ML20125D972

Forwards Draft Comments Re Environ Assessment,Remedial Action Plan & Processing Site Characterization for Gunnison, Co Site

ML20125D972
Person / Time
Issue date:	02/28/1985
From:	Hawkins E NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION IV)
To:	Themelis J ENERGY, DEPT. OF
References
REF-WM-61 NUDOCS 8506120542
Download: ML20125D972 (14)

Text

n.

• DISTRIBUTION 4 .URF0 s/f INO 8.2 URF0 r/f

~*

DBangart, RIV INO 8.2/SLW/85/02/26/0. SWastler

-1 WM DOCKET CONTROL MHaisfield d & ,17 CENTE" DMartin LHigginbotham FEB 2 c 73g5 *BS t:: n",,

URF0:SLW INO 8.2 040WM177101E 040WM177102E John Themelis, Project Manager Uranium Mill Tailings Project Office U.S. Department of Energy Post Office Box 5400-Albuquerque, New Mexico 87115

Dear Mr. Themelis:

Enclosed are our draft comments on the drafts of the Environmental

' Assessment, Remedial Action Plan, and Processing Site Characterization

' Report for remedial actions at the Gunnison UMTRAP site in Gunnison, Colorado. The coments are being forwarded to you in draf t form in order to meet your request for. receipt of the Gunnison comments by March 1,1985. The final comments will be sent to you directly from NP,C Headquarters by Mr. Mark Haisfield as soon as possible.

Should you require clarification or wish to further discuss these draft comments, we are prepared to discuss them in more detail. If we can be

. of further assistance, please contact Sandra L. Wastler on FTS 776-2811 or Mark Haisfield on FTS 427-4722.

Sincerely,Ac'k

.S/ A) .

aJ Edwa d F. Hawkins,. Chief Licensing Branch 1 Uranium Recovery Field Office Region IV WM Project- -

Cases Closed: 040WM177101E WM Record File 040WM177102E Docket No.

/

PDR.

LPDR B506120542 850228 PDR WASTE ~~

Dis lribylipn:

WM-61 PDR z .. .. ........._ . ... 3 d L .,@ G'.Oe's -

0FC :

p

)  : g ]- ::...-------- :....[

5 nNift .

2 55  :

---

: g -

'O g-----:-----------

ahyg[Oy,I ,Ma Q s,,,I,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,Y,[,,,,,f,,,,,,,,,,,

DATE :85/02/26

s: 0 /K : .....-...... -.-. _..... . ............

NRC Coments - Gunnison Site Surface Water Hydrology and Erosion DRAP

1. Appendi x E ,_P,MF, D,ete_rmi,n,a,t,ien Based on a review of the DOE flood analyses (provided by cover letter from M. Bone to T. Johnson, deted August 17,1984),we conclude that the Probable Maximum Flood (PMF) for the Gunnison River may not have been adequately or conservatively estimated. Our review of the analyses indicates that several important input parameters to the HEC-1 model have not been properly estimated.

Because the PMF on the Gunnison River determines the riprap size that will be provided to protect the toe of the tailings pile, it is important that an adequate estimate of the flow velocity be determined. Further, because the design must remain functional for a 1000-year period without routine maintenance, it becomes very important to conservatively estimate the flow velocities and shear forces associated with the PMF.

The DOE PMF estimate of 205,000 cfs for the Gunnison River was compared to regional maximum flood (RMF) estimates as presented in Crippen and Bue (1977). Based on this comparison, the RMF value of about 210,000 cfs exceeds the DOE PMF estimate. The NRC staff has concluded that RMF data, in most cases, do not represent corservative PMF values and that any PMF which coes not exceed these RMF values has probably been underestinated.

The DOE estinate for the Gunnisen River was also compared to the DOE PMF estinate for the Animas River (Durango EIA). The PMF for the Animas River, with a drainage area of 770 square miles was corputed to be 271,800 cfs. It is doubtful that the Animas River basin is drastically different fror the Gunnison River basir., in terms of flood reducing characteristics.

Further e>ar.ination of the HEC-1 analyses indicate that several paraneters in the PMF ar,alyses have not beer, preperly estimated.

IT A eel reduction of PP'F. The PMF for the Gunrison River cra"r. age casin TIFor Fi') was reduced for the cortined erair. age 1

I

2 area (2000 mi ) of the Gunnison Rivei and Tomichi Creek. This reduction should nc! be applied when dealing with only the peak PMF flow in the Gunnison River.

2) Rainfall Infiltration Rates. The SCS curve numbers that were seTe~c~t'eif Tn"tWelEC-1 model are not considered to be conservative, since little initial saturation of soils is assumed to be present, using these values. For these values to be acceptable, the 24-hour PMF should be preceded by a significant amount of rainfall (e.g., the rearrangement of the last two 24-hour segments of the 72-hour PMP to the first two 24-hour segments), to minimize later infiltration.

Alternately, more conservative estimates of the curve numbers (e.g., AMC-III) or saturation of the soils should be used in the analyses.

3) Tines of Concentration. The formula used to calculate times of concentrat%Eltc's), and resulting lag times, may not be appropriately reasonable or conservative when dealing with extremely steep mountain streams with very high flow velocities. Based on our exanination of topographic maps and portions of the Gunnison River basin, the staff recormends use of the stream hydraulics method (Ref. Design of Small Dams) to estimate tc's and resulting lag times, for streams this steep.

4) Stom Centering. It is noted that the HEC-1 analyses were performed using several sub-basins of the Gunnison River, but usirg a uniform FMF rainfall amount over the entire basin. A mere appropriate rethod of analysis is te use the procedures outlined in "Hydroreteorological Report No. 49" where varying rairfall depths are centered according to an actual isohyetal pattern over the sub-basins to produce the peak PMF in the entire basin.

Taking the above factors into consideration, a revised PMF analysis should be performed to conservatively determine the upper limit of flooding Octential for this site. An acceptable nethod for addressirg the above input paracters and computing the PMF nay be found ir ANSI N-170, " Standard for Determining Design Basis Flooding of Power Reactor Sites".

2

2. Appendix E__ ,_,R_iprap,,Tp,e_,Prptection A. Calculational Basis for Rock Size Provide the basis for the determination of D g riprap size of 6 inchesfortherocktobeplacedatthetoe6kthepileto prevent erosion during a PMF. Based on our review of the HEC-2 analyses, it is difficult to determine what methods were used to compute the required D rock size. Our independent EO analyses for a flow velocity of 15 fps and depth of 7 feet indicatetha$aDon an assumed I $failure size ofslope, approximately using Corps15"ofwill required for Engineers forvulae (Ref. " Hydraulic Design of Flood Control Channels").

However, DOE calculations indicate a required DRO of only 6" (page E-33). The basis for the 6" rock size should be further discussed and/or evaluated.

B. E_ffect of Revised PMF Analyses Using)the above , therevised riprap forPMF estimates the toe (asshould of the pile applicable be - See Question 1 re-evaluated, based on the velocities computed using the revised PMF. .

C. Effects of Pptential Chann,el Meandering and Mo,vem_ent The geor. orphic evaluation provided in Appendix B indicates that a potential for channel mover.ent exists at the site. Because o' the recognized uncertainty regarding channel stability, and the potential for migratior. (of unpredictat.le extent) of the Gunnison River channel toward the pile, we conclude that an additional margin of safety should be applied to the design of the rock toe. This would provide additional protection in the esert of a flood which channelizes flow and produces high velocities at the pile, similar to those velocities produced in the existing Gunnison River channel under FMF conditions. We recomend that redesign of the rock be perforred, using an additional margin of safety to account for this phenomena.

Tak ing irto corsideratior, the above concerns, the rock toe protectier should be re-evaluated.

3

3. Appendix E - Erosion Prot _ec,tjon_,f,or,,

o Top,and Si_de_of, Pjl,e The NRC staff has concluded that the design for long-term erosion protection for the top slopes of the reclaimed pile may not be adequate to meet EPA long-term stability criteria. We believe that the currently-proposed erosion protection could be damaged by rainfall much less intense than the PMP.

The erosion protection for the pile has been designed to resist only the effects of unconcentrated sheet flow, which may not be an accurate representation of conditions which are likely to exist in the future. We conclude that areas of flow concentration will form due to differential settlement, random flow spreading, and/or uneven grading. We conclude that erosive volocities may be produced by precipitation much less severe than the PMP, if such phenomena occur.

Based on independent analyses performed by the NRC staff, we conclude that rock with a D E0 f about 1-1/2 inches should be placed on the top at a thickness of two feet. We conclude that the proposed rock with a Dr0 of about 3-1/2 inches is acceptable for the 3

sides of at a thickness of one foot).

Accordingly, either revise the design of the rock for the top of the pile, or provide additional analyses to document that the design adequately accounts for flow concentration due to different settlerent and uneven grading.

4 Appenc h E , pan Failures identi'y any upstream dars whose failure could affect the site. If it is determined that there are upstream dans that are not desigr.ed for an occurrence of a PMF, provide analyses to determine the effects cf failures of such dams on the various site design features. Identify any future efforts that will be undertaken by the da owners to upgrade the dars or spillways to pass a PMF, Recog-iring that dan f ailure calculations are difficult to perfen-and are sensitive to certain assur.ptions such as time-of-failure, f ailure r. ode, etc. , you ray provide the results of conservative It'ut less rigerous) calculations used to estinate peak water levels arf velecities at the site. Corps of Engineers and U.S. Weather Bureau a

unsteady flow models may be used for such predictions, if a more rigorous approach is taken.

Hydrogeology_(Ground-Water Characterization)

DEA

5. The Gunnison UMTRAP site has contaminated ground water hydraulically downgradient from the site in the alluvial aquifer of the Gunnison Valley. Our review concentrates on the seven criteria provided by EPA in Subpart C of 40 CFR Part 192. These seven criteria are used by NRC staff to review the adequacy of hydrogeologic assessments.

Of the seven criteria, the hydrogeologic assessment in the Gunnison DEA and DRAP satisfies criterion 1, establishment of background ground-water quality, and criterion 2, identification of ground-water contamination. The other five criteria, however, have not been adequately satisfied.

Additional field characterization presently underway should satisfy criterion 3, characterization of the extent of ground-water contamination, including relative concentrations of contaminants within contaminant plumes, and criterion 4, characterization of the rate and direction of contaminated ground-water migration. Once these assessments are completed, DOE should be able to satisfy criteria 5, 6, and 7 and select appropriate actions for the Gunnison UMTRAP site. The NRC would recommend that new data and analysis derived from on-going investigations be submitted prior to issuance of a fir.al EA to allow the NRC to review the data to assure EPA's seven criteria have been satified and to avoid schedule delays The follewing paragraphs reiterate the NRC's concerns as originally expressed in our comments on the PDEA. Ground-water quality infomation in the DEA, DRAP, and SCR indicates that shallow ground water hydraulically downgradient from the site is contaninated, and that water fror this same aquitor is being consumed by residents of the Gurrison Valley. As reported in these documents, concentrations of urariun, arsenic, and seleniun in grcund water consured by the residerts are below maxirur allowatle concentrations for drinking water (e.c., MCL's under 40 CFR Part 141). Additional characterization of the exter.t of grc und-water cor.tanination is warrerted, however, because these cor centratiers nay increase as the contar'r.ated ground water ber.eath the site rigrates doegradiert teward the domestic users of the shallow acuifer. We reco-renced 5

l.

that DOE characterize the variation of ground-water quality with depth, the lateral extent of ground-water contamination, and the fate of contaminated ground water near the confluence of the Gunnisor. River and Tomichi Creek (i.e., does the contaminated ground water conpletely discharge into the surface water or does it migrate beyond the confluence).

Under criterion 4, the DEA, DRAP, and SCR should characterize the stratigraphy of sediments that conduct the contaminated ground water because stratigraphic heterogeneities may significantly affect rates and directions of contaminant migration. In addition, the documents should characterize the distribution of hydraulic head in the alluvial aquifer in the valley by measuring water levels in shallow monitoring and domestic wells in this aquifer. This information is necessary to (1) predict the rates and directions of contaminant migration, and (2) characterize recharge-discharge relationships of the alluvial aquifer with the Gunnison River, Tomichi Creek, and other surface water bodies in the valley (e.g., gravel pits, irrigation canals, ponds, etc.). This information will also aid in characterizing the potential for migration of contaminated ground water toward a municipal well in the alluvial aquifer, which is located several thousand feet north of the site.

Once this information is developed, we recommend that you pursue the strategy for selecting appropriate remedial actions for ground water as described in the DEA and DRAP, which includes characterization of the extert of contamination, identification of remedial action alternatives, and evaluation and optimization of these alternative actions. By pursuing this strategy and satisfying the EPA guidarce previdec for such decisions in Subpart C of 40 CFR Part 19?, DOE should be able to support its decisions.

6. More explanation is needed to adequately justify that the proposed stabilized er.bankment and liner system will protect ground water after re edial actions are completed at the site. As presently describe:' and evaluated, the proposed design does not denonstrate

-adequate protection of ground water beneath the site. Future versiers cf the RAP should include conservative predictions of post-re edial action performance to demonstrate cor.pliance with the EPA sta-dards in 40 CFR Part 192.

f

7. The hyd-: logic characterization of the Gunnisen site alluvial acuifer trovided in Appendix D of the DEA is inadequate. The Theis 6

l n

method used to analyze the test data is not valid because many of the assur.ptions made by Theis in developing the method are violated, Additionally, neither the early or late time / drawdown plots provided as Figures D.2.6 and D.2.7 fit the Theis type curve. Until a more detailed lithologic description of this unit is provided, one must assume that the aquifer is unconfined, poorly sorted, very course-grained and poorly compacted. Use of the Jacob straight line method on the late time drawdown data or the Theis method as modified for thick unconfined aquifers would have resulted in more representive hydraulic properties. Taking the above facts into consideration, a revised analysis of the pump test data should be performed to provide representative hydraulic properties of the alluvial aquifer.

8. Page D-55; It is stated that the difference in hydraulic head between the shallow and deep observation wells indicates that confining layers or lenses are present within the aquifer making it

" semi-confined." This observation by itself does not prove or disprove the existance of silt and clay lenses within the aquifer.

All _ it says is that there is vertical flow within the aquifer.

Unless additional geologic evidence proves the existance of confining layers, this paragraph should be deleted.

9. Page D-55; It is stated that there are higher concentrations of uraniur and iron present in the shallow wells than in the deep wells. The authors attribute this to sorption by clay layers present between the shallow and deep wells. It should be noted that uraniur is a fairly mobile species and would not be expected to be significantly sorbed by clay fractions. It is possibly more reasonaHe to attribute this to a redox phenorena, pH change or the fact that the shallow wells are closer to the contaminant source (bottor of the pile) than are the deep wells. Also, according to the authers, high concentrations of Fe are present in the natural ground water anyway. The text should be modified to reflect the possibility that other natural mechanisms may be responsible for this pheror.enon.

10. Page D-5*.; it is stated that the purp test data indicate the preser.ce of a layer or 2cne which acts like a confining layer. The data presented in Figure D.2.6 may indicate a boundary condition.

However, this rey also represent a no-flow boundary similar to the bedroth ':tarrel boundaries depicted in Figure C.2.5. L'nl e s s 7

ORA py additional geologic evaluation proves the presence of confining layers, this inference should not be used as proof of confinement.

11. Page D-67; According to DOE's plan for implemeting EPA standards for UMTRAP sites, ground water and contaminant modeling will use existing data from each site to explain historic contaminant transport and predict post-remedial action concentrations and rates of migration. The DOE modeling plan is predicated on two invalid assumptions; (1) that sufficient data exists to adequately model the sites, and (2) that the model can be used as a prediction tool.

In the case of Gunnison, we recommend that cherrical and transport models rot be used in a predictive manner because there is not enough data concerning:

a) Dispersivity and scale dependence of dispersivity. Also, TRUMP uses finite-difference approximation to advection-dispersion equations. The validity of this equation is questionable at the travel distances involved in this problem.

b) Uranium and heavy metals are the only contaminants of concern (50 does not appear to be a health hazard at the site). The red,uction in concentrations of these species is governed by cor: lex adsorption-coprecipitation reactions that are not adequately understood. No kinetic or thermodynamic data exist on these reactions. A source term for trace elements computed by PHREE 0E would be meaningless, c) Tre riethod of computing the source tern via PHREE 0E and the ecuation on page D-70 is over sinplified.

One alternative might be to model moisture flow and possibly mass transpert in one dimension through two hypothetical columns of tailings: 1) existing profile, and 2) remedial action profile.

Solute ' lux into the water table for the proposed remedial action would be corpared to that of the original profile. In this way, the effectiseness of the reredial action in isolating contaninants from the grc -d water syster could be evaluated. The flux of contar"irts into the ground water would be lowered by a factor equal 1: the ratio of the two fluxes. Due to the e>istence of retarde .'on and removal processes associated with uranium and other heavy r+ als, this evaluation could be considered censervative.

E e

~

12. Page D-67, 68; FEMFLO, TRUST, and TRUMP are extremely expensive models to utilire. DOE does not have sufficient data to adequately describe the site complexities that would justify the use of such models. We recomend that DOE use any analytical models which may provide valuable information at a fraction of the cost.

DRAP

13. Pages D D-55; The authors attempt to present a case to justify

" semi-confinement of the alluvial aquifer underlying the Gunnison site." However, on page 14 of the draft RAP, the authors are calling the aquifer unconfined. This inconsistency should be clarified.

14 Page 17; (DRAP) The alluvial aquifer is being defined as unconfined, which is inconsistent with data presented in the draft EA. Also, it is unclear as to the basis for detemining that the depth of the aquifer is 130 feet. Data presented in the DEA indicated that the aquifer riay be 200 feet thick under the site and cited literature sources as saying it could be as much as 400 feet thick.

Geology DEA

15. More regional information is needed to adequately evaluate the geology of the site. Supporting regional physiographic, geomorprelogical, sedimentological, stratigraphic, and structural infornation is needed. Evaluation of site-specific data alone (even when adequate local inforr.ation is provided) cannot result in a proper assessment of the long-term suitability of a site in relation to the EPA requirenents (40 CFR 192). Given the nature of geological processes, past, present and future geologic conditions at any given localized site are inextricably linked to the geologic character of the larger region (geologic province) in which it is located.

Standard professional nethods and procedures recuire that adequate geologic investigations cf a site include conplenentary infernation to describe:

1. Tre gencorphic cerditions at the site and their relatier to the regional geonorphic settino.

9 i _

?. The site stratigraphy and its relationship and correlation with the regional stratigraphy,

3. The structural geology of the site and the relationship of site structure to regional tectonics.

4. The geologic history of the site as it relates to the regional geologic history.

5. The hydrocarbon, metallic and non-metallic mineral resources of the site as they relate to regional or analogous trends.

The regional information required usually can be derived primarily from a review of the existing, pertinent geologic literature. The inforration should be documented by references to all relevant published and unpublished material. The Ut1 TRAP document review process will be expedited if the DOE submittals contain sufficient inforr.ation for the reviewer to make an independent assessment of the corclusions regarding the geologic suitability of the site.

16. With regard to the terrace at the level of the site, the location of the terrace scarps and the projected duration of the terrace relati ve to the 200- to 1,000-year stability requirement for the pile steuld be identified.

Geotechnical DEA

17. Fage 2C; Reference is made to a 2-foot thick clay filter" layer.

This terr. is anbiguous as to whether the layer is to function as a liner er as a filter for the capillary break layer. Please clarify the terr and indicete the purpose of the clay.

18. Pages 21 and 37; The construction activities listed for the two alterrative sites include placement of " clay filter" and capillary break layers. We are not convinced that it is appropirate to assure that tFese design aspects are necessary without further characterization of the ground-water regire at the sites. In additicr, there is insufficient data te support conclusions related to the asailability of site soils for use in liner or cover layers.

Sore tasic site characterizatior. work at the te alternative sites is netted to enable a F0re represertative Cost Corparison.

10 L

~

DRAFT

19. Page 24; The discussion on cover construction indicates that long-terr. moisture contents of 5 percent (top 1 foot) and 10 percent (bottor 4 feet) were assumed for radon barrier calculations.

However, the test data on the silty-clay borrow material (DSCR) indicates in-situ moisture contents ranging from 4 to 9 percent (5.8 percent average). Justify your use of 10 percent. In addition, Item 4 of E-17 states that long-term moisture contents are based on calculations using site-specific data. The method used for estimating long-term moisture contents should be clarified.

20. Page E-35; The rock durability criterion of less than 10 percent loss af ter 12 cycles of freeze / thaw appears to be too low. In addition, the criterion for the Schmidt Hamer test appears to be in error. (

Reference:

NUREG/CR-2642, Table 6.2.) Provide the basis used to determine the rock durability criteria and verify the criteria for the Schmidt Hammer test.

21. Page E-39; The determination of the need for a waste water treatrent plant cculd significantly impact the cost conparisons presented in the DEA. The determination should be included in future versions of the EA and RAP.

SCR

22. The DSCR does not adequately characterize the subsurface materials at the site. Provide the results of the 22 borings performed by Ford, Bacon and Davis (GC series) and any other new or historical data te characterize these materials. This information is necessary to evalt. ate the liquefaction potential of underlying foundation '

soils.

Radiological D_E A_

26. Page 77; Because of seasonal fluctuations in radon concentrations, the average outdoor background should be an annual average rador concertration. The average outdoor background radon concentratier provided in the DEA may not be representative of actual background.

In orde- to preside a representative average outdoor background radon c:ncentration an arruel average should be deterrined by taling three cr four 24-hour sarples at the five sarpling locatiors in differer t seasons of the year.

11 L

24 Page 79; Uranium, Pb-210, Th-232 and Th-230 concentrations in the tailings should be provided along with a discussion on why these radionuclides are or are not included for health effect considerations. A comparison of these values with natural backgrourd should also be made. The DEA should be modified accordingly in order to provide a complete analysis of the radiological health effects from radioactive particulates.

25. Page 97; Ra226 is a bone seeker and whole body risk estimates should include this fact rather than assigning a whole body risk factor on the basis of a dose to the bronchi only.

26. Page 101. Paragraph one refers to " airborne radioactive particulate concentrations yet the DEA only addresses Ra-226.

Therefore the analysis as presented in the DEA is incomplete. The effects cf airborne Th-230 and Pb-210 should also be addressed.

This lack of information affects the assumptions in the second paragrag+ as well.

27. Page H-E; thisThe

.a. DEArisk factor is 100 X f10-gr excess fatal lung cancer, which indeaths per pers general population and for the remedial action worker. The Evans et al (1961) reference, which gives the primary justification for using this risk factor, states that workers are a higher risk than the general population for equal exposures to radon daughters. A higher risk factor comparable to those recomended by UNSCEAR and used by the NCC should be applied to the remedial action worker,

b. Cor;aring total organ doses over 50 and 100 years for both workers and the general population would help to clarify the differerce when compared to expected background exposures rather thar corsaring only relative risk.

28. Page H-7; Second paragraph. The dose conversion factor for 1 WLM in BEIF :11 has been superceded by NCRP *77 (1984) which lists 14 REM /WLP.

29. Page F-9; The average wind speed of 2.1 neters per second from the 1sbill Associates source should be used rather than the average 2.7 nete-s per second fror the Fercheau reference. The Isbill wind data was collected over a 5-year period, whereas the Fercheau data 12 L.

F

~

DRAFT was collected over a 2-year period. Also, because the Isbill data is a product of NOAA's Environmental Data Service, it is probably a better data source. With the Ishill average wind speed, a 30 percent greater on-pile radon concentration remedial action worker health effect is obtained.

30. Page H15; Equation at top should list 730 hours0.00845 days <br />0.203 hours <br />0.00121 weeks <br />2.77765e-4 months <br /> /WLM for a member of the public.

DRAP

31. Page 8 Section 2.5 and Page A9; states that, when working levels are between 0.02 WL and 0.03 WL, the government will have the flexibility to decide if measures should be taken to reduce working levels. This is inconsistent with the EPA standard in 40 CFR 192.12(b)(1). The standard requires that a reasonable effort be made to reduce working levels to below 0.02 WL. A decision to take no action would constitute the application of supplemental standards. The DRAP should be modified to reflect compliance with the EPA standards.

32. Page 36; It appears that dust control will depend exclusively on spraying. The DRAP should recognize the possibility of extreme dust conditions and require more restrictive controls when warranted.

Controls such as reduction or stoppage of work should be considered.

The DRAP should be modidied accordingly.

I?

t