An examination of Post World War 2 Concrete Vaulting

CHAPTER I: INTRODUCTION

 

Introduction

A vault is defined as an architectural unit which is in the arched form so that it can offer a space with the slab or the roof. Concrete vaulting has a very old history. The Egyptians are considered to be one of the early civilizations which employed the usage of vaults in their architecture. However, they did not use vaults which were cloister or groined. They employed and the introduced the concept of arch. Mostly, they used it in tombs, storage rooms, drains, etc. During the Middle Kingdom, vaults were used extensively in tombs. However, it should be kept in mind that the vaulting system in the Egyptian kingdom was used as an architectural and structural element. However it was not the dominant and decisive factor to assist in planning the design. The concept of concrete vaulting arose and developed with the Roman Civilization. They developed and introduced vaulting systems in order to fit it into their architecture. Vaults became popular when the Romans introduced concrete as a one of the most important building material. Concrete vaulting systems were produced at accelerated rate and were economically available at large scale. With the development and innovation in the concrete vaulting systems, structures such as domes were developed. Research suggests that concrete vaulting system introduced and developed by Romans had an important place. In recent times, Arches and Vaults have become important in the new and modern architecture.

The growing amount of radioactive and hazardous wastes generated throughout the world is a concern for a number of reasons. For example, these wastes adversely affect human health and remain toxic to the environment for long periods of time. The wastes need to be treated or isolated in disposal systems both economically and safely.

One disposal option that meets these criteria is waste storage in below ground concrete vaulting systems. The key to long-term storage is immobilization of the waste. Since water entering the concrete vaulting system has the potential of dissolving and then transporting the waste out of the vault and into the environment, it is essential to minimize the amount of water passing through the vault. This amount of water is the hydraulic performance measurement for the concrete vaulting systems.

“Dissertation Master provides top-quality dissertation writing services. If you are having trouble writing a dissertation and are looking for a website to get cheap and top-quality dissertation help, then feel free to contact us.”

These concrete vaulting systems are beneficial to waste isolation because their low hydraulic permeability allows a very low amount of water into the vault. However, the vault design is an important consideration that affects the ability of the vault to isolate the waste and protect the environment. Factors related to enhancing the performance of concrete vaulting systems through design are investigated in this paper. Design factors of below ground concrete vaulting system in the unsaturated zone are evaluated by examining the hydraulic performance of the concrete vault. The hydraulic performance is determined by estimating the flow rate of water passing through the vault floor. A study by Nichols (1996)^[1] examined design factors of sloping sand and gravel capillary barriers on the amount of water passing through the underlying concrete vault.

Through evaluation of these design parameters, significant factors that determine the hydraulic performance of the concrete vault are identified not only for the concrete vault as designed (i.e., concrete is intact), but also for the vault during its service life (i.e., concrete is degraded). The latter is important because the vault is expected to meet performance requirements not only initially, but also throughout its lifetime. The relationship between the hydraulic performance and design is complicated because the concrete vault is situated in the unsaturated zone. In the unsaturated zone, flow is more complex than in the saturated zone and is subject to scale and gravity effects. The unique aspects of unsaturated flow have significant implications for concrete vaulting design and long-term hydraulic performance.

Additionally, the unsaturated zone can be an aggressive environment for concrete degradation and lead to leaching.^[2] In the unsaturated zone, concrete can be exposed to soil gases (oxygen and carbon dioxide). There is a potential for buildup of ion concentrations, such as sulfate and chloride, because evaporation can be enhanced by low permeability zones in the cover (ACI, 1984; ACI, 1990; Clifton and Knab, 1989^[3]; Walton et al., 1990).^[4] These conditions contribute to concrete vault degradation and consequently impact the hydraulic performance of the concrete vault.

 

 

Statement of the problem

Growing interest in the preservation of architectural heritage has created a need for examining the historical buildings architecture, especially, the use of concrete vaulting in order to see the significance of the architectural designs and the reasons for decline of the concrete vaulting structure in our times. Although concrete vaulting architectural methods were originally used in Roman period, the importance of the concrete vaulting in present times have grown due to many factors: the natural calamities like earthquakes require the building engineers to design such building architecture capable of sustaining any threat of destruction. Similarly, the radioactive proliferation has also threatened and alarmed the building architects to look for the designs that may improve the safety of buildings.

In the context of the growing rainfalls and other natural calamities like tsunami, below ground concrete vaults have been used and will continue to be used to store wastes over long periods of time. These concrete vaults are beneficial to waste isolation because their low hydraulic permeability allows a very low amount of water into the vault. However, the vault design is an important consideration that affects the ability of the vault to isolate the waste and protect the environment.

Design factors of below ground concrete vaults in the unsaturated zone are evaluated by examining the hydraulic performance of the concrete vault. The hydraulic performance is determined by estimating the flow rate of water passing through the vault floor. A study by Nichols (1996) examined design factors of sloping sand and gravel capillary barriers on the amount of water passing through the underlying concrete vault. The design factors evaluated are soil types and properties of those soil layers surrounding the vault, the vault size, slope of the vault roof, and the amount of concrete degradation under various infiltration rates.

The relationship between the hydraulic performance and design is complicated because the concrete vault is situated in the unsaturated zone. In the unsaturated zone, flow is more complex than in the saturated zone and is subject to scale and gravity effects. The unique aspects of unsaturated flow have significant implications for concrete vault design and long-term hydraulic performance In previous work, a number of researchers studied flow and transport through engineered covers (e.g., Booth and Price,^[5] 1989; Brun et al., 1994; Weeks et al., 1992).

However, less work has been conducted to evaluate factors affecting flow through below ground concrete vaults. Prior work examining flow through concrete vaults (e.g., Walton, 1991; Walton and Seitz, 1992) indicated the tendency for perched water to form on the roofs of large low permeability vaults and thus, yield an increased amount of water flowing through the vault. Because of the effects of gravity on the perched water, the hydraulic gradient through the vault is greater than unity. This contrasts with typical hydraulic gradients of 0.01 or less in the saturated zone.

Additionally, the unsaturated zone can be an aggressive environment for concrete degradation and lead to leaching (van der Sloot 2002). In the unsaturated zone, concrete can be exposed to soil gases (oxygen and carbon dioxide). There is a potential for buildup of ion concentrations, such as sulfate and chloride, because evaporation can be enhanced by low permeability zones in the cover (ACI, 1984; ACI, 1990; Clifton and Knab, 1989; Walton et al., 1990). These conditions contribute to concrete vault degradation and consequently impact the hydraulic performance of the concrete vault.

 

Purpose of the study

The main purpose of this study is to evaluate the use of concrete vaulting architecture design in post World War-II era. For this purpose, a systematic review of the existing studies on concrete vaulting will be made using diverse sources of information.

 

 

 

CHAPTER II: REVIEW OF THE LITERATURE

Deeply embedded and underground tanks storing high-level radioactive wastes are critical components in nuclear facilities, and it is important that they be designed to withstand safely the earthquakes to which they may be subjected. Failure of these systems, which may typically contain up to a million gallons of waste, may have catastrophic consequences, and their safety is of major concern at the present time. There are two types of underground tanks currently in use: the single-shell tank which is composed of a steel liner adjoining a concrete vault; and the double-shell system which is composed of a steel tank embedded in a nearly rigid concrete vault. The tank in the latter case may be either free-standing or constrained by the surrounding vault at the top.

Current understanding of the seismic response of liquid—containing upright cylindrical tanks is mainly derived from analyses of ground-supported, base-excited cantilever systems that contain homogeneous liquids. For a detailed account of this knowledge, the reader is referred to the state-of-the art report by Veletsos and to the references of more recent publications by Haroun et al, Lau and Zeng Veletsos et al and Malhotra et al. The rational analysis of the underground tanks that are of current interest requires the solution of a number of special problems :

1. Being embedded in the ground, the effects of soil-structure interaction may be important and must be considered. For the single-shell tanks, the interaction effects of interest are those of tl1e liquid-tank system with the surrounding soil, whereas for the double-shell systems, provision must also be made for the interaction of the concrete vault with the enclosed steel tank.
2. Many of the steel tanks in the double-shell systems are connected at the top to the surrounding concrete vault. The top constraint may affect significantly the magnitude and distribution of the resulting hydrodynamic forces in the tank and must be considered.
3. The contents of many of these tanks cannot adequately be modeled as homogeneous liquids but must be approximated as layered liquids with varying layer thicknesses and mass densities or as inhomogeneous liquids with continuously varying mass densities. The effects of this density stratification on the hydrodynamic effects may be important and need to be assessed.
4. The base plate of the tank-liquid system is usually assumed to be rigid. This may not be true in many cases, particularly when the base-plate to soil stiffness is small and the distribution of the superposed tank and liquid inertias is non-uniform. It is necessary to evaluate the adequacy of the rigid-base assumption and assess the effects of base – flexibility on the induced hydrodynamic response.
5. Surface-waves, with displacement amplitudes larger than the available free- board, will impact the tank-roof. The impact pressures and forces that are then transmitted to the side-wall of the tank can be significant and need to be evaluated.

The embedded tank-liquid systems of current interest have also been the subject of parallel studies at Rice University. Some of the issues related to item 1 have been addressed by these studies. Many of the remaining issues are addressed in this dissertation.

Effects of Tank-Base Flexibility

The foundation is modeled as a flexible circular plate supported on a soil represented by Winkler springs. The effects of the tank wall, roof and participating soil inertias are represented by a mass distributed along the periphery of the plate. Different conditions of rotational constraint are considered at the junction of the wall and plate, including the limiting cases of hinged and fixed edge-supports. The response quantities of interest include the natural frequencies of the system, the associated mode shapes, the modal damping ratios, and the hydrodynamic pressures that are induced on the tank-wall and base.

Effects of Roof-liquid Impact

The kinematics of a design impact wave are defined such that the wave can be presumed to produce the maximum impact effects during the time history of the liquid sloshing. The magnitude and temporal variation of the roof impact force due to the design wave are then derived by using an impulse-momentum relation. The added hydrodynamic mass of the impacted length of the roof plate is used in this computation. The effects are evaluated for both flat and sloping roofs and the validity of the results is established by comparing them with limited experimental/empirical data available in the literature. Procedures are then formulated for evaluating the impact pressures and forces transmitted to the tank wall. Comprehensive numerical solutions are finally presented for the roof and wall impact effects and their importance is ascertained by comparing them with previously established impulsive and convective effects.

Many tanks used to store liquid radioactive wastes in nuclear facilities are underground and are embedded in vaults. Furthermore, they are attached to the vaults in such a manner that during an earthquake, they are excited simultaneously at the top and base. Fundamental to the analysis and design of the inner tank of these double shell systems is a thorough understanding of the hydrodynamic effects in constrained systems. The objectives of this chapter are:

• To highlight the nature of the hydrodynamic effects induced in tanks that are supported at the top and experience the same input motion at that level as at the base; and
• To establish the interrelationship of the response of the top-constrained systems to the well-established response of free-standing, cantilever systems.

The governing expressions are presented in a form similar to those for the cantilever systems, and the effects of the top-constraint on the associated response coefficients are assessed by suitably using/ modifying existing methods and programs for cantilever systems.

The response quantities of interest include the natural frequencies of the vibrating tank-liquid system, the impulsive pressures exerted against the tank wall and base, the shears at the top and the base, the base moment and the foundation moment. Both impulsive and convective components of the induced tank forces are considered. Selected numerical data are presented in order to elucidate the action of the top constraint, and the effects and relative importance of the numerous parameters that influence the response.

 

 

 

CHAPTER III: RESEARCH METHOD

Methodology

There are a number of concrete vault design factors that may influence the amount of water passing through a vault. As stated previously, several design factors have been identified including the soil types and properties of those soil layers surrounding the vault, vault size, slope of the vault roof, and the amount of concrete degradation (i.e., either intact or degraded) at various infiltration rates. The infiltration rate can also be considered a design parameter since a decision needs to be made regarding selecting the location for the disposal site and whether the precipitation and potential infiltration conditions are acceptable. These factors are evaluated using the conceptual model and input parameters described in this section.

Conceptual Model

A schematic representation of the concrete vault conceptual model is provided in Figure 1. The vault is conceptualized as having a constant infiltration rate. This condition could arise as a result of an engineered surface cover above the concrete or because of assumptions regarding infiltration sources such as precipitation or flooding events. For example, a low infiltration rate can represent either a site that is relatively arid or a cover that exhibits a high performance at a humid site.

Figure 1: Schematic representation of the concrete vaulting system and the model domain

The dimensions of the nominal concrete vault are 10 m high, 20 m wide, and infinitely long (i.e., a two-dimensional simulation). In the simulations, vault symmetry is assumed and therefore the modeled vault is 10 m wide. The nominal concrete vault has a roof slope of 0 degrees. The concrete vault is located in the unsaturated zone and the water table is 4 m below the vault floor. Because of symmetry, the concrete vault is assumed to exhibit a no flux plane at the vault center. Figure 2 illustrates the modeled dimensions and boundary conditions of the nominal concrete vault.  Four different types of cover designs for materials surrounding the concrete vault are modeled. The covers consist of either one or two soil type layers as illustrated in Figure 2. The nomenclature utilized to designate the four covers lists the soil layers beginning from the concrete and moves outward. The designations of the four covers are clay-sand-loam, clay-loam, sand-loam, and loam. Each design assumes that the vault is buried in a loam soil (i.e., loam is backfill). The four designs are:

1. Clay-Sand-Loam: a 50 cm layer of clay and a 50 cm layer of sand surround the concrete vault.
2. Clay-Loam: a 50 cm layer of clay surrounds the concrete vault.
3. Sand-Loam: a 50 cm layer of sand surrounds the concrete vault
4. Loam: (default case) only loam surrounds the vault.

The concrete vault is assumed to be either intact with a concrete low permeability or degraded with a much higher concrete permeability.

 

 

Input Parameters

The hydraulic properties for the soil layers surrounding the concrete vaulting system/vault and for the intact and degraded concrete are provided in Table 1 (Carsel and Parrish 1988). The degraded concrete vault is assumed to have the unsaturated flow properties of coarse sand. This condition is equivalent to assuming fracture flow. In fractures, most of the flow resistance is from entrance and exit head losses. This assumption is acceptable because the effective permeability of fractured concrete is generally a function of the hydraulic conductivity of surrounding porous materials (Walton and Seitz, 1992). Permeability of layers adjacent to the concrete has been modified to reflect this condition. That is, the Table 1 permeability values reflect conditions of the surrounding porous materials with the top or bottom cover soils of the vault having a lower permeability than on the sides of the vault.

In addition to assumptions regarding two of the design parameters (i.e., whether the concrete is degraded or intact and the composition of the cover soils adjacent to the concrete vault), assumptions for other design parameters were made. These design parameters include the vault width (i.e., horizontal scale), roof slope, and infiltration rates. The horizontal scale of the concrete vault is assessed with seven values for the vault width. To take advantage of symmetry and reduce the computional requirements, vault half-widths of 2.5, 5, 10, 15, 20, 30, and 40 m are employed in model simulations, and hereafter, vault half-widths are used to refer to the horizontal scale. Four roof slopes are evaluated. These roof slopes are 0, 3, 6, and 10 degrees. Also, five different infiltration rates of 0.01, 0.1, 1, 10, and 100 cm/yr are employed to evaluate the vault performance through calculating the seepage rate of water through the concrete vault.  These five infiltration rates are indicative of both the effectiveness of the cover, if one is present, and the environmental conditions.

The ranges of the values specified for the concrete vaulting system design parameters (i.e., degree of concrete degradation, soils types and properties surrounding the vault, vault scale, vault roof slope, and infiltration rates) are considered to be representative of lower and upper limits. Consequently, seepage rates of water through the vault computed over these ranges of design parameters are considered bounding for evaluating the performance of the concrete vaulting system/vault.

 

Model Assumptions

Concrete cracks and degrades over time. As a result of cracking, a concrete vault or concrete canister containing waste becomes more permeable to water. Because of the concrete degradation, the hydraulic properties of the concrete vault will be highly variable and uncertain over time.

The assumptions in this analysis are formulated recognizing the variability and  uncertainty inherent in concrete vault disposal systems. Consequently, these assumptions rely on simplifications for the properties of intact and degraded concrete. The challenge is to design a vault that has acceptable performance not only with intact materials and but also with the degraded materials as those materials evolve over time.

All simulations represent flow through the concrete vault under steady-state conditions. That is, the infiltration rates and the hydraulic properties are assumed to remain constant in a simulation.

Governing Equations

There are a number of approaches are available that could be applied to model flow through a concrete vault (e.g., Kacimov and Obnosov 2000; Warrick et al 1997;  Zang et al. 2002). A numerical computer code that models steady-state flow is implemented to evaluate the concrete vault design factors. The code has been previously described in Walton (1991), validated, and used to model unsaturated flow (Shahjahan 1995; Ahmed 1995). The numerical method employs integrated finite difference for variably-saturated, steady-state flow. Van Genuchten curves combined with the Mualem integral are used to relate the hydraulic conductivity to negative pressure head and saturation (Jury et al., 1992).

Model Discretization

The discretization of the model domain in Figure 2 is provided in Figure 3. This discretization consists of variably spaced finite-difference grids with closer node spacing near material boundaries. The material boundaries are located at the vault-inner cover boundary, the outer cover-backfill boundary, and inner cover-outer cover boundary, if present. The boundaries are illustrated in Figures 1 and 2.

Nodal spacing is variable with closer node spacing used near high contrast boundaries to improve accuracy. The code automatically calculates the mass balance of water at each horizontal plane, which must be equal at steady state for the imposed boundary conditions. The solutions are considered to be converged after the mass balance error is less than 15%.

 

 

 

CHAPTER 3: RESULTS AND DISCUSSION

 

Simulations are conducted for the four different types of cover designs described in the previous section and the design parameters listed in section after that. The design parameters are for the vault half-width, roof vault slope, infiltration rate with intact and degraded concrete. Convergence is acceptable in simulations with mass balance errors less than 15%.

The simulation results are displayed as plots and contours in Figures 4 through 14. Tables 2 and 3 provides simulation results of seepage rates with degraded and  intact concrete for a 10 m wide concrete vault using the modeled vault roof slopes,  infiltration rates, and four cover materials. To illustrate the effects of vault scale, Table 4 presents seepage rates with degraded and intact concrete for the Clay-Loam cover layer and zero vault roof slopes at the modeled infiltration rates and vault half-widths. For these plotted results in Figures 2-4 through 2-15, the following sections provide a discussion of these results for intact and degraded concrete.

 

Figures 4 through 7 Showing Effects of Slope on Seepage

These four figures illustrate the trend in seepage rates through the concrete vault at different roof slopes for intact and degraded concrete. All four cover layers are evaluated. The vault half-width is 10 m in these simulations. Figures 4 and 6 correspond with each other for degraded concrete. The only difference in these figures is the former is a 3-dimensional plot and the latter displays results for constant vault half- width. Similarly, Figures 5 and 7 correspond with each other for intact concrete in the same manner.

For degraded concrete, both the Loam and Sand cover layers behave similarly and at infiltration rates equal to 0.1 cm/yr and greater allow all infiltrating water to pass through the vault. The Loam and Sand cover layers perform hydraulically the worst for degraded concrete and allow the highest seepage rate through the vault. In contrast, both the Loam-Clay and Clay-Sand-Loam cover layers behave similarly and exhibit the best hydraulic performance for degraded concrete, diverting infiltrating water away from the vault for all roof slopes. Moreover, in a degraded vault and for a given cover layer, these results suggest there are only small differences in seepage rates with varying roof slope for all infiltration rates. That is, the concrete vault roof slope for a given cover layer is not a sensitive design parameter with the degraded concrete vault since it has relatively small influence on hydraulic performance.

For intact concrete, seepage rates are low and all four cover layers have similar hydraulic performance. With all four cover layers, seepage through the vault generally reaches a maximum value at infiltration rates of 0.1 cm/yr and greater, except for Loam  and Sand cover layers with roof slopes of 6 and 10 which reach a maximum value at  infiltration rates of 1 cm/yr and greater. These maximum seepage values are attributable to the formation of perched water on the vault roof and near-saturation conditions existing inside the intact concrete vault. Additionally, in an intact vault and for a given  cover layer, these results suggest there are only small differences in seepage rates with  varying roof slope at infiltration rates greater than 0.1 cm/yr. At an infiltration rate of   0.21 cm/yr for Loam and Sand surrounding an intact vault, the seepage rate does increase with roof slope by about 4 orders of magnitude as the roof slope decreases from 10 to 0 degrees.

 

These results suggest a difference in the performance cover layers over time that is independent of roof slope. When the concrete is intact, all four cover layers have equivalent hydraulic performance. However, concrete degrades over time. Once the concrete degrades, there is a significant difference in the hydraulic performance of the cover layers. The Clay-Loam and Clay-Sand-Loam cover layers are preferred over the Loam and Sand cover layer.

Figures 8 and 9 Showing Effects of Vault Scale on Seepage

The simulations examine whether there is a scale effect associated with the hydraulic performance of a concrete vault. The scale effect involves changes in the capability of the concrete vault cover layers to divert water laterally as the vault half- width increases and thus prevent the seepage of water through the vault. Figures 8 and 9 illustrate the trend in seepage rates through the concrete vault at varying vault half- widths for degraded and intact concrete with the Clay-Loam cover. The Clay-Loam  cover layer was selected because this cover layer together with the Clay-Sand-Loam  cover layer have the best hydraulic performance when compared to Loam and Sand cover  layers as shown in Figures 4 through 7. The vault roof slope is 0 degrees in these simulations.

Comparing results in Figures 8 and 9 reveals that there is a scale effect for both degraded and intact concrete at infiltration rates less than 10 cm/yr. At infiltration rates of 10 cm/yr or greater for both degraded and intact concrete, water is perched on the vault roof and vault half-width does not affect seepage. However, with decreasing infiltration rates from 10 to 0.01 cm/yr, there is more of a scale effect. That is, over the range of vault half-widths (i.e., from 2.5 to 40 m), the differences in the seepage rates become greater as the infiltration rate decreases.   The scale effect is greater for intact concrete than for degraded concrete, because there is greater variability in seepage rates over the range of vault half-widths with intact concrete than with degraded concrete. These results suggest that although present for both degraded and intact concrete, the scale effect is greater initially and decreases as the concrete degrades.

Figures 10 and 11 Showing Effects of Infiltration Rate on Pressure Head

These two figures show the effects of increasing infiltration on seepage through and around the vault for degraded and intact concrete. The darkening blue contours represent increasing pressure heads which corresponds to increasing saturation. Both figures model a vault half-width of 10 m and a roof slope of 0 degrees for a Clay-Loam cover layer. Figure 10 corresponds with results in Figure 7 for degraded concrete at a  10 m vault half-width. Similarly, Figure 11 corresponds with results in Figure 8 for intact concrete at a 10 m vault half-width. In both Figures 10 and 11, as expected, the degree of saturation in the model domain increases with an increasing infiltration rate.  The results shown in Figure 10 indicate perched water exists on the vault roof at infiltration rates of 1 cm/yr and greater. The occurrence of perched water that is evident in this figure is also indicated in Figure 8 when the plot of the seepage rate at a vault half-width of 10 m flattens out at the 1 cm/yr infiltration rate. Once there is perched water, the seepage rate reaches a maximum and hydraulic performance is unchanged even though infiltration increases. Figure 11 for intact concrete displays a similar behavior except perched water occurs at infiltration rates of 0.1 cm/yr and greater. Again, the occurrence of perched water that is evident in this figure is also indicated in Figure 9 when the plot of the seepage rate at a vault half-width of 10 m flattens out at the 0.1 cm/yr infiltration rate.

Comparing Figures 10 and 11 also shows that the intact concrete vault has a higher saturation than the degraded concrete vault. That is, although the intact vault allows less seepage and has a better hydraulic performance than a degraded vault, the intact vault holds more water than a degraded vault. The increased amount of water will increase the concrete degradation, including rebar corrosion. That is, paradoxically, the low permeability of intact concrete leads to a greater degradation rate when compared to degraded concrete.

Figures 12 and 13 Showing Effects of Vault Half-Width on Pressure Head

These two figures show scale effects (i.e., change in the vault half-width) for the seepage through and around the vault with degraded and intact concrete. The figures are similar to Figures10 and11. The darkening blue contours represent increasing pressure heads which corresponds to increasing saturation. Both figures model an infiltration rate of 1 cm/yr and a vault roof slope of 0 degrees for a Clay-Loam cover layer. Figure 12 corresponds with results in Figure 8 for an infiltration rate of 1 cm/yr with degraded concrete and 2.5, 5, 10, and 20 m vault half-widths. Similarly, Figure 13 corresponds with results in Figure 9 for an infiltration rate of 1 cm/yr with intact concrete and 2.5, 5, 10, and 20 m vault half-widths. In both Figures 12 and 13, as expected, the degree of saturation in the model domain increases with an increasing vault half-width because the capability to divert flow decreases with increasing vault half-width.   The results shown in Figure 12 indicate perched water exists on the vault roof at  vault half-widths of 10 m and greater. The occurrence of perched water that is evident in  this figure is also indicated in Figure 8 when the plot of the seepage rate at an  infiltration rate of 1 cm/yr flattens out for vault half-widths of 10 m and greater. Once there is perched water, the seepage rate reaches a maximum and hydraulic performance is unchanged even though infiltration increases. At vault half-widths less than 10 m, the seepage rates have not reached a maximum at an infiltration rate of 1 cm/yr and there is no perched water above the vault roof.   Figure 13 is for a 0.1 cm/yr infiltration rate and intact concrete. The plotted results display a similar behavior as Figure 12 except perched water exists on the vault  roof at vault half-widths of 20 m and greater. The occurrence of perched water that is evident in this figure is also indicated in Figure 9 when the plot of the seepage rate at a vault half-width of 20 m flattens out at a 0.1 cm/yr infiltration rate. Seepage for vaults  with half-widths less than 20 m at a 0.1 cm/yr infiltration rate as shown in Figure 9  have not reached a maximum value and thus there is no perched water (though at 10 m,  the seepage is near the maximum value exhibits perched water in Figure 13).   Comparing Figures 12 and 13, just as with Figures 10 and 11, shows that the intact concrete vault has a higher saturation than the degraded concrete vault. Again, although the intact vault allows less seepage with decreasing scale and has a better hydraulic performance than a degraded vault, the intact vault holds more water than a degraded vault and the increased amount of water increases the degradation of the intact concrete.

Figures 14 and 15 Showing Effects of Vault Roof Slope on Pressure Head

These two figures show the effects of the roof slope on the seepage through and around the vault with degraded and intact concrete. The figures are similar to previously presented figures. The darkening blue contours represent increasing pressure heads which corresponds to increasing in saturation. Both figures model an infiltration rate of 1 cm/yr and a vault half-width of 10 m, and simulate roof slopes of 0, 3, 6, and 10 degrees.  Figures 14 and 15 represent Clay-Sand-Loam and Sand cover layers, respectively.  Figure 14 corresponds with results in Figure 6 for degraded concrete, a Clay-Sand- Loam cover layer at an infiltration rate of 1 cm/yr, and vault roof slopes of 0, 3, 6, and 10 degrees. Similarly, Figure 15 corresponds with results in Figure 7 for intact concrete,  a Sand cover layer at an infiltration rate of 1 cm/yr, and vault roof slopes of 0, 3, 6, and  10 degrees. In both Figures 14 and 15, as expected, the degree of saturation in the model domain increases with an increasing vault roof slope because the capability to divert flow increases with an increase in the vault roof slope.

The results shown in Figure 14 indicate perched water exists on the roof of a degraded vault at all roof slopes displayed in the figure. The occurrence of perched water that is evident in this figure is also indicated in Figure 6 since the plot of seepage rate flattens out for all modeled vault roof slopes at an infiltration rate of 1 cm/yr.

Figure 15 is for a Sand cover layer with a 0.1 cm/yr infiltration rate and intact concrete. The plotted results display a similar behavior as previous figures except perched water is evident on the vault roof at roof slopes of 0 and 3 degrees. This perched water behavior is similar to that discussed in previous sections. Both Figures 14 and 15 display a slight increase in saturation (or ponding) at the right hand boundary. This increase is an artifact in the results caused by the no flux boundary condition. Note that this boundary effect has little effect on the results for pressure heads and flow of water through the vault.

 

Implications for Concrete Vault Design

The results displayed in Figures 4 through 15 have implications in the planning and design of below ground concrete vaults. These implications are oriented toward designing below ground concrete vaulting systems with the best hydraulic performance as measured by exhibiting the lowest seepage rate of water through the concrete vault floor. Decreased seepage leads to lower releases from the disposal and consequently a lower risk of exposure to humans. The results indicate roof slope is not an important design consideration; however scale effects (i.e., varying vault half-widths) do affect hydraulic performance.

An alternative concrete vault design suggested by these simulation results is “double containment.” In this configuration, the outer vault would be similar to a large storage area that would be filled with containers holding the waste. A larger-scale concrete vault having a clay outer cover layer (minimally) could be utilized as the outer containment. The outer containment would initially reduce the flow of water through the vault. However, as discussed previously, the outer layer would degrade. Water passing through this degraded outer concrete vault would be conditioned (e.g., the pH would increase), thus protecting the inner containment that holds the waste from concrete degradation. Furthermore, as indicated by the results presented in this paper, the smaller scale inner containment would better divert flow towards the side of that inner containment.

Another design consideration is suggested by these simulation results.  Engineered covers are designed for placement near the ground surface and away from the below ground concrete vault (DOE 2000). Because of the location, these ground surface covers are susceptible to failure from activities related to plant growth, animals, and humans. Failure at the ground surface could redirect infiltration toward the below ground concrete vault. It is recommended that the cover by moved away from the surface and instead be placed adjacent to the concrete vault. The simulations results show the positive effects on hydraulic performance of using clay as a cover layer for the concrete vault and the negative effects on hydraulic performance of using loam and sand (i.e., backfill) as a cover layer for the concrete vault.

 

 

CHAPTER V: Conclusions

This chapter presents the conclusions reached from evaluating simulation results and the effects of model parameters on the hydraulic performance of the vault. A discussion of the conclusions is provided followed by a summarized list of those conclusions.   The roof slope for each of the cover layers is not a sensitive design parameter with degraded and intact concrete vaults since roof slope has a relatively small influence on hydraulic performance. For degraded concrete at infiltration rates equal to 0.1 cm/yr  and greater, both the Loam and Sand cover layers allow all infiltrating water to pass  through the vault at all roof slopes. Both the Loam-Clay and Clay-Sand-Loam cover layers exhibit the best hydraulic performance for degraded concrete and divert infiltrating water away from the vault for all roof slopes. For intact concrete, seepage rates are low  and all four cover layers and perched water forms above the vault roof at infiltration of   0.1 cm/yr and greater regardless of roof slope.

There is a significant scale effect for both degraded and intact concrete at lower infiltration rates. At higher infiltration rates water is perched on the vault roof and vault half-width does not affect the seepage rate. The magnitude of the scale effect increases with decreasing infiltration rate. The scale effect is greater for intact concrete than for degraded concrete, because there is greater variability in seepage rates over the range of vault half-widths with intact concrete than with degraded concrete. These results suggest that although present for both degraded and intact concrete, the scale effect is greater initially and decreases as the concrete degrades.

For both intact and degraded concrete, perched water exists on the vault roof at higher infiltration rates and is evident when the seepage rate reaches a maximum value as infiltration increases. Therefore, once there is perched water, water is diverted around the vault and the vault hydraulic performance is unchanged even though infiltration increases. Water perches for intact concrete at lower infiltration rates than for degraded concrete. Additionally, intact concrete vault has a higher saturation than the degraded concrete vault. Paradoxically, the low permeability of intact concrete and good hydraulic performance promotes a greater degradation rate.

A summary of above conclusions from these modeling simulations is provided below.

 

• Clay layers placed adjacent to the concrete were found to lower water flow through the vault and enhance hydraulic performance.
• Smaller vault sizes result in lower flow rates and indicate a scale effect.
• Roof slope has a relatively small influence on hydraulic performance.

 

Although not evaluated in this work, the simulation results suggest that a “double containment” system comprised of outer and inner containments would yield an enhanced hydraulic performance over a single concrete vaulting system. The outer vault would initially provide a low permeability barrier to flow. Once the outer vault degrades, the water passing inside of the outer vault would be conditioned and then contact a smaller- scale concrete vault. As indicated in the results from this paper for scale effects, the smaller vault would have more of a capability to divert flow towards the sides of the vault. Future work will evaluate this design.