^ % \

-*>. : A

^joj^^ SFUND RECORDS CTR

• ' ^ j ^ ^ 7 . 8 8 0 1 5 0 4 0 ^^^ * * ° RECORDS CTR 0222-00478

UNITED STATES ENVIRONMENTAL PROTECTION AGENCY

' ^ i pRdXt' *'' REGION IX

75 Hawthorne Street A R 0 6 9 1 . San Francisco, Ca. 94105

CERTIFIED MAIL P 347 535 715 RETURN RECEIPT REQUESTED

November 26, 1990

William R. Victor Errol L. Montgomery and Associates, Inc. 1075 East Fort Lowell Road, Suite B. Tucson, AZ 85719

Re: Comments on Remedial Investigation Report and Liquid Waste Evaluation Hassayampa Landfill Site, Maricopa County, AZ

Dear Mr. Victor:

Attached with this letter are the U.S. Environmntal Protec­tion Agency's (EPA's) comments on the following documents:

- Draft Remedial Investigation Report for Former Hazardous Waste Disposal Area at Hassayampa Landfill Maricopa County, Arizona, dated October 11, 1990, and

- Liquid Waste Evaluation - Hazardous Waste Area, Hassayampa Landfill Maricopa County, Arizona, dated October 9, 1990.

Also attached are comments submitted by the Arizona Depart­ment of Environmental Quality (ADEQ) and the Arizona Department of Water Resources (ADWR) regarding the same documents. For the most part, EPA's comments are inclusive of those provided by ADEQ and ADWR; although there are a few exceptions. The Remedial In­vestigation (RI) report was generally found to be well-written; however, the concerns presented in the comment letters must be resolved before EPA can approve the report. Several of the com­ments identify data gaps, which could potentially limit the ability of the Feasibility Study to develop and compare various remedial alternatives for the site. EPA would like to meet with the Technical Work Group as soon as possible to discuss this issue.

According to Subsection C.S of the Administrative Consent Order (U.S. EPA Docket No. 88-08), a fifteen (15) day period for the purpose of technical meetings between EPA and the Respondents is available begining the day after your receipt of this letter. Furthermore, according to the schedule presented in Table 3 of the Work Plan, the Respondents have a maximum of ten weeks in

Printed on Recycled Paper

-2-

which to finalize the RI Report. In order to expedite review of the final RI Report, EPA requests that you provide a brief narra­tive along with the revised RI report, which describes how each of the comments submitted by EPA, ADEQ, and ADWR were addressed.

The exact manner in which the Liquid Waste Evaluation (LWE) fits into the RI/FS framework established for the site is un­clear. If the LWE is to be referenced by the RI report and Feasibility Study report (and hence become part of the Ad­ministrative Record for the site), it must be revised in accor­dance with coinments provided by EPA and ADEQ. For reasons out­lined in EPA's comments, results presented in the LWE will not be used to support the Risk Assessment which is currently being con­ducted by EPA.

If you have any questions regarding the attached comments or would like to schedule a time to meet, please call me at (415) 744-2368.

Thomas J. Dunkelman Project Coordinator

Attachments

cc: Jackie Maye, ADEQ Grant Gibson, ADWR Robert W. Hacker David Machlowitz David P. Kimball, Esq. Charles A. Bischoff, Esq. Kim E. Williamson, Esq. Errol L. Montgomery Richard C. Keiffer Ronald Frehner Robert H. Brauer, Esq. Roger K. Ferland, Esq. G. Van Velsor Wolf, Esq.

Carl C. Meier, Esq. James Derouin, Esq. Alan Abbot Robert Cameron Lt. Col. Ray Swensen Kevin Milliken, Esq. G.S. Hagy Charles Geadelmann Cindy Lewis, Esq. G. Eugene Neil, Esq. William J. Cheeseman, Esq. Terry A. Thompson, Esq.

EPA Comments on the Remedial Investigation Report Hassayampa Landfill Site

EPA's RI comments are grouped into two categories: General Comments typically refer to a more general theme that is present throughout the report; while Specific Comments refer to a specific issue on a given page. In some instances, symbols are provided which allow cross-referencing between General and Specific Comments.

General Comments:

1. The RI report does a good job of presenting the results of the investigations conducted; however, the level of interpreta­tion and analysis needs to be expanded. Several examples of this are given in the Specific Comments section. In particular, in­terpretive analyses of the nature and extent of contamination and fate and transport of contaminants are lacking in detail. The inclusion of such information in an RI report is discussed in Sections 3.4.1.3 and 3.4.1.4 of the EPA manual "Guidance for Con­ducting Remedial Investigations and Feasibility Studies Under CERCLA" (October 1988). The RI should include a separate sec­tion, near the end of the report, which interprets the results presented, particulary with respect to fate and transport of con­taminants. Those Specific Comments which pertain to issues which should be discussed in this section are designated by the symbol llpm II

2. EPA has several concerns regarding data coverage for the site. The Agency is not anxious to request additional field work which will lead to an increase in the cost or duration of the RI/FS. Nonetheless, EPA feels that there are a few pieces of critical information which are lacking from the RI report. Without this information, EPA believes that it would be difficult to assess various remedial alternatives for the site. Those Specific Comments which pertain to potential additional data re­quirements are designated by the symbol "DR". Generally, the potential additional data requirements perceived by EPA are as follows:

Installation of additional angle borings beneath Pits 1 and 3 in order to better define soil contamination with depth. This information could be used to better es­timate the volume of waste in the pits, and determine whether deep soil contamination is acting as a continu­ing source of groundwater contamination. Installation of a monitoring well in Unit B, in the vicinity of MW-6UA. Soil gas survey to determine if unidentified con­taminant sources exist in the Special Pits Area.

3. The RI report should include a section entitled Data Limita-

tions and Data gaps which assesses the implications of any such limitations and gaps. Those Specific Comments which pertain to data limitations are designated by the symbol "DL". Furthermore, any issues identified by General Comment 2 which are not resolved by further sampling should be discussed in this section.

Specific Comments:

1. - p . 3 4 - 3 5 . Although it is mentioned on p. 22, the Waste Types and Quantities section should reference the types of wastes placed in Pits A and B. Furthermore, this section should include a more detailed description of the types and quantities of waste placed in the Special Pits Area. According to p. 5 of the LWE, a summary of the types and quantities of wastes disposed in the Special Pits is tabulated by the Bureau of Waste Control (1980).

2. - FT, p. 44. The RI report needs to expand upon its inves­tigation and discussion of upper alluvial deposits as interpreted from the wall of the large excavation located within the landfill. This excavation provides a valuable source of informa­tion which has not been adequately presented in the RI. This should include a detailed geologic description of coarse and fine-grained sub-units, a description of sedimentary structures present, and a descrip-tion of any other features that might in­fluence the hydrogeologic properties of this unit. In the opinion of EPA personnel and contractors, the observed lithology of the cut surface suggests that liquid wastes disposed onsite could have easily migrated vertically through this unit. The presence of lateral-vertical lenses of sand and gravel and thin, caliche-filled fractures would contribute to the ability of wastes to migrate through this unit. The reaction of acids from the pits with carbonate cement present in the upper alluvial deposits would could also increase permeability beneath the pits. If such models are accurate, it is possible that evaporation from the disposal pits may have been overestimated, and downward migration of contaminants may have been underestimated.

3. - FT, p. 48. The discussion of the basaltic lava-flow unit should include a description of those locations where this forma­tion was observed to outcrop. This should include a description of the nature of the outcrop, fractures, voids, or any other fea­tures of the basalt that could have a bearing on its hydrogeologic properties.

4. - FT p. 50. The structural contour analysis presented here is very interesting and should be expanded upon. If sufficient data exists, structural contour maps should be prepared for the top of Units A and B. Also, a structure-contour map depicting the thickness of the basalt unit would be helpful. To the extent possible, the structure contour analyses should be used to dis­cuss fate and transport of contaminants.

5. - Fig. 3 and Fig. G-1. Pit dimensions illustrated in these

two figures do not match very well, particularly for Pits 1, 2, and 3a. These figures should be revised so that the pits dimen­sions are the same or similar in both figures.

6. - p. 59. The depth of trenches ranged from 4-12 ft. below land surface, but generally did not penetrate more than 1 ft. into top of iDuried waste. If the soil cover across the site is 1-2 ft. thick, how can this statement be correct? This issue comes up several times throughout the report, and requires some clarification.

7. - DL p. 59. At most there is only one waste sample point per pit. Some pits have no such sampling points (i.e. Pit 3b, Pit 4a, Pit A, Pit B, and Special Pits). Provide rationale describing how the waste sample locations were selected. Also include a discussion of the likelihood that the waste sample is representative of the waste contained in the pit.

8. - DL p. 66-69. The comparison of waste data to angle boring data and the statements regarding lack of downward migration of contaminants may not be as significant as presented here. This is particularly true for Pit 1 and Pit 3c, where significant horizontal distances (up to 50 ft) separate the waste sample locations and angle boring locations. The RI assumes that the waste sample is representative of waste in the pit, but there has been no discussion of this. The RI should discuss this issue and whether it is appropriate to make such data comparisons.

9. - FT, DR, DL p. 66-73. EPA is concerned about the data coverage for soil and waste samples. As mentioned in previous comments the waste sample coverage is poor. This is also true for soil samples. Several significant data gaps exist with respect to soil and waste characterization.

a. The vertical extent of waste within the pits and shal­low soils is poorly defined, and as a result it would be difficult to estimate the volume of material poten­tially requiring excavation or treatment.

b. It is possible that contamination within the Special Pits Area could have gone undetected, especially since groundwater contamination at MW-IUA has not been ade­quately accounted for. A soil gas survey could be employed to detect any such sources of contamination.

c. The contaminant profile beneath Pits 1 and 3 is poorly understood due to the limited number of soil borings. The highest level of soil contamination was detected in the deepest sample (AB-3 - 60 ft). Additional soil borings should be installed in order to better under­stand how the contaminant profile changes with depth, and to determine whether soil contamination extends into the basalt unit and Units A and B. Furthermore, these issues need to be addressed in the Fate and Con-

taminant section. Questions such as the following need to be answered:

Are contaminants "pooling" on top of the basalt? Is it possible that paleo-relief along the basalt unit is influencing the patterns of soil and groundwater contamination? Does the soil contaminant profile extend through the basalt into the underlying units? Is soil contamination at depth acting as a continuing source of groundwater contamination?

10. - PT p. 70-71. The discussion of non-targeted compounds detected needs to be expanded. In several samples, high con­centrations of non-targeted compounds were detected - primarily hydrocarbons. The presence of these hydrocarbons should be noted in the text, and any implications regarding the presence of these hydrocarbons should be discussed here and in the Fate and Transport section.

11. - p. 73. This summary needs to be revised to reflect previ­ous comments.

12. - FT 75. The issue of vertical versus lateral spread of con­tamination needs to be discussed in more detail here and par­ticularly in a Fate and Transport Section. The fact that sig­nificant soil contamination has been detected immediately above the basalt, and that groundwater contamination has also been detected, indicates that significant vertical migration of con­taminants has occurred. References to the LWE must be removed from the RI report unless the LWE is revised to reflect EPA's comments.

13. - DL p. 81-85. The RI report should discuss the fact that air sampling was only conducted over a one day period, and that the data gathered is not necessarily representative of conditions at the site during different meteorological conditions. The report should also indicate that Stage II air data was collected under atypical conditions (cloudy, after a rain event). Further­more, the Stage II analytical results section should discuss the fact that contamination was detected, and that some compounds detected during Stage II air monitoring were present at higher concentrations than detected during Stage I air monitoring. It is possible that additional air monitoring may be necessary during the remedial design/remedial action phases.

14. - p. 89. The RI report should include a table with coor­dinates for monitoring well and soil boring locations.

15. - p. 106, Appendix L, p. 18. Provide further explanation for the estimate of storage coefficient of 0.10 under long-term pump­ing conditions, when the range under short-term pumping condi­tions was .000015 - .000096. A storage coefficient of 0.10 is typical of unconfined aquifers, while a storage coefficient of

.000015 - .000096 is typical of confined aquifers. Please clarify this.

16. - p. 105, p. 107, Appendix L. What interpretations can be made from the shape of drawdown and recovery curves for Units A and B?

17. - DL p. 105, p.107. The aquifer tests provide little or no information regarding the interconnection of Units A and B. Such information may be important during the remedial design/remedial action phases and should be identified as a data limitation/gap.

18. - p. 115. TCA was detected in MW-8UA. If this is a result of laboratory contamination, please clarify. If not, then this contamination should be noted in the text.

19. - Table 31. 1,2 DCE was detected in MW-6UA at 1.2 micrograms per liter according to Table 30, rather than ND as is shown in Table 30. Please clarify.

20. - DL p. 111. This section of the RI report should include further discussion of the groundwater sampling methods. Specifi­cally, discuss the impact of using submersible electric pumps as opposed to other non-aerating methods. If the sampling method introduces any bias into the data, then this issue should be dis­cussed in the data gaps and limitations section. In the future, it may be necessary to revise groundwater sampling methods.

21. - p. 117. There are several semi-volatile compounds presented in Table 3 4 which do not appear to result from laboratory contamination (eg. see MW-2UA, MW-6UA). If these con­taminants are a result of laboratory contamination, please clarify. If not, their presence should be noted and discussed in the text.

22. - FT, DL p. 115, p. 110. The RI should attempt to explain the presence of contamination in MW-IUA. Is this the result of contamination entering the water table from a point beneath Pit 1, or is it the result of contamination entering the water table from another source (Special Pits?)? Is there is enough informa­tion to answer this question? If not, what further information is necessary? Assuming that the source of contamination in MW­IUA is a point beneath Pit 1, would this conflict with calculated groundwater flow velocities? If so, please discuss.

23. - FT, DG, DL p. 116, p. 118. The detection of significant contamination in MW-6UA seems to confirm that contamination in HS-1 was not merely a result of improper drilling techniques. This issue needs further discussion. Furthermore, the presence of significant contamination in HS-1 and MW-6UA coupled with an observed downward vertical gradient suggests that groundwater contamination Unit B is likely; yet no monitoring wells are in­stalled in Unit B at this location. It cannot be argued that MW-4UB proves a lack of contamination in Unit B, since MW-4UA did

not reflect the contamination observed at MW-6UA and HS-1. This represents a significant data gap, and it is possible that groundwater contamination in Unit B could go undetected in light of the current monitoring network. This situation must be remedied.

24. - FT, DL p. 115 - 119. An expanded level of interpretation of the results of groundwater monitoring is needed here and in the Fate and Transport section. The RI should attempt to explain the spatial and temporal patterns of groundwater contamination. It appears that groundwater contaminant levels are increasing over time. This is a very important issue which is not discussed at all. If insufficient data exists to explain contaminant pat­terns, the RI should state what data would be required to com­plete this explanation. Furthermore, questions such as the fol­lowing should be answered:

How are the overlying units affecting spatial and tem­poral variations in groundwater contamination? Can the results of the structural-contour analyses be used to explain variations in contaminant patterns? Why is groundwater contamination generally increasing over time, and what are the implications for site remediation? Is it possible that the basalt unit is introducing "lag time" into detection of groundwater contaminants? Is any vertical speciation of groundwater contaminants occurring?

25. - p. 129. As is discussed in the General Comment 2 and many of the above Specific Comments, EPA does not necessarily agree with the statement here that "Additional testing for the Feasibility Study outside of the Remedial Investigation is unnecessary." As is discussed above, in General Comment 2 and in several Specific Comments, EPA has identified several potential additional sample requirements:

26. - Executive Summary. This section should be revised to reflect the comments provided above. At a minimum, the following conclusions need revision:

p. 5, no. 6. This conclusion needs to be rewritten in light of previous comments; p. 6, no. 7. As is discussed in previous comments, references to the LWE must be removed from the RI report unless the LWE is revised to reflect EPA's com­ments ; p. 11, no. 14. Based on observed patterns of con­tamination, it appears likely that groundwater con­tamination may exist in Unit B in the vicinity of HS-1 and MW-6UA. This issue needs to be addressed here. Also, the fact that groundwater contaminant concentra­tions are generally increasing over tirae needs to be addressed.

p. 12, no. 15 and 16. EPA disagrees with these conclu­sions. p. 12, no. 17. The air monitoring program did detect contamination, and this fact should be noted in this section. Questions pertaining to the relative risk posed by these contaminants will be addressed by the Risk Assessment, and as a result discussion of this issue is not appropriate in the RI. Conclusion 17 should be removed. Those issues raised in the General Comments must be ad­dressed in the Conclusions.

EPA Comments on Liquid Waste Evaluation Hassayampa Landfill Site

The bulk of EPA's comments on the LWE are provided in a November 20, 1990 letter from Randolph Fish to Thomas Dunkelman, a copy of which is attached. A few additional comments are provided below.

1. The amount of subsurface data available significantly limits the capability of the LWE to accurately model site conditions. The issue of limited subsurface data is discussed extensively in the RI comments.

2. p. 26, par. 2. The statement, "A representative estimate of evaporation and infiltration is considered by CRA to the the range represented by Case 1 and Case 2," is not adequately sup­ported and appears to be no more than a guess. Unless this statement can be further supported, it should be removed. Fur­thermore, any conclusions in the LWE and RI report should cite the full range of evaporation and infiltration as represented by Cases 1 through 4.

3. p. 26, par. 3. The following sentence must be removed from the LWE: "Further, the results of the RI for the Site do not in­dicate the presence of significant quantities of waste material in the groundwater." Groundwater contamination at the site ex­ceeds MCLs for several contaminants, thus indicating the presence of significant contamination. Furthermore, given the current network of monitoring wells and the fact that the basalt unit is likely to slow contaminant movement, it is possible that larger amounts of liquid waste have gone undetected.

4. p. 27, par, 2 and 3, As was discussed previously, EPA does not agree with the representative range concept. Also, EPA dis­agrees with the statement that the representative ranges are an overestimate. Based on on comments provided it is possible that the LWE may have significantly underestimated infiltration at the site. The fact that significant soil contamination has been detected immediately above the basalt, and that groundwater con­tamination has also been detected, indicates that significant vertical migration of contaminants has occurred.

TO -fOTTi/) l77ju^xi7^k7bU7uuayiA:>

Bf>/^ ^

ijsfv^ AK- hKfi'' , SUBJECT T^E^lA/Ey^ ^* CJJfy)fy^S7^7TS <n^ DATE / A f ? f ? a

//AiSSAYAWA)A>S ^ / ^

LJ&ru J j72/j77^ SL Jju4A>£3U n±tA\jr^Ao_nlJ ^l^u.tAy77»$j / r u 2 y ) J x j j . TPlAJA^ jTJTDJdb ^ A : ^ Q J - 76Ai>^

\^i£yiAi7t i t ^ l A ^ jTS^ry^/yja4^ciaJ C ^ A A ' Q J C ; ^ : : ^ 7^ ^ ^ ' ^ ^ ^-^

SIGNED

R E P L Y

sJflT cu ^mTU^ ^ - ( ^ ^

4

f

SIGNED DATE / / 1 REDIR^JIM, 4S ^72 SEND PARTS 1 AND 3 INTACT -

PART 3 WILL BE RETURNED WITH REPLY. POIY PAK f50 SETS) 4P472

ARIZONA DEPARTMENT OF ENVIRONMENTAL QUALITY

Inter-Office Memorandum

H-4994.01

November 9, 1990

TO: Jacqueline Maye, Project Manager Superfund Coordination Unit

THRU: Michele Kennard, Acting Manager"^)^-^ Superfund Hydrology Unit

FROM: Michael Leach, Hydrologist/vij^ Superfund Hydrology Unit

RE: Hassayampa Landfill Remedial Investigation and Liquid Waste Evaluation Reports

The draft Remedial Investigation Report (RIR) and Liquid Waste Evaluation (LWE) report for the above-referenced facility have been reviewed. In general, the information contained in the RIR is simply a compilation of past data and there has been no change in the conclusions drawn from this data or the recommendations which call for no further site investigation activities. The LWE attempts to show that a large amount of the liquid wastes disposed at the site were evaporated and that very little of these liquid wastes could reach the water table. Specific comments have been generated for each of these reports which will be discussed separately.

A. Remedial Investigation Report

1. As I have commented before in previous submittals (copy of May 30, 1989 memo to Kristie Kilgore of ADEQ from myself is given as Attachment 1) regarding the fate of liquid wastes upon disposal at the Hassayampa Landfill, it has long been my opinion that the estimates of liquid waste evaporation, retention in the vadose zone, and lateral migration in the vadose zone are all overestimated. As a result of the Technical Work Group Meeting on August 8, 1989 at EPA Region 9 headquarters in San Francisco, California, memb^s of Errol Montgomery and Associates (EMA) and myself once again agreed on very few aspects concerning the fate of liqui 'wastes upon disposal at Hassayampa Landfill. However, we did a^ee that the proposed slant borings (at the time slant borings had not been installed) beneath each pit would determine what had actually happened to the liquid wastes discharged into the pits.

ADEQ/ADM/Stock (12/88) — IPS 3171

Laboratory results of the samples obtained from slant boring completion beneath each disposal pit have been summarized on page 5 of the RIR. As is evidenced by the data, a great deal of both HVOC and semi-VOC contamination exists to the bottom of the angle borings completed beneath Pits 1 and 3 (both of which received a large amount of liquid wastes). Based upon the site-specific facts, it appears that a great deal of soil contamination beneath Pits 1 and 3 has occured along with the non-detect of these contaminants in soil borings surrounding these pits. These facts suggest that very little lateral migration of contaminants occurred in the vadose zone upon percolation.

2. My opinion on the occurrence of HVOC contaminants detected in ADEQ monitor well HS-1 remains the same as is contained in Attachment 1. Subsequent detection of high concentrations of HVOCs in MW-6UA (which was installed in place of HS-1 after HS-1 was abandoned in 1988) would appear to confirm that improper drilling techniques for HS-1 were not responsible for the presence of HVOCs in this well, and would also tend to support the theories of subsurface contaminant migration presented in Attachment 1.

3. It is unclear whether or not page 35 of the RIR has taken into account the several thousand gallons of solvents which were disposed into the Special Pits Area.

4. My opinion of the locations of monitor wells MW-7UA (well B) and MW-8UA (well C) remains the same as stated in Attachment 1.

B. Liquid Waste Evaluation

The purpose of this report is unclear. The report attempts to show that much of the liquid waste disposal into the pits should have evaporated, resulting in very little or no contaminant migration to the underlying aquifer. However, since there is not enough subsurface soil contamination inforraation to verify the results of the raodeling, the raerit of the raodeling effort is questionable. It is tirae to face the facts, soil and groundwater contamination does exist at the site as a result of waste disposal activities, and further efforts (whether site investigation or remedial) should focus on the eventual clean up of the site. Specific comraents on the LWE are given below:

1. Page 20 of the LWE states that the vertical hydraulic conductivity value used in the evaporation model was obtained from a single boring (SB-14) located on the eastern boundary of the landfill. Basing a raajor coraponent of the evaporation model on a permeability test from one discrete depth (25 feet) at a single point in my opinion is quite inadequate. Additional permeability tests on shallow soil saraples are needed from various areas at the site in order to determine such an important component to the evaporation model.

2. Page 22 of the LWE states that the migration front of contaraination in the vadose zone would move at a rate of 2:1, vertical to horizontal. It is my opinion that very little horizontal movement would occur in the vadose zone for the reasons given in Attachment 1. In addition, it is well-known that upon discharge into a horaogeneous media a liquid will move primarily in a vertically downward direction with very little horizontal moveraent , regardless of higher permeabilities in the horizontal direction. The volurae of contaminated soil predicted for the model is therefore too high and greatly overestimates the amount of liquid waste retained in the vadose zone, and consequently greatly underestimates the ability of liquid waste to reach the underlying aquifer.

3. Page 26 of the LWE states that the results of site investigation activities do not indicate the presence of significant quantities of waste raaterial in the groundwater. Firstly, significant quantities of waste raaterial have been detected in the groundwater since MCL's have been exceeded for sorae constituents. Secondly, due to the slow rate of contaminant raovement in the aquifer along with the fact that monitor wells are not properly situated to intercept areas of the contarainant pluraes which would be expected to have the highest HVOC concentrations, it is still unknown whether or not large araounts of liquid wastes have reached the underlying aquifer.

C. Recommendations

1. My recommendation previously given in Attachment 1 regarding soil boring completion remains the same except for the notable exclusions given below.

The slant borings appear to be adequate for detecting contamination directly beneath the pits, and therefore nothing should be altered from the way that the original slant borings were installed. Delineation of the pit dimensions helped greatly in the effectiveness of the original slant borings to detect subsurface contamination from liquid waste disposal in the pits.

Of the four pits identified, only 1 and 3 require additional subsurface soil investigations. This determination is based upon the results of the original soil boring program.

2. My recoraraendations previously given in Attachment 1 regarding monitor well installation remain the sarae. Due to the detection of the HVOC contarainants in MW-6UA it recoraraended that raonitor wells D and E also be completed at their originally prescribed locations.

Source(s) of HVOC contamination in the Special Pits Areas needs to be determined. A soil gas and/or soil boring prograra needs to be implemented in order to effectively site remedial facilities for eventual vadose zone clean-up.

ML:smp

•^-

ARTi' ) A DEPARTMENT OF ENVIRONMENTAL Qi> lilTY

Inter-Office Memorandum

•/ H-4994 J

DATE: May 30, 1989

TO: Kristie Kilgore, EHS III Remedial Projects Unit

THRU: Chuck Graf, Manager 7^ iT t i -Superfund Hydrology Unit ^ - ^ ^ ^

FROM: Michael Leach, Hydrologist ^^)L Superfund Hydrology Unit

RE: Hassayampa Landfill - Review of Stage I Report for the RI/FS and Supplemental Task F Work Plan

The Stage I Report for the RI/FS study and the Supplemental Task F Work Plan for Hassayampa Landfill have been reviewed. Comments generated for both reports are presented in the following attachment.

MLrblv

Attachment

(memo/lndf.has/g)

ADEQ/ADM/Slock (12/88) — IPS 3171

Ms. Kristie Kilgore May 30, 1989 Page 2

ADEQ Groundwater Hydrology comments on Stage 1 Report for the RI/FS and Sup­plemental Task F Work Plan.

A. FATE OF LIQUID WASTES UPON DISPOSAL:

Several parts of the Stage I RI/FS report mention that, due to the presence of low-permeability sediment layers (predominantly clays) which exist in the vadose zone at the base of the Upper Alluvial Deposits (UAD), any percolating liquids would tend to mound over these layers and result in a significant amount of lateral migration. However, evidence of this phenomenon was not observed in soil samples obtained from the 12 on-site soil borings. These soil samples exhibited low soil moisture contents and laboratory analysis did not detect the presence of any contaminants. The report concludes that much of the hazardous liquid must have evaporated in the disposal pits when exposed to the atmosphere or were absorbed by sedi­ments in the vadose zone, and never reached the water table.

However, several lines of evidence indicate that liquids disposed at the .. -" site can, and in fact, have reached the water table without being detected

in the vadose-zone by soil borings.' Sections A.l through A.4 discuss this" reasoning below:

A.l Liquid Waste Evaporation:

•" Although there is little doubt that some evaporation of the' liquid wastes did occur in the pits, the importance placed on evaporation as a cause of solvent removal from the pits is probably significantly overestimated. This is due to the fact that, although halogenated volatile organic compounds (HVOCs), the targeted contaminant for this site, are by nature prone to evaporation, a great quantity of HVOCs and a great variety of liquid wastes other than HVOCs were disposed into the pits. These include waste oils and water-diluted liquid wastes (primarily Pits #1 and #3). Since HVOCs are almost without exception more dense than both oil and water and were disposed in quantities that indicate that the solubility limits of the HVOCs were exceeded, the HVOCs would be expected to sink into the liquids and not be exposed to significant evaporation at the liquid surface. Such a layering effect would also increase the percolation of the HVOCs into the subsurface due to the increased pressure of the over­lying, less dense liquids.

Ms. Kristie Kilgore May 30, 1989 Page 3

A.2 Retiention of Liquid Waste in the Vadose Zone:

Although some of the liquid waste is undoubtedly retained in the vadose zone as it percolates down, the report overestimates the degree of liquid retention. It states that a large amount of liquid retention would occur in the vadose zone, but no calculations are given which would even roughly approximate the ability of vadose zone sediments to retain liquids. A review of the lithologic descriptions of the on-site soil borings (Appendix F) indicates that the upper 25 to 40 feet of the UAD consists primarily of gravelly sands, sands, and somewhat silty sands. Percolation through such high permeability sediments would be primarily in the downward direction in response to gravity, with \er-y little or no lateral migration (which would expose other dry soils to the liquids and possibly result in the retention of additional water in the vadose zone). In addition, sandy sedi­ments, even those that are initially dry, retain very little moisture subsequent to saturation and internal drainage. Therefore, only small amounts of liquids could be retained in the upper, coarse­grained sections of the vadose zone and any significant storage would have to occur in the lower, fine-grained sections (primarily silts and c} ays) ofthe vadose zone. __-_„

A.S Lateral Migration of Liquids in the Vadose Zone:

The report appears to exaggerate the potential of the fine-grained sediment layers of the lower UAD to promote a significant amount of lateral migration of liquids in the vadose zone. Given a constant liquid surface source, the extent to which lateral migration of a mound will occur depends primarily upon the saturated and unsaturated properties of all sediments involved, the liquid percolation rate, the total volume of liquids available for percolotion, and the an­tecedent soil moisture conditions in the vadose zone. Unsaturated flow of these liquids may laterally extend the effects of mounding by increasing soil moisture contents out to a certain distance as governed by capillary soil forces. However, upon reaching a maximum lateral extent, liquids within the mound (especially near the center of the mound where overlying liquid pressures would be at their highest) will move down through the low-permeability layers in response to gravity. Due to the fact that each disposal pit repre­sents a finite quantity of liquid and because the seepage rate, q, from each disposal pit was probably not large, little lateral migra­tion of liquids would be expected which would explain why ncne of the soil borings encountered soils with high moisture contents.

1^^ Ms. Kristie Kilgore May 30, 1989 Page 4

.r:

A.4 An On-Site Example:

An on-site example appears to demonstrate that migration of fluids from the pits to the groundwater is possible, and in fact, has ap­parently already taken place. Although this example does not involve hazardous liquid wastes, it still indicates that disposed liquids can reach groundwater and not be detected in nearby soil borings. The example derives from the hydrogeologic data obtained from two soil borings (SB-4 and SB-5) and a downgradient monitor well (MW-5UA) in the Pit A area.

Pit A was constructed simultaneously with the other disposal pits in the hazardous waste area and began receiving liauid septic tank and cesspool wastes in 1979 or 1980. Prior to September, 1988, each of the on-site monitor wells had undergone two sampling rounds and the chemical quality of groundwater obtained from MW-5UA was typical of groundwater obtained from the monitor wells in the hazardous waste area. In the following two groundwater sampling rounds (9/13/88 and 12/5/88), however, concentrations of sodium, chloride, nitrate, and total dissolved solids (TDS) were at least twice as high as those ob­tained from' the two. original sampling events. In addition,''detec-'" table levels of phosphorous were found in the 9/13/88 sample. These chemical constituents are all inherent in domestic wastes and indi­cate that MW-5UA had intercepted a contaminant plume emanating from Pit A. Other explanations for this phenomenon such as incomplete well development or well purging at MW-5UA are unlikely since no other well at the site showed this dramatic increase in TDS con­centrations.

Lithologic descriptions of the drill cuttings obtained from SB-A and SB-5 (both located neat Pit A) indicate that within the lower silt and clay layers of the UAD, the soils are described as being only "slightly moist". Also, a laboratory soil moisture measurement of the silt layer situated directly above the basal clay layer of the UAD (above which is presumed that any downward migrating liquids would mound and being to migrate laterally) in SB-5 only had a soil moisture content of 18.5 percent. Silt and clay mixtures will typi­cally have porosities ranging anywhere from 40 to 50 percent. An 18.5 percent moisture content indicates that the soil is indeed only "slightly moist" and has never been saturated since fine-grained sediments typically have ve ry high field capacities (27 to 35 percent; Dunne and Leopold, 1978). Because moisture contents in SB-4 and SB-5 are much lower than the field capacity values and since these soils are situated too deeply to be affected by evapotrans­piration, the SB-4 and SB-5 soils apparently have never been saturated by water or liquid wastes in the vadose zone.

Ms. Kristie Kilgore May 30, 1989 Page 5

It is evident from the site hydrogeologic data surrounding Pit A that vertical migration of liquid wastes from Pit A to the water table can (and has) occurred without significant lateral migration of con­taminants in the vadose zone.

B. MONITOR WELL HS-1:

Abandoned ADHS monitor well HS-1 has had the highest HVOC concentra­tions detected at the site. There have been different hypotheses concerning the source of the HVOCs detected in this well. The report states that reaming operations or sloughed material from the UAD gen­erated during drilling operations which may have been contaminated with hazardous waste could be responsible for the groundwater con­tamination. This is unlikely, however, since no hazardous substances were detected in soil samples obtained from the UAD in soil boring SB-6 which is located between HS-1 and Pit #3. Also, any con­taminants migrating laterally in the vadose zone from Pit #1 should have been detected in SB-6 if contamination of the UAD was present at HS-1. Therefore, it appears unlikely that the groundwater con­tamination present in HS-1 was due to anything but predominantly ver-tically downward migration from a nearby contaminant source which has migrated to HS-1 by way of groundwater flow.

The argument is also made that the contamination encountered at HS-1 is not from an upgradient groundwater source, because, based on the calculated groundwater flow velocity in. Unit A of 220 ft/yr, the groundwater contaminants initially detected in HS-1 in September, 1984 should have been detected in downgradient monitor wells by 1988. Laboratory analysis of four sets of groundwater samples obtained from monitor wells MW-2UA, MW-2UB, MW-4UA, and MW-4UB between April and December of 1988, indicate that no HVOCs were detected in these downgradient wells. Therefore, it is concluded that the contamina­tion at HS-1 must have been a drilling problem and did not result from detection of a contaminant plume in the ground-water. Examina­tion of this claim, however, indicates that groundwater veloc­ities were calculated for geologic units A and B in using the specific yield of the aquifer in the velocity equation instead of the aquifer porosity (effective porosity).

Ms. Kristie Kilgore May 30, 1989 Page 6

The long-term specific yield of an aquifer consisting primarily of fine-grained sediments cannot be accurately estimated by the results of a 72 hour aquifer test. This method may be acceptable for es­timating drawdowns from a pumping well, but it underestimates the correct figure for effective porosity. This is due to the fact that, in fine-grained sediments, the redistribution of water upon source removal can persist at an appreciable rate for many days and have durations of a few months (Hillel, 1971). In addition, it should be remembered that calculations of sediment permeabilities are simply approximations and the permeability calculated for Units A and B could be higher than actual conditions. For example, utilizing a Unit A permeability only half as great as calculated, or 6 ft/day (which still may be too high for such fine-grained sediments), and an effective porosity of 0.2 (which may be too low), the groundwater velocity is about 50 feet per year. Other reasonable values could yield even lower velocities. In any case, these smaller groundwater flow velocities could account for why no HVOC contaminants have been detected as of yet in downgradient monitor wells. Regardless of the source of HVOC contamination in well HS-1, there has been an observa­tion of HVOCs from-MW-IUA which appears to be independent of the groundwater contamination in well HS-1, based on the presence of other monitor wells between wells MW-IUA and HS-1 and knowledge of groundwater flow directions.

Other scenarios explaining the presence of HVOCs in HS-1 might in­clude incorrectly estimating the locations of Pits #1 and #3 and/or the fact that the clay layer, which lies directly above the basalt, appears to become much thinner in the northeast portion of the haz­ardous waste area and in some borings appears to be non-existent. Any liquids percolating from the base of Pit #1 or Pit ?3 could mound over this clay and migrate a short distance until the clay was absent and then percolate through the basalt (basalt was mentioned as being permeable on page 24 of the report), and then directly to the water table. Such thinning of the clay layer would significantly cut down on vadose zone travel times and could explain why HVOCs were encoun­tered at such an early.date (1984) in HS-1.

C. PROPOSED SITE INVESTIGATION ACTIVITIES:

The site investigation activities proposed in Supplemental Task F Work Plan appear to focus on establishing a perimeter outside of which no HVOC groundwater contamination is present and then initiat­ing site cleanup activities. The proposal also places a lot of em­phasis upon the results of the proposed soil sampling beneath each

I- 1 i l

Ms, Kristie Kilgore 0 May 30, 1989

Page 7

pit (a single boring to each pit) as an adequate means of identifying contaminant sources. ADEQ is concerned that in order to effectively lo­cate remedial devices (such as vapor extraction systems and recovery wells) a more thorough knowledge of contaminant sources and the dis­tribution of contaminants within the vadose zone and the underlying aquifer is necessary. Upon review of both reports, it appears that this information is not known and although the proposed site investigation activities (additional soil borings and monitor wells) may add some new insight on contaminant distributions, they are insufficient and often situated in locations which are incapable of identifying con­taminant sources.

C-1 Soil Borings:

The slant soil borings proposed in the Task F Work Plan attempt to identify the occurrence of any liquid contaminant migration beneath Pits #1, #2, #3, and #4. A single boring at a 10-15 degree angle from vertical would be drilled beneath each of the four pits to

A^_.. _.,:'_.:_.:. :_... . . auger refusal ,;:••_.At. an_ estimated depth. of_ 60...feet_to_ auger refusal_ '-:7 (estimated depth to the top of the basalt layer) and drilling at a

maximum angle of 15 degrees, the boring would have moved only 16 feet inside the perimeter of the pit. Much of the boring would only be a few feet within the pit.

Although the pits are not of great areal extent, their dimensions are large enough that a single boring, which is only going to ex­tend a few feet into the pit interior, would not be considered rep-resentatiye of subsurface conditions for the entire pit. It is therefore proposed that multiple slant borings be conducted at each pit, except for Pit #2 which is small enough to warrant only a single boring. Although Pit #1 is reportedly small in size, it was designated as a disposal area for liquid organic wastes and accord­ing to the manifests received several thousand gallons of PCE and TCA. Therefore, it is a likely location to search for HVOCs in the subsurface. Also, because HVOCs have been detected in nearby monitor well HS-1, it is suggested that Pit #1 should have at least three slant borings drilled at various pit perimeter locations. At Pits #3 and #4, it is recommended that at least four slant borings be drilled due to the large size of the pits and the amount of liq­uid waste disposed into each (about 1,3 million gallons).

( I

Ms. Kristie Kilgore May 30, 1989 Page 8

The increased number of borings is believed to be necessary, espe­cially considering that the report claims that a majority of the contaminants have been absorbed within the soil matrix of the vadose zone.

It should be stressed, once again, that due to the very short dis­tance that the slant borings will extend into the pit interior, even increasing the number of soil borings may not be adequate in detecting contaminant sources if percolation of liquids through the base of the pits was localized and not evenly spread over the en­tire pit. In other words, the non-detection of contaminants in the proposed borings' does not exclude the pit as a potential con­taminant source. Perhaps soil gas analysis could be used as a com­plementary technique to determine source locations.

For the slant borings, it is suggested that soil samples should be obtained at five foot intervals until auger refusal, and analyzed for the constituents listed in the Task F Work Plan.

C.2 Monitor Wells:

Proposed monitor wells A through E in the Task F Work Plan, attempt to provide additional groundwater quality data in the vicinity of the two on-site wells which have detected HVOCs, MW-IUA (proposed wells B and C) and HS-1 (proposed wells A, D, and E).

Monitor well A is proposed to be completed only 10 feet from aban­doned well HS-1 and should answer the questions surrounding the original detection of HVOCs at HS-1. In addition, optional monitor wells D and E appear to be properly located and will be useful in identifying any migration of contaminants from the HS-1 area. However, if HVOCs are detected in well A, it would be useful to in­stall an additional monitor well(s) between well A and Pits #1 and r3 to determine the magnitude of groundwater contamination near the potential sources. It should be noted that detection of HVOCs in well A does not imply that well HS-1 is a contaminant source. A common upgradient source, such as Pit #1 or Pit #3 is far more likely.

Well C has been proposed to be located on the border of the land­fill and about 700 feet to the southwest of MW-IUA for use as a downgradient monitor well to MW-IUA. Due to this great distance.

f

Ms. Kristie Kilgore ! ^ May 30, 1989

Page 9

the utility of this well is questionable if its purpose is to determine the downgradient extent of the HVOC contamination detected in MW-IUA, However, if the purpose of well C is to ensure that contaminants have not migrated off-site, the well is properly located, but it should be understood that additional wells located between MW-IUA and well C might be required to delineate the downgradient extent of the HVOC plume detected at MW-IUA.

Proposed monitor well B has been tentatively positioned 280 feet upgradient of MW-IUA to define the source area of the contamination detected at MW-IUA, A likely source for the HVOCs detected at MW­IUA are the Special Pit Areas which have been ignored by the report as a potential contaminant source during the siting of potential monitor well locations. The Special Pits received many types of liquid organic wastes, including over 7,000 gallons of TCE. Well B would be located crossgradient to the the Special Pits Areas, and if anything, would be monitoring groundwater quality

.downgradient of Pit A (which received non-hazardous domestic ;; y;,.-.:-_::-_i'.__•:-:—,_.... waste).—- Well. B-should be. located to monitor the Special- Pits-'•' Areas, along with at least two other monitor wells, one of which

should be positioned as a downgradient monitor well between MW-IUA and MW-2UA on the southwest corner of the 1979 Special Pits Area.

'7) (memos/lndf.has/g)

November 7, 1990

ARIZONA DEPARTMENT OF WATER RESOURCES

Rose Mofford, Governor N. W. Plummer

Director

15 South 15th Avenue Phoenix, Arizona 85007

Thomas Dunkelman (H-7-2) Remedial Project Manager U.S. Environmental Protection Agency 75 Hav/thorne S t r e e t San Francisco, CA 94105

Re: Draft Remedial Investigation Report and Liquid Waste Evaluation; Hassayampa Landfill

Dear Mr. Dunkelman:

The Arizona Department of Water Resources (ADWR) has reviewed the following Hassayampa Landfill documents:

• Draft Remedial Investigation Report for Former Hazardous Waste Disposal Area; dated October 11, 1990

• Liquid Waste Evaluation - Hazardous Waste Area; dated October 9, 1990

ADWR has found the draft Remedial Investigation Report to be a detailed and v/ell v/ritten summary of the results from the investigation conducted at the landfill. Our main concern is that treated groundwater from the selected remedy be put to a beneficial use that is consistent v/ith the water management plans and goals of the Phoenix Active Management Area. Our specific comments on this document are attached for your consideration (Attachment 1).

ADWR does not have any comments on the Liquid Waste Evaluation Report

We appreciate the opportunity to review these documents. If you have any questions, please contact Grant J. Gibson, Project Manager, at (602) 542-1552.

Sincerely,

Bruce S, Davis Chief Water Quality Division

BD/GG/sg

cc: Jackie Maye, ADEQ Mason Bolitho, ADWR

ATTACHMENT 1

ADWR Review Cominents on the Draft "Remedial Investigation Report for Former Hazardous Waste Disposal Area at Hassayampa Landfill, Maricopa County, Arizona"; October 11, 1990

1. Page 12, First Paragraph:

Nitrate ion may not be related to past disposal activities at the Hassayampa Landfill, however, the treatment of nitrate ion must be at least considered in the feasibility study alternatives if groundwater treatment is necessary.

2. Page 48, Basaltic Lava-Flow Unit:

Were any tests perforraed to determine the porosity of the basaltic lava flow? Do the fractures within this unit serve as pathv/ays for contamination migration?

3. Table 40, Preliminary Evaluation of Remedial Action Technologies Hassayampa Landfill RI/FS, Page 2, Treated Groundwater Disposal:

Some of the disposal options outlined in this section (such as discharge to surface v/ater and discharge to the sewage effluent pipeline for the Palo Verde Nuclear Generating Station) are not consistent with the water management plans and goals of the Phoenix Active Management Area,

pr© PRC Envgronmentall Plamiinimg liVilaniagemenit, Dme. Suite 700 120 Howard Street San Francisco, CA 94105 415-543-4880 FAX: 415-543-5480

November 20, 1990

Mr. Tom Dunkelman, H-7-2 U.S. EPA - Region 9 75 Hawthorne Street San Francisco, CA 94105

Contract No. 068-W9-0009

Work Assignment No. 012-C09035

Re: Comments on Liquid Waste Evaluation

Dear Mr. Dunkelman: Based on a preliminary review of the Liquid Waste Evaluation submitted to you on

October 9, 1990 by Conestoga-Rovers and Associates (CRA), PRC has prepared the following general and specific comments. References noted as Heath (1987) are to R.C. Heath's Basic Ground-Water Hydrology, U.S. Geological Survey, Water-Supply Paper 2220, 1987.

GENERAL

In determining the amount of waste available for migration to the water table, CRA estimated the amount of waste expected to evaporate from the pits, seep from the pits, and adsorb to soil in the unsaturated zone.

These phenomena compete simultaneously for the wastes, and should be modeled as occurring simultaneously in an integrated model. However, each phenomenon was treated separately. A more detailed evaluation is needed to determine the effect of separating these phenomena on the result.

In general, the models used are too simplistic and based on unsupported assumptions. Therefore, the results will not be used to support the risk assessment.

SPECIFIC

Page 10. Section 2.4. Paragraph 2

CRA accurately states that "the amount of adsorption onto subsurface soils is a function of soil type and the amount of natural organic carbon in the soil." The organic carbon fraction is probably more important in determining adsorption, however, this parameter is never mentioned again in the text or used in calculation of adsorption.

Mr. Dunkelman November 20, 1990 Page 2

Figure 3

The amount of sand in the shallow UAD unit and in Unit A (between 40% and 60% at 70 feet to 80 feet BLS in well 40UA) should be reflected in the descriptions, similar to the siltstone interbeds described in the upper UAD unit.

Table 3

The heat of vaporization values used are questionable. The values presented for different compounds are not standardized to a reference temperature. Reference temperatures for the heat of vaporization values range from 13°C to 145°C. The reference temperature should correspond to a typical average temperature at the site.

Section 4.2. Seepage Model

It is not clear why so few models were screened for applicability. Each of the three analytical models are very simple and do not consider the alluvial heterogeneities present at the site. If the present site characterization does not support more rigorous modeling activities, modeling should be based upon very health conservative assumptions to fill data gaps.

Page 19. Section 4.2

Although the equation used to calculate the seepage rate is given as:

fp = K(H + L -f S) L

where fp is the seepage rate and K is vertical hydraulic conductivity, the relationship really used to calculate the seepage rate is:

fp = K

This should be made clear in the report.

Page 20. Paragraphs 1 and 2

The text maintains that the inwash of fine materials will seal the bottom of the impoundments preventing further infiltration of contaminants. This inwash effect probably occurred at the site, however, if any of the pits were allowed to dry completely the formation of desiccation mud cracks would significantly enhance infiltration. Therefore the ability of the inwash effect to prevent significant amounts of infiltration is suspect.

Page 20. Paragraph 3

Using a hydraulic conductivity value of 4.5 x 10'* cm/s derived from laboratory tests may underestimate the true value for hydraulic conductivity. Values derived from pump tests indicate that laboratory values are significantly lower than actual values. For example, values provided for hydraulic conductivity of Unit A differ considerably between those obtained in

Mr. Dunkelman November 20, 1990 Page 3

the laboratory and those derived from the pump test. Tests of a sample (25% sand and 80% silt and clay) obtained from between depths 74.5' - 90.5' at well MW5UA resulted in a hydraulic conductivity value of 1.99 x 10* cm/s (Pg. 97, RI), whereas pump tests for MW5UA resulted in a value reported as 35 gpd/ft^ which when converted is 1.66 x 10" cm/s (Page 105, RI). According to Heath (1987), these values represent the low and high end respectively of hydraulic conductivity values for silt. Since there is so much variation between values, it seems that the value reflecting more realistic conditions should be used. The larger value may reflect a sandy component which correlates well with the lithology description. Thus, if laboratory tests underestimate the true hydraulic conductivity, then the calculated seepage rate used in the model could be underestimated by three orders of magnitude. Heath (1987) also indicates that hydraulic conductivity in a basalt lava flow varies greatly depending on the amount of fracturing, but can be as high as 1 cm/s.

Detection of VOCs in HS-1 in 1982 indicates that a higher seepage rate than the 0.15 in/day ( or, metrically, 4.5 x 10* cm/s) calculated by CRA may be more realistic. Also, the heterogeneity of the sediments in the UAD unit may be important in evaluating the variation of hydraulic conductivity in the formation. Coarser-grained channel deposits can provide pathways of enhanced migration. Also, the hydraulic gradient is assumed to be 1.0 without any explanation.

Section 4.3

This section is labeled "Adsorption Model". This model does not simulate adsorption or any other chemical process. It is applicable only to aqueous system. The model consists of:

(saturation porosity - specific yield) - moisture content = potential storage (specific retention)

For an aqueous system this model assumes that the moisture content of the soil is not in equilibrium with a passing wetting front. If a wetting front has passed through these sediments within the last several decades their moisture content may approximate the specific retention. If so, the specific yield of the sediments ranges from 4% to 30% with an arithmetic mean of 16.44% excluding the sample from less than 38 feet in depth. Therefore the specific yield values assumed by the model are low and the potential storage calculation is meaningless. If a specific yield value of 16.44% % is used in the model, 5 of the 10 samples would have negative potential storage. The concept of potential storage may be applicable to surficial sediment subject to desiccation but it is not applicable to sediments at depth.

The model does not simulate contaminant transport or calculate the amount of contaminant storage. There are a number of reasons for this.

• Some organic contaminants will displace adsorbed water off the soil adsorption sites. The model assumes that water will out-compete all the organics for the adsorption sites. Preferential adsorption of some contaminants could increase the amount retained within the unsaturated zone.

Mr. Dunkelman November 20, 1990 Page 4

• The model assumes that the soil will have the same specific retention of contaminants that it does for water. Many organic contaminants, especially solvents, have lower viscosities than water. This means that some contaminants will flow through pore space where water will become entrained due to capillary force. The net results is that the specific retention by the soil of the contaminant will be smaller, by volume, than the specific retention of water.

The model assumes that the contaminants will migrate as a uniform wetting front, however many contaminants will migrate as ganglia. Contaminant migration as ganglia will significantly reduce the amount entrained within the unsaturated zone.

A number of other factors will influence the migration of the contaminants including pooling of contaminants on impermeable lenses, cosolvation effects, and fracture flow effects within the basalt. The amount of uncertainty introduced by the factors described above render the model used unreliable. A more reasonable approach may be to chose several indicator chemicals, model their potential for migration and then generalize the results to other contaminants.

Also, based on Heath (1987) the specific yield of 0.01 used in the model is half the typical value used for clay. Therefore, based on the lithology, use of this value probably leads to an underestimate of the amount of contamination available for migration to the ground water. The higher specific yield value of 0.10 used in Cases #3 and #4 is more appropriate, but may still be an underestimate.

Page 24

If the more volatile compounds were also more dense, the assumption of mixing may also have acted to overestimate the amount of wastes evaporated.

If you have any question in regard to these preliminary review comments, please call me at(415) 543-4880.

Sincerely,

A. Randolph A. Fish Project Manager

RAF/jvg