| Abstract: | This meta‐analysis based on 19 studies from Finland comprising 43 grass silages was undertaken to evaluate the effect of silage quality on liquid yield, liquid composition and retained compounds in liquid using four different liquid–solid separation methods. Silages were classified according to species (grass, clover or a mixture of them), additive treatment (no, biological or formic acid‐based additive) and harvest (primary growth or regrowth). A mixed model regression analysis with random study effect was used to evaluate the impact of silage characteristics on biorefinery efficiency. There was a large variation in silage quality in the data set. Silage dry‐matter concentration was the characteristic most highly correlated with liquid yield for all separation methods, and when used as an independent variable in the model, it resulted in the best predictions. The liquid–solid separation methods presented a great variation in the liquid yield, ranging from 0.26 to 0.56 when silage dry‐matter concentration was standardized to 250 g/kg. There was no effect of additive treatment and harvest in the estimation of the biorefinery potential, but species was a significant variable in predicting liquid yield for the laboratory‐scale presses with higher liquid yield for mixed grass and legume. The high correlation between silage quality and liquid yield and liquid composition provides potential to predict the biorefinery potential based on equations developed for each separation method. This information can be used to modify the silage production systems so that they best meet the requirements of a green biorefinery process. |
