Of all the engineering issues associated with bioreactor landfills, slope stability is among the most important.

By R.C. Bachus, M.F. Houlihan, E. Kavazanjian, R. Isenberg, and J.F. Beech

Designers and regulators often express concern that the introduction of water or other liquids to enhance the degradation of the waste will reduce the stability of the waste mass. Although the introduction of liquids has several potentially destabilizing effects, they can be mitigated through sound design, construction, and operating practices. This article summarizes the state of the practice for bioreactor landfill design, construction, and operation and their influence on bioreactor landfill slope stability. Based on the information presented in this article, bioreactor landfills can be designed, constructed, and operated in compliance with regulatory requirements and standards of practice for slope stability.

Bioreactor Stability Overview

In a typical bioreactor landfill, leachate and/or other liquids are introduced into the waste mass to enhance waste decomposition, resulting in increased moisture content compared to a conventional, dry landfill. There are two stability-related technical issues that must be considered to address the introduction of liquids: (1) the impact of the presence of liquids; and (2) the impact of the accelerated degradation of the waste. To understand the impact of the presence of liquids, one must understand the migration of liquids within the landfill. The migration of introduced liquids into relatively non-homogeneous waste is often thought to result in random (and uncontrolled) migration of liquids through preferential flowpaths in the waste. Detailed analyses and field observations indicate that this is not the case. Actually, liquids first are absorbed by the waste until the waste reaches its field capacity and then migrate along the path of least resistance, which is predominantly downward through the waste but with a lateral component if lower permeability layers (e.g., daily cover) are present. The effects of accelerated waste degradation include increased gas generation, decreased organic content, decreased waste permeability, and increased settlement compared to a conventional landfill. These effects have, until recently, been difficult to predict, but studies have recently been published (as described later in this article) that provide data for designers to use when considering these effects in stability analyses of bioreactor landfills.

When performing a slope stability analysis, three issues must be considered to address the flow of the introduced liquid through the waste: (1) increased weight of the waste compared to the weight of "dry" waste; (2) the possibility of perched liquids, which h would cause a localized buildup of pore-water pressure; and (3) liquid migration along a layer, which could break out on the face of the slope and induce a veneer stability failure. A related stability issue involves the accelerated rate of landfill gas generation; if this gas is not collected, then gas pressure could build up within the waste mass and increase the pore pressure, contributing to instability and increasing the potential for sideslope seepage and uncontrolled gas migration.

The primary defense against these effects is the use of sound design, construction, and operational practices. The primary design practice is to promote efficient and uniform distribution of the liquid (and, concurrently, good gas collection) using vertical and/or horizontal piping networks within the waste. The primary construction practice is to provide hydraulic connection between the distribution/collection system and the waste by constructing the system using high-permeability materials and by preventing continuous, low-permeability layers within the waste that could restrict downward flow of liquid through the landfill. The primary operational practice is monitoring of the landfill to confirm that liquids are not building up in the landfill at levels that would cause excessive pore pressures. Use of these practices is common at both leachate recirculation and bioreactor landfills in the United States. Table 1 provides a list of bioreactor landfills that have been successfully designed, constructed, and operated in the United States. These bioreactors have been successful because they were intentionally designed as bioreactor landfills, rather than converted to bioreactors after the fact without attention to bioreactor design or construction considerations.

Stability: Conventional Versus Bioreactor Landfills

Field experience confirms that stability can be maintained for bioreactor landfills by using the design, construction, and operation practices previously described. Still, the designer must demonstrate that the bioreactor should be stable under the permitted operating conditions by performing a comprehensive slope stability analysis. This analysis can be performed using the same analytical tools as those used for landfills where liquids are not intentionally introduced. These analyses typically consider the following three potential failure modes (see Figure 1): (1) overall global stability of the waste mass, (2) local and/or deep stability within the waste or along discrete interfaces, and (3) veneer stability of the cover system. Limit-equilibrium methods, which are common in geotechnical engineering practice, are typically used. Commercially available computer programs are available (e.g., XSTABL, PCSTABL, SLIDE, SLOPE/W, UTEXAS, etc.) to analyze stability. For the third mode, which is specific to shallow, linear failure surfaces for cover systems and along soil-geosynthetic interfaces, closed-form analyses based on infinite or finite-slope evaluations incorporating force equilibrium considerations can also be used [Giroud et al. 1995; Matasovic 1991].

When performing a slope stability analysis for a bioreactor landfill, the designer must consider the differences between bioreactor landfills and non-bioreactor landfills. The key elements of the stability analysis, and considerations that should be made specific to bioreactor landfills, are described as follows.

Selection of Critical Cross Sections
This first step involves identifying the sections that have the lowest calculated factor of safety. The designer must consider the construction conditions, interim conditions, and final closure conditions when selecting the critical sections because the slope geometry and the buttressing influence at the toe of slope may be different in each of these conditions. Because construction and interim conditions may involve slopes are that are steeper than, but not as high as, the final closure slopes, slope stability must be considered at all phases of operation. The designer should also evaluate the length of time that interim conditions will exist. For "interim" slopes that are left for several years, the factor of safety should be nearly equal to that for final conditions.

Foundation Conditions
The location and extent of each type of material beneath the ground surface that could affect the stability analysis needs to be identified. In addition to the geologic materials that serve as the foundation for the landfill, the presence of geosynthetic interfaces must be considered, as geosynthetic interfaces are continuous planar surfaces that may be weaker than other surfaces passing through the foundation. It is common to perform circular and non-circular analyses that analyze the stability along the most critical interface and through the waste mass itself. Geosynthetic interfaces are also important because they can prevent downward migration of liquids, which could cause a buildup of pore pressures.

Shear Strength
The selected values of shear strength of waste and soils are critical to the calculated factor of safety because stabilizing forces are primarily a function of material shear strength and both the driving and stabilizing forces are a function of the weight. Extensive testing has been conducted on the shear strength of soil/geosynthetic and geosynthetic/geosynthetic interfaces that commonly exist in modern landfills. These data indicate that the shear strengths of these interfaces are typically lower than the shear strengths of waste or soil and that the presence of liquids may reduce interface shear strength. Fortunately, laboratory interface tests have long been conducted considering "wetted" or "hydrated" interface conditions in order to represent worst-case conditions. More recently, studies have been undertaken to assess the shear strength of solid waste. Figure 2 provides a summary of the recommended Mohr-Coulomb strength envelopes for MSW from several recent investigations. The bi-linear envelope from Kavazanjian et al. [1995] was developed to represent an approximate lower bound of measured laboratory strengths and back-calculated shear strengths as performed by the authors and as reported in the literature. Recent data from large-scale direct-shear tests conducted on waste recovered from bioreactor landfills [Kavazanjian et al. 2001], as well as laboratory testing and back-calculated shear strengths from landfills having highly degraded waste and zones of high liquid content [Isenberg 2003] indicate that there is little to no difference between the strength envelopes for "dry" waste from conventional landfills and "wet" degraded waste from bioreactor landfills. Still, because all landfills are different, waste shear strength properties should be selected to reflect the specific waste composition, density, placement and compaction methods, daily cover, and other operational practices.

Unit Weight
Recent field data indicate that the unit weight of MSW is higher than the values commonly used in practice. The authors have observed that "typical" values of MSW unit weight used in practice today are in the range of 55 to 65 lb/ft³. This range of values appears to be based primarily on old literature and reports from landfill owners based on the total weight of waste placed in the landfill, and does not account for the impact of cover soils and absorbed liquids on total unit weight. Figure 3 shows field data from recent in situ testing for unit weight performed at Operating Industries Inc. (OII) Landfill in southern California [Matasovic and Kavazanjian 1998] and at Tri-Cities Landfill in northern California [UC Berkeley 2003], two relatively dry MSW landfills. These data indicate "typical" unit weight values in the range of 80 to 110 lb/ft³, with the higher values occurring in areas having a high construction-and-demolition debris content. The addition of liquid to waste is likely to increase the unit weight to even higher values. In fact, field testing in saturated-waste zones at OII (which are not reported in Figure 3) yielded unit weight values as high as 135 lb/ft³ [Matasovic and Kavazanjian 1998]. These values of both wet and dry unit weight are significantly higher than the values typically used in engineering practice today. While an increased unit weight has little to no impact on the internal stability of a frictional material (because the increased weight proportionally increases the driving forces and the shear strength), it would reduce the stability of a cohesive material and could reduce foundation stability and seismic stability. Therefore, careful consideration of unit weight is critical in bioreactor stability analyses.

Assessment of Piezometric or Phreatic Surfaces:
Although the Mohr-Coulomb shear strength parameters of solid waste may be unchanged by the addition of liquids, addition of liquids could raise the phreatic surface, which could decrease the effective normal stress and decrease the shear resistance of the waste. Therefore, consideration of the liquid level within the waste is critical. For the past 15 years or so, the design and operational philosophy for conventional landfills has been to keep liquids from contacting the waste; therefore, landfills that do not add liquids and that have a functional leachate collection system are typically assumed to have no phreatic surface within the landfill. While the introduction of liquid has the potential to create a phreatic surface, the authors' experience is that bioreactors that are operated as recommended later in this article have no significant increase in phreatic surface and no buildup of head on the landfill liner system attributable to liquids addition [see Morris, et. al., 2003; Yazdani, 2004; and Waste Management, 2003]. This is because, unless liquid is introduced continuously (which is not recommended and would only be possible at landfills having substantial sources of leachate and make-up water), most liquid is absorbed by the waste. When liquids are introduced under pressure, the dominant migration pattern is vertically downward. When the pressure is reduced, downward movement of the wetting front essentially stops. Numerical simulations of liquids migration confirm this observation, as illustrated in Figure 4, which shows (1) little increase in liquids pressure in the waste mass beyond the immediate vicinity of an infiltration trench that is operated as recommended in this article, and (2) little tendency for liquids migration after the pumping cycle stops [also, see Reinhart and Townsend 1999]. Simulations and practice also indicate that, where infiltration trenches are widely spaced (see Table 1) and where adjacent trenches are not charged simultaneously, there is little potential for a widespread increase in liquid pressures in the waste mass.

Seepage Forces

Flowing liquids can induce seepage forces on the waste mass. This is of particular concern at the face of the landfill, where emerging seeps can induce veneer stability problems [see Giroud et al. 1995]. Operational practices that prevent continuous low-permeability layers within the waste and that control liquid injection rates and volumes can mitigate the potential for this type of stability failure.

Gas Pressure
Due to the high compressibility of landfill gas, internal gas pressure is generally not a significant consideration when evaluating landfill stability. However, recent studies of the Payatas Landfill failure in the Philippines [Fitz 2003] and the Dona Juana landfill failure in Bogotá, Colombia, [Gonzalez-Garcia and Espinosa-Silva 2003] indicate that, when the waste is close to saturation, landfill gas generation may induce significant increases in the internal water pressure due to gas-pore water interaction. Although this effect is the topic of ongoing studies, the authors believe that efficient collection of landfill gas and minimization of internal gas pressure should be considered an important operational consideration in maintaining bioreactor landfill stability.

Maintaining Bioreactor Landfill Slope Stability

Bioreactor landfills that are well designed and constructed can be as stable as conventional dry landfills. However, proper design and construction cannot, by themselves, ensure the stability of a bioreactor landfill. It is also critical that the bioreactor be operated in a manner that is consistent with the assumed design conditions. Recommended operational practices that minimize the potential adverse impacts on slope stability include the following:

Promote Absorption of Liquids
Liquids should be injected in a manner that promotes absorption, not just "recirculation." Absorption promotes wetting of a greater volume of waste than recirculation, and is therefore preferable to recirculation. Injecting liquids periodically (instead of continuously) tends to promote absorption and helps to prevent a buildup of excess pore pressure within the waste.

Control Liquids Injection Pressure
It is certainly possible to "overwhelm" the infiltration trench by injecting liquids under high pressure, which could effectively overcome the overburden pressure imposed by the waste and cause excessive lateral migration of the liquids. In addition, excessive injection pressure can cause unintended pore-pressure buildup. Therefore, injection pressures should be limited (e.g., typically less than approximately 5 pounds per square inch gauge) and liquid pressures should be allowed to dissipate between injection events.

Control Infiltration Trench Geometry
When horizontal trenches are used for liquids injection, the geometry of the infiltration trench needs to be matched to the operating conditions. Specifically, the slope, width, and depth of the trench should be designed to allow injection of the desired quantity of liquid under the prescribed pressure. If a trench is too large, then there will be insufficient liquids to allow the trench to properly operate under pressure and liquids may simply remain in the trench, restrained from vertical infiltration. Similarly, if the trench is not maintained a sufficient distance from the landfill sideslopes, breakouts could occur, which could adversely impact the stability of the sideslopes and the cover system.

Monitor for Indicators of Bioreactor Performance
In addition to the operating practices described above, the operator should monitor the performance of the bioreactor to confirm that the conditions assumed in the stability analyses are present in the landfill. This includes monitoring not only of liquids injection pressures and volumes but also of leachate generation rates. It is also valuable to monitor the changes of these parameters over time, as they may serve as early indicators of potential problems. Most important, however, is the development and implementation of a site monitoring plan that can detect flooded gas wells, poor landfill gas recovery rates, lateral seepage from the sideslopes, distressed vegetation, and odor problems. Collection and documentation of this information requires a commitment to regular, programmatic inspections by the operator.

Recommended Approach to Evaluating Slope Stability

Information presented in this paper suggests that the stability analysis of properly operated landfill bioreactors is not significantly different from the stability analysis of conventional modern landfills in which liquids are not introduced. Based on the discussions of the previous sections, the most critical issues related to slope stability of bioreactors, and the recommended approach for assessing them, include the following:

Slope Geometry
Assess the most critical operating, interim, and final closure grades/geometry and explicitly conduct analyses for these conditions.

Stratigraphy
Consider the specific subsurface stratigraphic conditions that are anticipated at the site, including stratigraphy both above and below the lining system.

Waste Properties
Consider site-specific interface direct shear strength tests conducted along wetted interfaces under the anticipated field normal stresses for any soil and/or geosynthetic interfaces that are introduced into the design; use the bi-linear Mohr-Coulomb failure envelope and the appropriate total unit weight for the waste (which considers daily cover and additional moisture, and which also accounts for settlement).

Operating Conditions
Develop project operating plans to control liquids injection pressures, injection schedule, and geometry of the infiltration trenches so that liquids are introduced in a manner that promotes absorption and downward vertical infiltration of liquids, not lateral migration. To prevent excessive liquids addition, establish a target maximum waste moisture content, such as field capacity, and calculate the amount of liquid required to achieve that moisture content and then monitor the amount of liquid actually added.

Monitoring
Monitor bioreactor performance to confirm that the observed field conditions match those that were assumed during the analysis.

Conclusions

This paper was prepared to present the key factors in analyzing slope stability of bioreactor landfills and the operating procedures for maintaining stable bioreactor landfill slopes. In summary:

Stability of a bioreactor landfill is controlled by the same parameters that control the stability of non-bioreactor landfills and non-waste fills, including material shear strength, the presence of pore pressure, unit weight of materials, and geometry. Available data document the range of values of these parameters that allow proper and safe bioreactor operation. Therefore, the stability of bioreactors can be evaluated using the same analytical tools as stability of non-bioreactor landfills.

The calculated slope stability factors of safety for bioreactor landfills that are operated as recommended in this article are not significantly different from those of conventional, non-liquid addition landfills. This is based on data that show that (1) the shear strength of waste from a bioreactor does not appear to be significantly different from the shear strength of waste from a non-liquid addition landfill, and (2) pore-pressure increases are not typically observed in bioreactor landfills that follow the operational practices recommended in this article.

Operations at bioreactor landfills may have a greater impact on stability than operations at non-bioreactor landfills, so additional performance monitoring is recommended, using the current state of the practice for monitoring, to verify that operations are not having an adverse impact on slope stability.

Robert Bachus and Jay Beech, P.E., are principals with GeoSyntec Consultants in Atlanta, GA. Mike Houlihan, P.E., is a principal with GeoSyntec in Washington, DC. Ed Kavazanjian, P.E., G.E., is an associate professor at Arizona State University, and Bob Isenberg, P.E., is a vice president and project director at SCS Engineers in Reston, VA.

MSW - September/October 2004