Further comment on ‘Landfill closure’ column

To the editor:

I want to express further thoughts on the Designer’s Forum column in the June/July 2005 issue of GFR (now Geosynthetics), titled “Landfill closure: A lesson in crisis management” by Gregory Richardson and William Chicca. This article is interesting but does not present the real reason for implementing the crisis management procedures described therein, namely the problematic design and lack of testing to confirm the design of the cover system.

The initial geosynthetic cover specifies a single-sided drainage composite with the geonet in contact with the underlying 20-mil PVC geomembrane on slopes that were designed to be 3H:1V (18.4°) but ended up to be as steep as 2.5H:1V (21.8°) after cover soil placement by the county. Because of the low normal stress present during cover soil placement, the geonet did not develop enough friction with the underlying geomembrane to produce a stable slope.

Besides the inclusion of the weak geonet/PVC geomembrane interface, other troubling aspects of the design were the stability analyses being conducted after the construction drawings were distributed for bid purposes. However, stability analyses were prepared prior to issuance of the Permit-to-Construct.

Second, the interface strength parameters used in the stability analyses were obtained from a geosynthetics seminar instead of site-specific interface testing. In addition, the assumed interface strength parameters were not verified prior to bid or construction. In fact, no interface testing was conducted until after slope instability developed.

Richardson and Chicca (2005) suggest that the designer used data from Stark and Hillman (2001) to develop the design interface friction angle of 23° but this paper was not published until after the date of the stability calculations. Even if the designer somehow obtained an “advanced copy” of the paper, it is recommended that designers not use information until it is published or contact the author to assess the status/usefulness of the data. For example, Hillman and Stark (2001) present interface data for a 30-mil PVC geomembrane and not a 20-mil PVC geomembrane as was used in this case. Hillman and Stark (2001) clearly show PVC geomembrane interface strength is a function of thickness and surface texture.

The selection of interface strength is especially critical in the cited case because the regulatory agency only required a state factor of safety of 1.3 for final cover slopes instead of the typical value of 1.5. Thus, the calculations performed using the “seminar” strength parameters only achieved a factor of safety of 1.3.

In conclusion, I recommend that:

1. design entities not “recycle” designs from prior sites and assume that the design will be suitable for another site;
2. slope inclinations are verified to match design assumptions before allowing bid and/or construction to proceed so oversteepened slopes are not encountered;
3. stability analyses be conducted before requesting bids;
4. site-specific interface testing be conducted to verify the design strength parameters prior to bid and construction; and
5. Maryland raise the required static factor of safety for landfill covers from 1.3 to 1.5 to prevent future landfill cover problems.

Timothy D. Stark
Professor of Civil and Environmental Engineering,
University of Illinois

Sources

Hillman, R.P. and Stark, T.D., 2001, “Shear Behavior of PVC Geomembrane/Geosynthetic Interfaces,” Geosynthetics International Journal, Industrial Fabrics Association International (IFAI), Vol. 8, No. 2, June, pp. 135–162.