Geotextiles in dams: Forging federal guidelines

By Douglas A. Crum

Abstract

The major building era for large dams in the United States, from the 1940s to the 1970s, ended about the time geotextiles were becoming popular. For the last 40-plus years, geotextiles have been used in many dams worldwide for various applications.

During this period, federal agencies have resisted experimentation with geotextiles in critical applications, agency policy has been to generally restrict use, and there have been minimal criteria governing use in federally funded water resource projects.

Geotextiles have been used for various purposes on levees and dams, sometimes (by some interpretations) against agency policy. Geotextiles are often recognized to provide economical solutions. In some cases, geotextiles have been reported to provide the only viable solution, which can be especially useful in remediation of existing structures. Current projections are for an increase, perhaps an explosive growth, in the rehabilitation of levees and dams.

On the other spectrum, internal filters for large dams are a critical application and a controversial issue. The federal approach for utilization of geotextiles should not condone use through absence of criteria. Compromise solutions based on technical merit alone are not adequate to resolve restrictions governing geotextile use. For critical applications, criteria must be based on design resiliency and risk tolerance. This article offers a perspective concerning pragmatic use of geotextiles for water resource projects, especially in critical applications and high-hazard structures.

Introduction

The purpose of this article is to promote establishment of criteria concerning use of geotextiles in dams. Criteria establishment has a dual purpose:

• to increase use of geotextiles where acceptable.
• to avoid misuse of geotextiles where unacceptable.

Example uses

There is a variety of potential applications for geotextiles in dams. The most common uses of geotextiles in embankment dams in the U. S. are: as a riprap filter placed on the upstream slope or in a downstream discharge area, as a riprap filter used to line watercourses, or as a filter in a downstream trench drain (often remedial work after construction).

Less common embankment dam applications include: as a protective layer and drain placed in contact with an upstream waterproofing geomembrane, as an internal filter, and as an internal drain.

Internal drains and filters are the most controversial. The applications shown in Figure 1 are not intended to define acceptable uses but to describe some of the potential uses that are contemplated, for the purposes of this article (most of these have been used and are in existence).

Performance requirements for dams

Misuse of geotextiles primarily focuses on the potential problems associated with using a geotextile as a critical and nonredundant design element deeply buried in a dam. It is the position of the National Dam Safety Review Board that geotextiles should not be used within a dam where they are both critical for dam safety and inaccessible for repair or replacement. This view was formulated during a workshop in October 2000, where the participants reached a consensus that:

• geotextiles should not be used in applications that are both critical to safety and inaccessible for replacement.
• geotextiles can possibly be used in locations that are critical for safety but accessible for replacement.

However, the workshop participants qualified the second case with the caveat that the engineer must assess the potential hazard posed by failure of the geotextile and the time available to respond and repair or replace the geotextile.

Case histories

In 1986, the International Conference on Large Dams prepared a special publication to address use of geotextiles in dams (ICOLD 55). It concluded: “The cautious adoption of geotextiles in dam engineering properly reflects the uncertainties involved in using a new and novel material.” Furthermore, “Successful performance of geotextile filters in noncritical locations cannot be used to validate use in more highly stressed locations where behavior mechanisms may be different. It has not yet been demonstrated that geotextiles can provide an equivalent to properly designed granular filters for this purpose, particularly when protection against cracking or the effects of earthquakes is required.”

Although this document is now dated, and there have been some pioneering efforts in both research and construction toward breaking this barrier, these concerns are still relevant.

There is a study awaiting publication (FEMA, draft document) that includes a literature review of geotextile applications in dams. The use of geotextiles for embankment filtration and drainage is mainly evident in France, Germany, China, and South Africa. A common factor in foreign practice regarding the use of geotextiles as filters and drains is the requirement that design must include rigorous performance testing to evaluate filtration and permeability using the proposed geotextile materials and actual soils from the project site. The findings evident in the literature are almost entirely positive, where geotextiles appear to be functioning adequately.

While the apparent successful use of geotextiles in constructed dams cited in published literature would suggest acceptance of geotextiles, the conclusions derived need to be taken cautiously. Often, technical publications are submitted within a few years of construction, and clogging may occur over much longer periods of time. Bad experiences are much less likely to be published for various reasons connected with litigation or designers’ reputations. Undocumented experience from the grapevine often alludes to bad experiences.

The past success stories alone do not define a threshold of acceptable use. Failure analyses are often more informative but they, too, have shortcomings. Failed designs are difficult to obtain and designers are understandably reluctant to advertise their mistakes. If the findings from postmortem analyses are that the geotextile was used with little or no testing and analysis, then the conclusions are trivial and may not help to define the threshold envelope. While the geotextile industry criteria and experience may be substantial for conventional applications, the reviewed case histories are not substantial enough to defend a higher degree of confidence desired for high-hazard structures.

Concerns regarding long-term performance of materials, products, or innovative construction methods are not attributable to geosynthetics alone. As an example, drain pipes embedded within embankments have proven to be problematic. Corrosion of the pipes can lead to loss of embankment and sinkholes. Clogging can lead to elevated water levels. The liability for these risks is not fully appreciated until performance assessments identify increased risks from low probability but high consequence failure events. Interim risk reduction measures imposed as mitigation to these concerns can be very costly. Also, the time to verify unsatisfactory performance, plan/design repairs, obtain funding appropriations, and construct the repairs can be unacceptable.

Existing agency criteria

The status of existing criteria at some agencies with missions in design, construction, and operation of dams was reviewed in a recent FEMA-funded study (FEMA draft document). A summary of the findings is included below.

Bureau of Reclamation

Geotextiles can be used in embankment dams but not as a sole element in a critical application (Bureau of Reclamation, 1992).

Corps of Engineers

Many documents provide a brief mention of geotextiles but sometimes lead to conflicting criteria. The most representative excerpt pertinent to dams is: “Since long-term experience is limited, geotextiles should not be used as a substitute for granular filters within or on the upstream face of earth dams or within any inaccessible portion of the dam embankment. Geotextiles have been used in toe drains of embankments where they are easily accessible if maintenance is required and where malfunction can be detected” (TM 5-818-8).

Natural Resources Conservation Service

The use of geosynthetics is interpreted as a variance to design policy since it is not explicitly allowed, so the use in dams is deferred to agency review by national technical staff. Unwritten policy does not permit the use of geosynthetics as an internal design component of a dam, particularly as a filtering/drainage function. Geotextiles have been used extensively in a separator function, placed between rock riprap and soil in rock-lined plunge pool basins and outlet channels downstream of concrete stilling basins. Geotextiles have also been permitted beneath riprap on the upstream slope of dams for wave protection.

Federal Energy Regulatory Commission

The unwritten policy is that geotextiles are not allowed in inaccessible areas of embankment dams and are not accepted as a filter for riprap. Geotextiles have been permitted in trench drains that are accessible for repairs.

State dam safety programs

In a poll of state dam safety engineers, only 8 of 40 respondents indicated that their states had implemented standards or guidelines that prescribe the use(s) of geosynthetics in the design or construction of earthen embankment dams. Fourteen respondents expressed reservations about using geosynthetics in the design and construction of earthen dams. The respondents ranked their reluctance based on: (1) absence of design standards, (2) poor performance, (3) too few examples, (4) lack of experience, or (5) insufficient support. This showed a wide dispersion of answers, leading to the interpretation that all of the above are considered valid concerns.

Private practice

Based on a poll of design firms, the consensus is that geotextiles should not be used for critical applications such as filtration and drainage, where their failure could affect the integrity of the dam. Liability is a primary concern to consultants.

Discussion

In conclusion, the only federal agency that has a comprehensive policy on use of geotextiles in dams is the USBR, but that document was last updated in 1992 and does not reflect current practice. Currently, both the USBR and the Corps of Engineers have works in progress to update criteria.

In the nonfederal sector, note that reliance on professional judgment creates the potential for a conflict of interest. While large, established engineering firms may take liability very seriously, small design firms with limited assets and history can afford considerably more risk. Without state policy on the use of geotextiles in dams, the regulating criteria depend heavily on ethical conduct for professional practice. Ethical conduct intertwined with innovative technologies leads to subjective design details without the hindsight of failure. The liability to the consultant is not necessarily linked with the liability to the public.

Performance concerns

Performance issues with geotextile products fall into 3 common categories:

1. limitations that can be accommodated by design, conservative assumption of material properties, and use of safety factors.
2. harsh environments that may be recognized and avoided.
3. inherently high-risk applications.

Geotextiles have numerous potential applications in civil works. Some applications in water-retaining structures have high consequences if they fail and should be avoided or used with caution and full understanding of the risks.

Under the first category, uncertainties may include ultraviolet degradation, installation damage, and design methods for using geotextile products. These concerns are normally addressed in design conservatism and construction quality control. These common performance issues are relevant to all applications of geotextiles in civil engineering and construction.

Harsh environments have contributed to documented failures of geotextiles. However, these conditions can usually be recognized through site investigainvestigation and experience gained from past performance problems. Environmental degradation includes the effects of ultraviolet light, high temperatures, hydrolysis, chemicals, radioactive materials, and biological organisms. Degradation of the polymers used in geotextiles has been shown to be inconsequential for typical soils. Difficult applications may also contribute to geotextile failures. Filtration in dispersive soils and filtration in flow reversal are applications that can cause clogging. These issues require increased attention to detail and in some cases, high-risk environments can be avoided by early recognition.

The first two categories require attention to detail, but do not preclude use of geotextiles for classes of applications, such as in dams. The last category does. Perhaps the biggest concern is use of geotextile filters for drainage in portions of a dam that are not readily accessible for removal and replacement. Failure of an internal feature of a large dam is not only very expensive to repair, it could jeopardize the safety of the dam. Designers are cautioned to evaluate the consequences of failure during the design process and decide if the risks are worth taking.

Clogging potential for granular and geotextile filters

The advantages of geotextiles in conventional civil engineering practice typically outweigh the risks associated with clogging. In critical applications, however, it is a cause for concern. It is generally known from experience that geotextile filters have, on occasion, experienced significant clogging problems. It has also been determined that generally these failures could have been avoided by increased attention to detail and design rigor. However, there are some concerns that suggest geotextiles are inherently more susceptible to clogging than granular filters:

Blinding and particle deposition. There are 2 competing requirements for design of granular filters: retention and permeability. The filter must be sufficiently large to pass flow, but sufficiently small to retain the base soil. There are 3 requirements for design of geotextile filters: retention, permeability, and clogging. Particulate clogging of granular filters is generally limited to screening for problematic conditions such as internally unstable soils (gap graded), dispersive soils, or mixed/stratified zones of variable soil types.

These same issues are also problematic for geotextiles, but geotextile design additionally includes explicit clogging criteria in the design process. The retention at a plane (or sheet) in the soil concentrates particles that migrate in the soil but are retained on the filter. To avoid particle clogging, the number and size of openings in the filter are increased to pass the small particles not necessary to retain the skeletal structure of the base soil. When the design criteria for clogging is more stringent than for permeability, then the design sensitivity (range of admissible sizes) is tighter and the construction tolerances are tighter. It is a physical limitation that controlling the opening size for geotextile filters will always have tighter tolerances than granular filters. Conversely, geotextiles are manufactured products with better concontrol of parameters than natural materials. The impacts of this on filter reliability have been debated in the geosynthetics industry with no clear consensus.

Intimate contact at the discharge face. It has been shown that fabrics that are not confined (i.e., those that move transverse to the flow) are more susceptible to clogging. Presumably this is because a moving fabric does not form a natural filter, so over time suspended particles eventually find a hole to become lodged in, contributing to blinding or particle deposition within the fabric. If the fabric is confined, and if the apparent opening size (AOS) is appropriately sized, the fabric replicates the soil structure close enough so a natural filter forms.

Growth medium. In addition to particle clogging, the polymer sheets can provide a medium for chemical precipitation or biological growth.

Integrity

There are some applications when 100% integrity of the geotextile is important. Internal zoning of embankments (e.g., impervious cores, rockfill, chimney drains) require a high confidence of filter integrity, which is difficult to assure with geotextiles. Generally, flaws fall into 2 categories:

1. Installation damage. Constructed features with flaws from installation damage have a potential to propagate or expand over time, causing a crisis or failure. To a degree, the QA/QC can be commensurate with the risk from installation damage. However, there is a practical limit for the intensity of QA/QC that is productive or effective. The practical limit of QA/QC may be governed by cost-effectiveness, destructive sampling, or sampling theory that an infinite number of samples are required to achieve 100% confidence. For high-hazard structures, the practical limit may occur before the acceptable risk is achieved.
2. Post-construction damage. Internal filters in earth embankment dams are, to a limited degree, self-healing. This resiliency has an important benefit when considering consequences from remote events but high-risk consequences. Granular filters may be resilient if they were to be penetrated by displacement from slope instability, seismic events, hydraulic fracturing, root or animal penetration, or some other circumstance.

Accessibility

The ability to access portions of a structure is relative. Accessibility should be determined on the basis of a plan that shows access for replacement is practical under the basis of cost and normal operating budgets. Given unlimited time and funding, almost everything is accessible. However, the practicality of repairing features that are difficult to access must be considered. If the cost to complete the repair exceeds the budget available for routine operations and maintenance, then it might not get accomplished. If the time for planning, design, and construction exceeds the interval between when the problem is recognized and when a design loading event occurs, then it is too late. If the performance is first recognized as distress to other portions of the facility that are more costly and to which the feature is ancillary, then, again, it is too late. Deeper than 20ft is generally not accessible for removal and replacement. Replacement may not be possible during certain loading conditions. Work that is intrusive sometimes requires pool restrictions. Since performance limitations are most noticeable during record pool levels, this is a particular concern for dams with storage allocated to flood control.

Restrictions and criteria

There are primary concerns where use is cautioned:

• geotextiles used within any inaccessible portion of dam embankments because of residual risk for clogging.
• geotextiles used in toe drains of embankments may be related to the embankment stability, but they are easily accessible if maintenance is required. This assumes that malfunction can be detected before unusual (flood or seismic) events.

The preceding sections allude to the necessity of providing restrictions and criteria, but do not provide exact terms. There is reason for difference in the criteria between agencies, depending on size and purpose of dams, hazards presented by typical dams, risk tolerance, and postconstruction rehabilitation needs. Criteria can be prescriptive based, performance based, or a combination of both.

Prescriptive criteria

Some of the possible applications for geotextiles in dams include ancillary and temporary features. These applications are generally more acceptable since they do not generally lead to highconsequence failure scenarios.

Ancillary features: Features that do not affect the dam performance are acceptable use.

Such features may include:

• drains for ground maintenance
• pavement applications
• landscaping

Temporary works: Caution is advised, as large cofferdams may be in place for several years and can incur significant damages. Temporary works may include:

• cofferdams
• construction dewatering
• construction access roads
• inverted filters
• sandbags

Rehabilitation of existing dams: There are cases where alternatives for rehabilitation of existing dams using geotextiles have been presented as the only feasible alternative. The only technically feasible alternative would be difficult to justify for new projects. In structural repair or rehabilitation, there are cases where funding limitations merge economic alternative issues into technical feasibility issues. A granular filter zone and a geotextile filter are generally economic alternatives that are compared using life cycle cost analysis. But if a granular aggregate source is not locally available, and the cost to import materials renders the project infeasible or delays it indefinitely, then the technical alternatives are to compromise the design parameters by reducing the filter zone or layer thickness, or using an alternative approach such as geotextile filters. From a risk-based approach, it may become justified to use geotextile filters, even in critical applications.

Risk methodology should not be used to lower standards, or to circumvent codes, regulations, or criteria. However, risk assessment is highly useful for prioritizing remediation activities in a resource-constrained environment. Allowance of geotextile filters in critical applications is perspective dependent, bearing on competing goals of maximizing use of resources or maintaining high standards. If the risk minimization goal is favored, then allowing geotextile filters in critical applications should only apply to existing projects, where the do-nothing and decommissioning alternatives are not acceptable.

Low-hazard dams and agricultural levees: Low-hazard structures (FEMA 333) that do not impose a health and safety risk are not a primary concern.

Crack stoppers: Transverse embankment cracking is a concern in arid climates.

Geotextiles have been proposed as a membrane to prevent soil erosion through open cracks. The issue of embankment cracking is a concern for existing dams constructed with natural materials, so the use of geotextiles for mitigation is not a substitution for a preferred alternative.

Performance criteria

Because of the vast alternatives for incorporating geotextiles into dams in various geometries and functions, the prescriptive criteria may not always apply or will be subject to interpretation.

The hazard of a geotextile in a dam can be identified in an analogy to the Fracture Critical Member concept in structural steel design. If failure of a certain component can lead to the collapse or failure of the structure, then it should be labeled as a Failure Critical Component (FCC).

For example, if a drain is installed for landscaping or grounds maintenance concerns, and if the embankment stability meets all factors for safety with the drain 0% effective, then the drain would be non-FCC. However, if a drain is installed for the purpose of reducing the phreatic surface and the design computations for the structure’s stability relies on this drain, then the drain is a FCC.

If there are 100% integrity concerns (where holes, flaws, or imperfections in the component may render it ineffective) then it is probably a FCC. If engineering analysis is not required by regulation, code, or funding authority, then it is probably not FCC.

Conclusions

While it is acknowledged that geotextile filters have been used in critical applications, particularly in projects outside the United States, this should not directly lead to acceptable industry practice. The amount of risk imposed by such projects has not been substantiated; endeavors by industry experts do not justify common practice for all practitioners.

Guidance regulating the use of geotextiles in dams is currently lagging behind the industry knowledge base and technical capabilities. Agencies that design, operate, or regulate high-hazard dams should establish explicit written criteria for the use of geotextiles.

Clogging of geotextile filters is a paramount concern governing geotextile use in dams. The implications regarding the use of geotextiles as filters and drains in embankment dams needs to be fully considered. In some applications, the zerodefects concern may be paramount.

While a properly designed geotextile filter is preferred over an inadequately designed or improperly applied granular filter, increased emphasis on design does not change the limits of what is obtainable. Granular filters have a longer history, suggesting a more reliable design than geotextile filters, given a similar level of design effort, expertise, and quality control/assurance.

Risk management in dam safety delves into consideration of low frequency/high hazard events. Such considerations lead to standards that favor very conservative design and are not tolerant of innovative technologies. It should not come as a surprise that regulators will impose some restrictions on use of geotextiles in dams.

There are many applications associated with dams with varying implications to dam safety. Some applications are appurtenant to the dam stability, some are redundant, and some are amenable to monitoring and maintenance. Criteria for use of geotextiles in dams should be formulated in consideration of critical and noncritical applications.

Doug Crum is the Dam Safety Program Manager, U. S. Army Corps of Engineers–Kansas City District, Kansas City, Mo.; douglas.a.crum@usace.army.mil. This article is an edited version of a paper presented at the 2008 USSD Conference in Portland, Ore., April 28–May 2, 2008.

References

Bureau of Reclamation, Design Standards, No. 13 – Embankment Dams, Chapter 19 –Geotextiles. U.S. Department of the Interior, Bureau of Reclamation, Denver, Colo., 72 p., (1992).

FEMA 333, Federal Guidelines for Dam Safety: Hazard Potential Classification Systems for Dams (1998).

FEMA (draft document), Geotextiles in Embankment Dams: Status Report on the Use of Geotextiles in Embankment Dam Construction and Rehabilitation, (pending publication).

ICOLD 55, Geotextiles as filters and transitions in fill dams, International Commission on Large Dams, Bulletin 55, 129 p., (1986).

TM 5-818-8, Engineering Use of Geotextiles, U. S. Army Corps of Engineers, (1995).