1). Each was fed with animal effluents and energy crops
at different ratios. At A.D. plants A and C, several industrial
by-products, such as slaughter waste, tomato skins, and glycerol
were also employed as feedstock. Plant B was fed with very fibrous
feedstock, comprised primarily of straw and cut maize while plant
D was 90% swine effluents. The highest loading rate (2.25 kgVS/m3)
existed at plant A, whereas plant C had the lowest one (0.96 kgVS/
m3).
All samples were collected at the exit of the post-fermenter and
before their inlet into the storage tank. Samples were stored in
gas-tight containers for transport to the lab and used for batch
trials within five hours of their collection. Batch trials were
performed in 2 L anaerobic batch reactors according to Verein
Deutscher Ingenieure (VDI) 4630 (2006) and following the experimental
procedure described by Dinuccio et al. (2010). Specifically,
at the beginning of the trials, each reactor was filled with 0.9 L of
digestate and sealed with glass taps connected to Tedlar gas bags.
Samples were incubated at 40 C (±1 C) in a temperaturecontrolled
room for 70 days. All reactors were manually stirred
twice a day. Each trial was performed in three replicates.
We regularly measured biogas with a drum-type gas meter
(Ritter, Bochum, Germany) and analysed methane content by an
infrared analyser (Draeger XAM 7000). The recorded data were
normalized at standard temperature and pressure (0 C and
1013 hPa) according to German Standard Procedure (VDI 4630,
2006). Since reactors were not completely filled, the normalized
volumes of biogas and methane were corrected with the equation
reported