In the last Waste Wise column, I highlighted some of the issues with LandGEM, the model used to estimate how much methane is produced in a landfill. I also indicated the need for better science in estimating landfill gas production. This month, we continue this conversation by highlighting some of the key efforts in research and data development aimed at addressing the acknowledged deficiencies in estimating landfill gas production.

The primary limitations in the LandGEM model center around the accuracy of the two variables used, k and Lo. To recap, Lo is the ultimate methane production potential in a given volume of waste. The widely held view is that Lo is an inherent property of the waste itself and its variability is primarily affected by waste composition. The decay rate, k, represents the speed at which organics decompose in a landfill. The speed at which gas is produced is extremely important as this dictates gas flow rates from a landfill that may be used for gas-to-energy systems or influence uncaptured or “fugitive” emissions. 

In the last column I noted the relative uncertainty of this variable. Current research efforts have focused on ways to increase the accuracy in estimating k by evaluating the potential for developing field-based values that account for site-specific variables such as waste characteristics, temperature, rainfall and site configuration.

EPA guidelines suggest using a k-value of 0.04 yr-1 when modeling landfill gas production. However, this estimate is based on a limited data set and has been demonstrated as having a high level of uncertainty. In 2013, research articles were published by the University of Central Florida (“Comparison of first-order-decay modeled and actual field measured MSW landfill methane data”; Waste Management) and North Carolina State University (“Using observed data to improve estimated methane collection from selected U.S. landfills,” Environmental Science & Technology) that tackle these issues head on.

The UCF study evaluated 3 landfills that had been receiving waste for 9, 10 and 34 years. All sites had gas collection systems installed with some type of cover system. Site specific data on waste composition, collected landfill gas and uncaptured emissions were quantified and used to develop field-based estimates of k-values for these landfills. In the N.C. State study, 11 landfills were evaluated in non-arid U.S. regions and, similar to the UCF study, site specific data on rainfall, ambient temperature, waste quantities and landfill gas volumes were captured. These 11 sites had been receiving waste for periods ranging from 15 – 55 years.

The UCF study estimated k-values of 0.04, 0.06 and 0.09 yr-1 for the landfills studied. The N.C. State study gave k-values of 0.04 yr-1 for two of the sites, but k ranged from 0.09 to 0.17 yr-1 for the other nine sites. The fact that these two independent studies gave similar results in their estimates of k is notable. Collectively, 11 out of 14 sites had k-values higher than the EPA guideline of 0.04 yr-1, with most of these estimated being at least two to three times higher.

This finding is significant and means that, if the landfills studied are representative of most landfills in temperate regions of the developed world, landfill gas production for active or recently closed landfills could be double to triple the flow rates predicted by LandGEM using default values. 

While this demonstrates the need to update the EPA default values, this also could indicate that landfill-gas-to-energy systems may be underdesigned to handle all the gas generated at a landfill if LandGEM and its default values are used as the sole tool to estimate landfill gas production. The other impact these findings have relates to the accuracy of greenhouse gas inventories, which use results from models like LandGEM to estimate the impact of landfills on global warming potential and direct federal, state and regional policies in this area.