4. Conclusions
Multiple Linear Regression (MLR) analysis was used on the lab
scale data to quantify the effect of waste composition, rainfall
and ambient temperature on the first-order decay constant (k).
The best model was selected using the backward elimination
method, best subsets method and stepwise regression method,
such that all parameter were significant at a = 0.1. The best model
was found to have an adjusted R2 of 0.75. A second regression
equation (R2 = 0.79) was developed from field data from 11 land-
fills in high-income countries with conventional operation, in
order to scale-up the k estimates from the first MLR equation
(based on the lab-scale data) to field-scale values; field-scale
methane generation rates are lower because conditions are not
ideal.
The Capturing Landfill Emissions for Energy Needs (CLEEN)
model was developed by incorporating both regression equations
into the first-order decay based model for estimating methane
generation rates from landfills. The CLEEN model can be used for
predicting methane generation rate from landfills in high-income
countries with conventional operation, receiving rainfall between
2 and 12 mm/day, annual ambient temperature from 20 C to
37 C (and perhaps lower), and biodegradable waste components
ranging from 0% to 100%. Future work will develop scale-up factors
to allow the model to be applied in low-income countries and to
landfills with bioreactor/enhanced leachate recirculation
operation.
CLEEN model values were compared to actual field data from 6
US landfills, and to estimates from LandGEM and IPCC. For 4 of the
6 cases, CLEEN model estimates were the closest to actual.