Comments on "Draft Final One-Dimensional Vadose Zone Modeling"
for the LEHR National Superfund Site
Prepared by Weiss Associates
Dated April 9, 1997

April 20, 1997

Julie Roth, Exec. Director
DSCSOC
Route 2, Box 2879
Davis, CA 95616

Dear Julie:

Please find enclosed my comments on the "Draft Final One-Dimensional Vadose Zone Modeling" for the LEHR site prepared by Weiss Associates. As discussed, I find, based on my experience on vadose zone transport of chemical constituents, that the DOE/Weiss Associates modeling approach is highly superficial and most certainly unreliable. DOE/UCD were warned about proceeding with this modeling approach by DTSC and DSCSOC representatives. It was pointed out before it was undertaken that such an approach cannot be used to predict with any degree of reliability that a constituent in LEHR site waste can be left at its current location based on the predicted slow rate of movement. Such models can, for certain constituents, only be used to confirm that exhumation of the waste management unit and soils contaminated by it must occur.

If DOE/UCD wish to persist with this modeling approach it will be necessary that they significantly improve the model to more properly represent what is known about unsaturated transport of chemical constituents. Further, they must acquire the necessary expertise to incorporate the chemistry of the constituents of concern into the modeling effort. To attempt to model the aqueous environmental chemistry of these constituents as a simple distribution coefficient is naive at best.

Also, DOE/UCD will need to conduct a far more intensive, comprehensive field site investigation to provide the necessary data needed for development of the model. DOE/UCD cannot follow the approach of a limited field investigation and then claim that they have adequately characterized the site so that an overly simplistic modeling approach of this type can be used to predict that the transport of a constituent is sufficiently slow as to enable it to be left at site without remediation. Basically, I recommend that DOE/Weiss Associates abandon attempting to reliably model unsaturated transport of LEHR site wastes.

If you, the RPMs or others have questions on these comments please contact me. I will be able to address them after I return from Puerto Rico next week.

Sincerely yours,

Fred

G. Fred Lee, PhD, DEE
Technical Advisor, DSCSOC

GFL:djc
Enclosure

Comments on "Draft Final One-Dimensional Vadose Zone Modeling"
for the LEHR National Superfund Site
Prepared by Weiss Associates
Dated April 9, 1997

Submitted by

G. Fred Lee, PhD, DEE
Technical Advisor, DSCSOC
G. Fred Lee & Associates
El Macero, CA 95618
Ph: (916) 753-9630
Fx: (916) 753-9956

April 20, 1997

Overall Comments

Weiss Associates, while following traditional unsaturated transport modeling approaches, have not incorporated what is known about the real world of vadose zone transport of either water or chemical constituents into their modeling of contaminant transport in the vadose zone that is applicable to properly assessing the transport time from the waste management unit to the groundwater table at the LEHR site. Basically, the Weiss Associates modeling of unsaturated transport of constituents is extremely rudimentary and grossly over-simplifies the real transport that has and will continue to take place at the LEHR site. It has been known for at least 15 years that unsaturated transport of water and constituents does not occur as a uniform front where a simple unsaturated transport model can describe either the rate or the amount of transport. Unsaturated transport from rainfall events occurs as short-term pulses which likely become saturated transport for short periods of time. It is during these pulses that most of the water and constituent transport occurs.

It has also been well-known for many years that the transport of water and contaminants in the vadose zone occurs primarily as fingers of higher degrees of saturation than the average saturation of the vadose zone and not as predicted based on the modeling approach used by Weiss Associates.

Another deficiency with the modeling approach used by Weiss Associates is assuming that the chemistry of the transport of constituents and especially the exchange reactions with solid surfaces that occur in a film of moisture on the surfaces of aquifer solids are the same as those that occur in bulk solutions. The assumption that a Kd developed in bulk solutions applies to a Kd for thin film adsorption-desorption reactions is highly tenuous and likely to be in significant error. The concentrations of constituents in the liquid phase of the film can be significantly different than that predicted based on a overly simplistic Kd developed for situations in which there are large amounts of liquid compared to the solid surface area.

The Weiss Associates' modeling approach is an academic model that likely has little applicability to the real world of the LEHR site. The rate of migration under real world conditions can be, and almost certainly is, greater than the academic model predicts. This means that the Weiss Associates modeling approach is not a reliable modeling approach for determining, as proposed, whether a constituent present in a waste management unit or contaminated soils will move from its current location to the groundwater table. If this modeling approach predicts that transport will occur for a "conservative" constituent, such as nitrate under some conditions, then it is likely that such transport will in fact be found. However, a modeling result prediction that transport will not occur does not mean that this prediction is reliable because of the inherently overly-simplistic approach that is was used in modeling transport through the vadose zone by Weiss Associates.

Specific Comments

Page V of the "Summary" states,

"The modeling was limited to the U.S. Department of Energy areas of concern at the LEHR site and may be used as a decision-support tool for evaluating site characterization data and removal actions, and as input to the human health risk assessment."

That statement is not technically valid. Vadose zone modeling of this type can only be used in the decision making process to confirm that areas where wastes have been deposited as well as contaminated materials associated with the site must be removed because of the threat they represent to groundwater. This type of model cannot be used reliably to predict that the transport of hazardous chemicals will not occur for as long as the wastes represent a threat.

Page V in the "Summary," paragraph four indicates that a three cm/yr average infiltration rate was used. The use of average infiltration rates is technically invalid in predicting rates of transport. The soil moisture at times immediately during and following a rainfall event should be used since this drives the movement of constituents in the vadose zone.

The statement on page VI,

"Based on this and other conservative assumptions, the ground water concentrations calculated by these models should represent an upper limit of potential ground water impact."

is not appropriate. The modeling approach is not conservative with respect to protecting groundwaters. It does not incorporate what is known about how constituents are transported in the vadose zone.

Page V, "Summary," indicates that certain COCs were selected as indicator compounds (ICs). A VOC should also have been selected as a constituent of concern for modeling since the VOCs' behavior in the vadose zone will be different than the chemicals selected. This is justified since VOCs are an important pollutant at the LEHR site.

One of the fundamental problems with the modeling approach is that it is based on very limited field data. Once again, DOE and apparently UCD are attempting to use their limited LEHR site characterization studies which were designed only to provide some exploratory data on some of the characteristics of some areas of the LEHR site as a definitive characterization of the site. An example of this occurs in Figure 8 where Weiss Associates claims that the data for nitrate in Figure 8 proves that a three cm/yr infiltration rate is appropriate. Examination of this figure shows that apparently this assessment is based on one data point. Those familiar with transport of nitrate in vadose zones know that in a climate of the type in the LEHR region, nitrate tends to accumulate in bands during the dry parts of the year. At the first rainfall event, large pulses of nitrate can pass through the vadose zone in saturated transport for short periods of time. At one of the sites that I worked on in western Texas, pulses of nitrate were detected over a one to two day period following a rainfall event that had concentrations at many 10's of mg/L nitrate nitrogen. One of the samples had over 100 mg/L nitrate nitrogen.

Another factor to consider is that nitrate is not necessarily a conservative chemical in the vicinity of waste management units of the type present at the LEHR site. The nitrate is likely being formed from the hydrolysis and oxidation of organic nitrogen-ammonia. Further, nitrate can be denitrified under low oxygen conditions and, based on a paper that I reviewed this past week at the ACS national meeting in San Francisco, can be abiotically chemically reduced to ammonia under certain conditions. I have made repeated recommendations to UCD, DOE and the RPMs that oxygen should be measured in all groundwater samples taken at the LEHR site in order to determine the redox status of the sample. DOE representatives claimed that such measurements were not necessary. The situation now exists where key information pertinent to the transport of nitrate and other constituents in the groundwater system are not available because UCD and DOE did not make the measurements they should have made in their groundwater investigation. This is another example of the short sighted approach used by UCD and DOE in conducting the LEHR site investigations where now important information is not available because an inadequate investigative approach was followed even after the errors in establishing the approach were pointed out. In order to detect the actual behavior of nitrate at the LEHR site a far more comprehensive soil column investigating program would have to be conducted.

Beginning with Figure 9 and for several subsequent figures, Weiss Associates have included a "25.2 mg/L UTL for Nitrate" value. If I understand the origin of this value, it is based on a statistical average background. As discussed in other comments, such an approach is technically invalid. The issue that must be address is what pollution of groundwater has or will occur due to management of wastes at the LEHR site? This issue cannot be reliably addressed by examining the average concentrations of nitrate in soils of the region. Appropriately conducted studies of the LEHR site need to be undertaken in order to establish whether UCDs mismanagement of wastes at a particular location at the site represents an increased pollution of the groundwaters over that which would occur from other causes. DOE and Weiss Associates should stop referring to the "25.2 mg/L UTL for Nitrate" value as having any technical validity for establishing clean up objectives; it does not.

Figure 14 presents so-called breakthrough curves for strontium 90. It appears that Weiss Associates used an overly simplistic approach is describing the chemistry of strontium 90 where a single Kd value was used. The transport of strontium 90 through aquifer systems includes a variety of factors in addition to simple sorption. In hard water systems, strontium will precipitate with carbonates. Its solubility, however, is somewhat greater than calcium so calcium will replace the strontium. Further, strontium should have fairly strong tendencies for ion pair formation which would tend to enhance its transport in groundwater systems. The behavior of strontium at the LEHR site cannot be reliably addressed through the overly simplistic chemical approach used by Weiss Associates.

The same problems occur for radium 226 as presented in Figure 15. The chemistry of radium is influenced by a variety of factors other than simple sorption-desorption which apparently have not been considered by Weiss Associates in their modeling.

Figure 16 presents the alleged chlordane breakthrough curve. The movement of chlordane will be highly dependent on the organic carbon content of the aquifer solids. It is not clear that this was considered by Weiss Associates in developing their breakthrough curve.

Another factor that is just beginning to be understood important to the transport of constituents in saturated and unsaturated aquifer systems is the colloid formation. It is now well understood that colloids are transported through both unsaturated and saturated systems and that materials such as chlordane could be sorbed to the surface of a colloid and be transported with it. Under these conditions the overly simplistic modeling approach used by Weiss Associates would significantly overestimate the real travel time of chlordane.

Figure 18 presents the hexavalent chromium breakthrough curves. The chemistry of chromium is extremely complex and poorly understood. The breakthrough curves presented in Figure 18 are largely academic and have little bearing on what will really occur at the LEHR site.

Table 3 presents the so-called rationale for excluding contaminants as ICs for vadose zone modeling. For most constituents, the rationale is "limited vertical and horizontal distribution." This issue needs to be discussed if there is any attempt to assume that any of the constituents listed have no potential to pollute groundwaters at the LEHR site. Part of this limited vertical and horizontal distribution could be attributed to the inadequate investigation program that has been conducted thus far by DOE and UCD.

Monitoring of Waste Management Units

One of the issues that needs to be clearly understood by UCD and DOE on the transport of constituents from waste management units is the need for ad infinitum detailed monitoring of all waste management units to ensure that any estimates of transport of constituents are, in fact, reliable. Any attempt to leave measurable concentrations of potentially hazardous constituents in waste management units or soils associated with waste management units above a properly evaluated background will require intensive, in perpetuity monitoring to first establish that the estimates that the constituent would not pollute groundwaters for as long as the constituent is a threat is reliable and to ensure the conditions do not change in the future which would mobilize constituents which are predicted, based on the models, to be immobile or slowly mobile constituents.

It will almost certainly be cheaper for UCD and DOE to totally remove the constituents from a waste management unit and contaminated soils than attempt to leave them in place based on the results of the unsaturated modeling that predicts that it will be a long period of time before the constituents reach the groundwaters. This monitoring cannot solely rely on a ring of groundwater wells around the downgradient edge of where the constituents from the waste management unit would be expected to enter the groundwater, but also must include unsaturated transport monitoring using vacuum cup lysimeters in the flow path of the area where the constituents would be expected to move. Since the sphere of capture of a vacuum cup lysimeter is small, there would be need to install a large number of them under each waste management unit at various depths below the unit to collect percolation associated with infiltration events. Such monitoring should be initiated immediately in order to verify that there is not much more rapid transport of constituents below areas where UCD - DOE plan to leave waste-derived constituents.

As I have discussed in previous correspondence, the conventional RCRA or some ill-defined, less-than-RCRA cap for waste management units where UCD - DOE plans to leave waste-derived constituents in the unit or contaminated soils associated with the unit will obviously not prevent moisture from entering the wastes or the soils for as long as the wastes or the waste-derived constituents represent a threat. UCD - DOE must commit to installing, operating and maintaining in perpetuity a leak detectable cover above these units/areas where on-site capping of wastes is proposed based on low predicted transport through the vadose zone. Failure to adopt this approach will not be protective of public health, groundwater resources and the environment.

Return to Document List

View Related Document List

Return to DSCSOC Home Page