 3 u Y O ai y C ++ O tl7 A L co CO C O A L i A • T O � 41 r O C Y r VVI O! U O r U + L mCLF V V 7y V V P ql r1• N L S. C6 1-8 C O V O Ym u V rN N VY a N O M L Y . F- I a r O O 4 Ln A V7 REFERENCES Clement Associates, 1986. Endangerment assessment for the Smuggler Mountain site, Pitkin County, Colorado, draft final report. Arlington, VA. 5 May. Dunlop, Tom, 1986. (Director, Aspen/Pitkin Environmental Health Department). Personal communication with Scott Mernitz, CDM. 24 June. Engineering Science, 1985. Hunter Creek soils investigations and corrective measure recommendations. Denver, CO. 14 June. Hart, F.C. Assoc., 1986. Remedial investigation/feasibility study, Smuggler Mountain site. New York, NY. March. Hillman, Juanita, 1986. (EPA Region VIII Quality Assurance Officer). Memorandum re: XRF methodology. Denver, CO. Nyberg, Phil. 1985. Memorandum re: radiological considerations for the Smuggler Mountain Site. 16 May. Parkhurst, David L. et al. 1980. PH REEQE - a computer program for geochemical calculations. U.S. Geological Survey Water -Resource Investigations 80-96. Reston, VA. November. Porter, J. Winston. 1985. Memorandum re: CERCLA compliance with other environmental statutes. 2 October. Staible, Tom, 1986. (EPA Regional Project Manager, Smuggler Mountain site, CO.) Memorandum re: applicable or relevant and appropriate Federal requirements, Smuggler Mountain site. Denver, CO. 12 June.