PROBABILITY ESTIMATION OF SILAGE EFFLUENT FROM HORIZONTAL SILOS

A mathematical model was developed to predict daily flow and total acc umulated silage effluent from a mechanically compacted horizontal silo (bunker or clamp). Generated daily rainfall over a 50-year period was used to estimate effluent flow and year-to-year Variations from a 200 t dry matter (DM) grass crop (1250 t silage at 16% DM) harvested by t hree different systems. A direct-cut non-stop system (DCNS) re suited in a total of 1600 L of effluent or less per t DM in a wet climate (14 40 mm rain per year) and 1120 Wt DM or less in a dry climate (720 mm r ain per year) at 95% probability (19 years out of 20). Daily maximum f lows at 95% probability were 185 L/t DM per day in the wet climate and 119 L/t DM per day in the dry climate. A design effluent storage capa city of 3 m(3) per 100 t silage was found adequate for one-day storage but could result in effluent overflow after two days under very wet c onditions. A second harvest system of direct-cut restricted to non rai ny days only (DCNR) reduced total effluent to 1150 L/t DM in the wet c limate and 670 Lit DM in the dry climate. Daily maximum flows at 95% p robability were 109 L/t DM per day in the wet climate and 54 Wt DM per day in the dry climate, a reduction of 41 to 55% compared to DCNS. A third harvest system that included field wilting (WS) during 6 h resul ted in total effluent of 68 L/t DM in the wet climate and 36 L/t DM in the dry climate. Daily maximum flows at 95% probability were 3 L/t DM per day in the wet climate and 2 L/t DM per day in the dry climate. M oderate wilting almost eliminated the problem of silage effluent but i t delayed the harvest period from 5 days with the DCNS system to up to 18 days with the WS system.