Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 123
Ward Valley: An Examination of Seven Issues in Earth Sciences and Ecology 4 INFILTRATION AND LATERAL FLOW Issue 2 The Potential Infiltration of the Repository Trenches by Shallow Subsurface (Lateral) Flow THE WILSHIRE GROUP POSITION According to Wilshire et al. (1994) the potential for shallow lateral flow of water downslope into the waste trenches and from the trenches was not addressed in any of the site evaluation documents. Specifically, the Wilshire group suggested that available data show that shallow, low-permeability layers may exist in the alluvial fan slope beneath the site and toward the main valley drainage at Homer Wash. In their view, these could promote lateral flow, leading to excess water leaking into the trenches and migration of contaminants from the trenches to Homer Wash. Once in Homer Wash, these contaminants could be redistributed into the general environment by wind and water erosion much faster than by percolation to the water table. THE DHS/U.S. ECOLOGY POSITION The position of the California Department of Health Services (Brandt, 1994) is that ''there is little, if any, deep percolation of water through the thick, highly moisture-deficient sediments in the vadose zone underlying the Ward Valley site.'' According to DHS, "the infiltrating precipitation is either removed by evapotranspiration or held in storage in the highly moisture-deficient soils near the ground surface until such time as it is removed by evapotranspiration." DHS further stated that their findings showed "the caliche layers at the site to be discontinuous and permeable such that one would not expect them to perch infiltration water in undisturbed areas where they are overlain by sediments. However, where exposed on the surface in the abandoned highway borrow pit, the shallow caliche layers act as a form of discontinuous pavement and the resulting rapid runoff of surface water has led to ponding in the low end of the borrow pit" (Brandt, 1994). THE COMMITTEE'S APPROACH The committee reviewed the data and information pertinent to the issue to evaluate subsurface lateral flow in natural desert soils under natural and enhanced rainfall conditions, ponded conditions, and in engineered systems. Data from Ward Valley and other pertinent
OCR for page 124
Ward Valley: An Examination of Seven Issues in Earth Sciences and Ecology literature, including results of field studies on subsurface flow at sites other than Ward Valley, were used by the committee to clarify the issue. Two basic scenarios are examined: lateral flow under rainfall conditions and lateral flow under ponded conditions. Under rainfall conditions, large areas of soil would be wetted, but the slope of the calcic soil horizon would be an important factor in controlling the extent of lateral flow. Although lateral flow is more likely under ponded conditions because of the higher fluxes, the localized nature of the ponded conditions, the amount of water ponded, and the lithologic continuity of layers are important in determining the extent of lateral flow. LATERAL FLOW UNDER NATURAL AND ENHANCED RAINFALL CONDITIONS AT ARID SITES Lateral flow under natural conditions in add soils depends on many factors, including (1) the lateral continuity of a perching (or low-permeability) horizon, (2) the relative permeabilities of the soil horizons at the site, (3) the magnitude of the rainfall events, (4) the storage capacity of the soil overlying the less permeable layer, and (5) the slope of the less permeable layer. 1. Lateral continuity: The development of a low-permeability soil layer depends on the long-term stability of the geomorphic surface. Roy J. Shlemon's soil report (LA Appendix 2310.A) states that the site was typified "by geomorphic stability throughout late Quaternary time." He states that the age of the surface that is reflected in the soils is at least 35,000 to 40,000 years. This stability would allow the development of a laterally continuous unit. The trench logs are ambiguous. They mention "pervasive" carbonate and show a laterally continuous white carbonate layer (North Wall of WV-TP-1), while the South Wall was not excavated to sufficient depth (>0.6 m) to reach the depth of that horizon. The trench log for WV-TP-2T shows a discontinuous white carbonate-cemented horizon at the same depth, although a slightly deeper horizon (Qf2v) is described as containing pervasive carbonate cement. A still deeper (approximately 2.4 m depth) paleosol is described in Shlemon's soil log as "very hard" and "violently effervescent." Both trenches were located downslope from the proposed waste repository site. Brief field observations of the borrow pit located upslope of the site indicate that the calcrete1 is a continuous horizon. It also appears to be continuous on aerial photographs, at least in the western portion of the pit up to the "bench" east of which it has been excavated. It is possible (more supported than contradicted by available information) that the horizons are laterally continuous. 2. Relative permeability of soil horizons: For ponding to occur, the calcrete layer must be less permeable than the soil above so that it acts as a (leaky) impedance layer which could temporarily perch under high infiltration rates during prolonged rainfall. Two observations in the borrow pit suggest that it is less permeable. First, the fact that the calcrete 1 See Box 2.2 in Chapter 2 of this report for a discussion of soil carbonates.
OCR for page 125
Ward Valley: An Examination of Seven Issues in Earth Sciences and Ecology layer can sustain fractures over seven feet long indicates that it is highly cemented. The second is that evidence of ponding is present at both the western and eastern ends of the borrow pit (Harding Lawson Associates, 1994). At least under heavy rainfall conditions the calcrete layer, when exposed to the surface, is sufficiently impermeable to create temporarily standing water. It is not known, however, for how long bare calcrete horizons can sustain standing water, as calcrete layers can be slightly permeable. Gile (1961) measured infiltration rates with a ring infiltrometer directly on calcrete layers classified as weak, moderate, strong, and very strong. Carbonate contents ranged from less than 1 to 93 percent. Infiltration rates of the calcic horizons ranged from 0.13 cm/hour to 15.0 cm/hour. Lowest infiltration rates were obtained where laminae occur in the uppermost part of the calcic horizon. Further evidence for water movement through a dense calcrete horizon under natural conditions was provided by information on the depth distribution of bomb-pulse 36Cl at the Nevada Test Site (Gifford, 1987). Some of the bomb-pulse tracer was found within the dense calcrete, which indicates water movement into the calcrete during the past 30 years. Experiments at the Las Cruces trench site have also demonstrated that calcrete layers are not impermeable and that significant vertical water flow through calcrete layers is possible (Wierenga et al., 1991). These field experiments showed that only limited lateral water movement occurred under enhanced rainfall conditions in a layered soil system that contained calcic horizons. In addition to the possibility of soil layering and resultant changes in permeability causing lateral flow, another process called tension-dependent anisotropy can also promote lateral flow. Yeh et al. (1985) used a stochastic analysis of the flow equation to demonstrate that the ratio of horizontal to vertical permeability depends on the wetness of the soil profile, increasing as the wetness decreases or as the tension increases. Such an increase in the permeability ratio should promote lateral flow. The Las Cruces experiments were designed to test, under unsaturated-flow conditions, the extent of lateral flow resulting from tension-dependent anisotropy and also from layering. The field experiments, however, showed limited lateral flow. The tests were conducted at the Las Cruces test site in the Jornada Range in southern New Mexico. Three comprehensive tests were conducted on this well-instrumented and characterized site. The results of the first experiment showed water fronts moving mostly downward with time after artificially being rained upon at a rate of 1.8 cm/day for 86 days (Figure 4.1). The area being rained upon was 4 m wide by 9 m long. A trench was dug to 6 m depth perpendicular to the long axis of the trench to observe and measure the water front penetrating the soil. The data in Figure 4.1 show that lateral flow at this site is limited. After 50 days and 91 cm of water application (7 times the total annual rainfall at Ward Valley), the water from had reached the 5.3 m depth. Lateral spreading after 63 days was only 1 m outside the rained-upon area. The extent of lateral water movement was also limited during the second experiment, which used an artificial rainfall rate of 0.5 cm/d. The third experiment again used an artificial rainfall rate of 1.8 cm/d for 86 days, similar to the first experiment. A total of 155 cm (61 in) of water was applied. Even though the soil was highly heterogeneous with several calcrete layers, three buried soil horizons, and a surface slope of 4 percent, lateral spreading was limited to 3 m
OCR for page 126
Ward Valley: An Examination of Seven Issues in Earth Sciences and Ecology Figure 4.1 Advance of the wetting front with time following wetting of a 4 m wide experimental plot at the Las Cruces Trench site. The plot was rained upon for 86 days at a rate of 1.82 cm/day. Solid lines show distribution of water from on trench face from 7 to 63 days after initiation of rain event (Wierenga et al., 1991). beyond the rained-upon area after 525 days (Vinson et al., 1995). Furthermore, saturation was not observed anywhere in the soil profile during the three experiments, and no water was observed to seep from any of the four trench faces. 3. Magnitude of rainfall event: Under normal rainfall conditions, it is unlikely that sufficient water would infiltrate to create saturated flow. Computer simulations with 5 inches of rain falling over a 2-hour period in a soil with similar hydraulic characteristics as the Ward Valley site and under a slope of 2 percent has shown no saturated conditions over the less permeable layer located at 50 cm below the soil surface (Pan and Wierenga, 1995). Only when a 5-inch rainstorm was followed by another 5-inch rain fall the following day was saturation observed for less than 12 hours above the less permeable calcrete layer. Two consecutive 5-inch rainstorms within a 24-hour period are highly unlikely. 4. Storage capacity of the soil: The storage capacity of soils was discussed earlier in Chapter 3 of this report. A surface layer at least 50 cm thick can store enough water to prevent ponding above the less permeable layer. Highly unlikely rainfall conditions such as simulated above, however, could induce subsurface ponding. Only in case of a very thin
OCR for page 127
Ward Valley: An Examination of Seven Issues in Earth Sciences and Ecology surface layer or no surface layer at all, is ponding likely to occur under natural climate conditions. The combined effects of relatively permeable surface soil with a somewhat permeable calcrete layer, dry initial conditions of the surface layer, and the lack of extreme rainfall events (e.g. two 5-inch storms in a 24-hour period) make saturation above the calcrete layer unlikely. 5. Slope of the less permeable layer: Slopes along alluvial fans, such as at Ward Valley, could possibly promote lateral flow. Hillslope lateral flow may be observed in road cuts where water seeps out of soil layers. This is often caused by rainwater percolating into soil upslope, reaching a low-permeability layer, and nearly saturating the soil above the low-permeability layer. If the low-permeability layer has a significant slope (e.g., 10-30 percent), water will move downslope above the layer. This downslope flow is the immediate consequence of the downslope component of gravity. As explained by Philip (1993), "when you have a car in neutral with the safety brake off, it rolls down hill." However, in flat lands, defined by Miyazaki (1993) as land having a slope of less than 3 percent, flow is driven by capillary diffusion, and is predominantly downward (Figure 4.1). At Ward Valley, the surface slopes about 2 percent to the east. Near-surface units are likely to have similar slopes, and sub-soil horizons should closely parallel the exposed surface. Drilling and test pits reveal local heterogeneity which makes it difficult to establish broad patterns of slope in the near-surface sediments. However, observations within about 1 to 2 meters of the surface in many areas of the fan indicate that the overall slope of the calcretes is not much different from the present slope and dose to 2 percent. Such small slopes are too low (˜ 2 percent) for the process of lateral flow in the upper unsaturated zone to be significant. Water will move down through the variously textured layers, rather than laterally. In order to constrain the maximum flow rate that could be expected into the trench by lateral flow, let us assume that a completely impermeable layer is located at some shallow depth below the land surface. Overlying this impermeable layer is a layer of soil with hydraulic conductivity similar to that found at the Ward Valley site. From the infiltration test conducted during the characterization efforts, we can assume that the near surface soils have a saturated conductivity of approximately 1 m/day. We then take this layered soil system to be inclined such that the slope of the impermeable layer is similar to that found at the proposed sate (2 percent) and intersected by a trench wall. If we ignore the effects of capillarity, a simple estimate of the amount of water which can flow along the top of the permeable layer can be calculated from Darcy's law (Equation 3.1) as: where Q is the volumetric flow, Ksat is the saturated conductivity of the soil, A is the cross sectional area of the flow, and dH/dz is the hydraulic gradient. If we assume that the trench face is 100 m wide and that the depth of saturation above the impermeable layer is 5 cm (corresponding to approximately 2.5 cm of precipitation being ponded on the layer), the cross section available for flow is 5 m2. Since the flow is ponded on the layer, and ignoring capillary forces, we can assume that the gradient is equal to the slope
OCR for page 128
Ward Valley: An Examination of Seven Issues in Earth Sciences and Ecology of the layer or 0.02 (2 percent). The resulting volumetric rate into the trench is therefore (1 m/day) (5 m2) (0.02) = 0.1 m3/day or approximately 100 liters per day. The actual rate would probably be much less as capillary forces would tend to hold much of the water within the soil matrix. If such water were to flow into the trench continuously, an additional 36,500 liters of water would be available each year to contact the waste and potentially move downwards towards the water table. Assuming this water were spread over the entire area of the trench floor (approximately 400 m long by 400 m wide or 40,000 m2), the resulting recharge rate through the trench floor would be 36 m3/yr/40,000 m2 or approximately 1 mm/yr. While this represents a considerable percentage increase in the recharge rate over what is proposed in the license application, the rate is still extremely small and is unlikely to affect the overall ability of the unsaturated zone to isolate waste. The above model and discussion represent a simplified model of lateral flow. We have ignored the effects of capillarity and assumed that ponding occurs on a shallow, continuous, and completely impeding layer. Under the more realistic field conditions, capillarity and finite permeability of any impeding layer will significantly reduce the amount of water available for lateral flow. As a result, ponding above a caliche layer over an extended time period and significant lateral flow into the trench is highly unlikely. Therefore, in the committee's opinion, hillslope flow is not significant at the Ward Valley site because the slopes are too low (˜ 2 percent) to cause large amounts of lateral flow. In addition, as discussed in Chapter 3 of this report, water content in the unsaturated zone at the site is low. Under low water-content conditions, the very small downslope gravity component of subsurface flow is negligible compared to the diffusion component. LATERAL FLOW UNDER PONDED CONDITIONS AT ARID SITES Although the natural interstream setting at Ward Valley shows no evidence for ponded conditions, information from ponding experiments provides conservative estimates of the degree of lateral flow if ponding occurs. The infiltration test conducted at Ward Valley consisted of ponding water in a 4.5 m by 4.5 m infiltration pond. The pond was located in an excavation constructed to a depth of 1.8 m below ground surface (Ohland and Lappala, 1990). A total of 97 cm depth of water was added, most of which infiltrated in approximately 25 hours. Thus, while the water at the Las Cruces site was added relatively slowly, water at Ward Valley was ponded and allowed to infiltrate at maximum rate. After 139 days the water front had reached a depth of about 5 m below the surface of the pond. The initial soil-water content at the Ward Valley site was 6 percent by volume. Wetting of the soil, following ponding, was observed 1 m from the pond boundary, but no wetting was observed at 5 m from the edge of the pond. The limited extent of lateral water movement is attributed to the dominant effect of gravity and to variation in hydraulic conductivity with water content. While a large matric potential gradient appears at the margin of the wetted region, the exponential decrease in hydraulic conductivity with a reduction in matric potential results in low lateral fluxes from the wetted region.
OCR for page 129
Ward Valley: An Examination of Seven Issues in Earth Sciences and Ecology A detailed ponding experiment was conducted at Yucca Mountain to evaluate subsurface flow in a layered desert soil system (Guertal et al., 1994). The experiment consisted of ponding 10 cm of water within a 3.5 m diameter ring for 14 days. Water content was monitored with a neutron probe installed in the center of the pond. Neutron probe access tubes were not installed to monitor lateral flow. Three distinct calcrete horizons were identified in the upper 8 m section. Subsurface water movement was restricted at each calcrete horizon. The downward flux was greater than the permeability of the calcrete, and water moved laterally on these low permeability horizons until the permeability of the calcrete could accommodate the downward flux. The data showed that it took only three hours to penetrate the top one meter thick calcrete. Results of this experiment indicate that calcrete will, upon wetting, begin to accept water, thereby limiting lateral flow. The experiment also demonstrates that the calcrete horizons are not impermeable. LATERAL FLOW IN ENGINEERED SYSTEMS AT ARID SITES In many instances trench caps are specifically designed to promote lateral flow. Clay layers or native soil amended with bentonite, as proposed for the Ward Valley B/C trench covers, are used to minimize downward water movement. Clay soil barriers are generally wetted and compacted to achieve permeabilities of approximately 10-7 cm/s. In arid settings, however, this wetted clay layer may become ineffective with time because it will dry and crack, creating potential preferred pathways for flow. This could enhance rather than minimize problems associated with downward water flow or upward radon diffusion because it would concentrate movement in the cracks. Many trench covers, including the proposed B/C trench cover at Ward Valley, are designed with capillary barriers. The capillary barrier consists of coarse-grained sediments overlain by fine-grained sediments (Ross, 1990). The controlling mechanism is the capillary potential of the fine-grained soil, which prevents water from entering the larger pores of the underlying coarse-grained soil. At equilibrium, the matric potential of the two layers will be equal, and if the soils are sufficiently dry, the hydraulic conductivity of the coarse layer will be much less than that of the overlying fine layer, which will cause water to accumulate at the interface of the two layers. In humid sites, steep slopes (20 percent) are used to promote lateral water movement to a drainage collection system. Although the details vary, the concept of capillary barriers has proved valid through field studies and numerical modeling (O'Donnell and Lambert, 1989; Nyhan et al., 1993; IT Corporation, 1994). In add sites, because of low fluxes, capillary barriers can be used to retard subsurface flow and allow greater time for water to evapotranspire. The slopes used at some of these sites are on the order of a few percent (Cadwell et al., 1993). However, by changing the composition of the fine layer over the lower layer, so that the horizontal conductivity is much larger than the vertical conductivity, lateral flow could be enhanced, even for gentle slopes (<5 percent) (Stormore, 1995). Capillary barriers also exist in the natural system. The concept of capillary barriers explains why buried gravel paleochannels, which could constitute zones of preferred lateral
OCR for page 130
Ward Valley: An Examination of Seven Issues in Earth Sciences and Ecology flow under saturated flow conditions, do not act as preferred pathways in unsaturated systems. This is because, in unsaturated systems, the permeability of coarse sediments is less than that of fine sediments and gravel paleochannels retard downward flow, except near saturation. The disposal facility at Ward Valley is designed so that the trench floor is sloped and any water that accumulates will be removed. If water ponded on the trench floor and was not removed, the infiltration experiments described above suggest that this water would move predominantly in a vertical direction. However, if the trench were to intersect a horizontal coarse-grained layer and became ponded, lateral flow would occur along this coarse layer. The extent of such lateral flow would be limited by the horizontal continuity of buried alluvial fan systems that produce these conductive zones. Even if the unit were laterally continuous, the extent of lateral flow would be controlled by the volume of ponded water, and by the hydraulic conductivities of the coarse and underlying fine-grained layer. CONCLUSIONS Low water contents, low water potentials, and high chloride concentrations indicate that subsurface water fluxes are negligible in interstream settings under natural conditions in arid regions. The permeability of calcrete is high enough that such horizons probably do not prevent downward movement of water at these low fluxes. This is supported by chemical tracer data at the Nevada Test Site where bomb-pulse 36Cl is found within a calcrete horizon. Experiments conducted at Las Cruces show that even under conditions of artificially enhanced rainfall (7 times the annual rainfall at Ward Valley) the permeability of calcrete horizons was sufficient to allow water to move predominantly downward and lateral flow was limited. The committee concludes that the slope at the Ward Valley site at ˜ 2 percent is too low to produce significant volumes of lateral flow in the unsaturated zone under natural conditions. The horizontal component of the gravity vector is simply too small on a 2 percent slope to overcome the resistance to flow through soils found at the site and to cause significant lateral flow as compared to downward flow that responds to the much larger vertical component of gravity. Under localized ponding conditions, lateral water movement was limited during the ponding experiment at Ward Valley. This is attributed to the dominance of gravitational flow and to the exponential decrease in hydraulic conductivity laterally from the wetted zone. A ponding experiment conducted at Yucca Mountain showed that calcrete horizons restricted downward flow, although the data showed that in only 3 hours water penetrated the upper 1 m thick calcrete. In summary, the committee concludes that the moisture-deficient nature of the soils at Ward Valley, the absence of an effective slope factor, the very low subsurface water fluxes, and the limited ability of the calcareous horizons to impede vertical flow, eliminate lateral flow as a significant issue at the Ward Valley site. The committee
OCR for page 131
Ward Valley: An Examination of Seven Issues in Earth Sciences and Ecology emphasizes, however, that conditions that could cause local lateral flow, such as ponding and enhanced percolation through the runoff-control structure, should be avoided, particularly in and immediately surrounding the trenches. REFERENCES Brandt, E. C. 1994. Summary submittal of the California Department of Health Service. October 6, 1994. pp. 43. Cadwell, L. L., S. O. Link, and G. W. Gee. 1993. Hanford Site Permanent Isolation Surface Barrier Development Program. Fiscal Year 1992 and 1993 Highlights, Report No. PNL-8741/UC-702, Pacific Northwest Laboratory, Richland, Washington. Gifford, S. K., III. 1987. Use of chloride and chlorine isotopes to characterize recharge at the Nevada Test Site. Masters thesis, University of Arizona. 73 pp. Gile, L. H. 1961. A classification of Ca-horizons in soils of a desert region, Dona Ana County, New Mexico. Soil Science Society of America Proceedings. 25:52-61. Guertal, W. R., A. L. Flint, L. L. Hoffman, and D. B. Hudson. 1994. Characterization of a desert soil sequence at Yucca Mountain, NV. Proceedings, International High Level Nuclear Waste Conference, Las Vegas, NV. May 22-26, 1994. American Nuclear Society, LaGrange Park, IL pp.2756-2765. Harding Lawson Associates. 1994. Summary of Borrow Area Investigation, Low Level Radioactive Waste Disposal Facility, Ward Valley, California. Report to Ina Alterman, NRC. Dated September 30. IT Corporation. 1994. Use of engineered soils and other site modifications for low-level radioactive waste disposal. National Low-Level Waste Management Program. U.S. Dept. of Energy. DOE/LLW-207. License Application. 1989. U.S. Ecology, Inc. Administrative Record, Ward Valley Low-Level Radioactive Waste Disposal Facility, Section 2310.A Miyazaki, T. 1993. Water flow in soils. Marcel Dekker Inc., NY, 296 pp. Nyhan, J. W., G. J. Langhorst, C. E. Martin, J. L. Martinez, and T. G. Schofield. 1993. Field studies of engineered barriers for closure of low-level radioactive waste landfills at Los Alamos, New Mexico, USA. Proc. of the 1993 International Conference on Nuclear Waste Management and Environmental Remediation, Prague, Czechoslovakia. O'Donnell, E. and J. Lambert. 1989. Low-level radioactive waste research program plan . NUREG-1380. Washington, DC. Ohland, G. L. and E. G. Lappala. 1990. Supplemental ground water flow and transfer mechanisms report. HLA Job. #171G0,068.11. Appendix 6151 B. U.S. Ecology, Inc. Pan, L. and P. J. Wierenga. 1995. Unpublished modeling study for this report. Philip, J. R. 1993. Comment on "Hillslope infiltration and lateral downslope unsaturated flow" by C.R. Jackson. Water Resources Research 29:4167. Ross, B. 1990. The diversion capacity of capillary barriers. Water Resources Research. 26:2625-2629.
OCR for page 132
Ward Valley: An Examination of Seven Issues in Earth Sciences and Ecology Stormont, J. C. 1995. The effect of constant anisotropy on capillary barrier performance. Water Resources Research. 31:783-785. Wierenga, P. J., R. G. Hills, and D. B. Hudson. 1991. The Las Cruces Trench Site: Characterization, experimental results, and one-dimensional flow predictions. Water Resources Research 27:2695-2705. Wilshire, H. G., K. A. Howard, and D. M. Miller. 1994. Ward Valley proposed low level radioactive waste site. A report to the National Academy of Sciences . Vinson, J., P. J. Wierenga, R. G. Hills, and M. H. Young. 1995. Flow and transport at the Las Cruces Trench Site: Experiments IIb. University of Arizona, Department of Soil and Water Sciences Contract Report. 216 pp. Yeh, T. C. J., L. W. Gelhar, and A. L. Gutjahr. 1985. Stochastic analysis of unsaturated flow in heterogeneous soils. 1. Statistically isotropic media. Water Resources Research 4:447-456.
Representative terms from entire chapter: