Economic viability of implementing an infrastructure for recycling bedding sand from a free-stall facility for dairy cows

This study proposes to examine the economic viability of implementing the necessary infrastructure for the recycling of bedding sand from a free-stall facility in a milk production system in southern Minas Gerais, Brazil. In specific terms, the total production cost (TC), total operating cost (TOC) and effective operating cost (EOC) of a cubic meter of recycled sand were estimated in order to estimate the total sand consumption for the free-stall system and per bed year-1 as well as the equilibrium point of the amount of recycled sand, in cubic meters. The experiment was carried out on a farm located in the south of Minas Gerais from January 2016 to December 2017. Three scenarios were analyzed by the tree-point estimation method (MOP - most likely, optimistic, and pessimistic). Utilization of 85%, 95% and 75% of the recycled sand was considered for scenarios 1, 2 and 3, respectively. In all of them, the value charged per cubic meter of sand by a supplier close to the farm was considered. Monte Carlo simulation was also carried out with hurdle rates (HR) of up to 90%. Under the studied conditions, sand recycling showed to be economically viable in all scenarios, with positive net present values (NPV), internal rates of return above the HR, simple and discounted payback below the 10-year horizon, and satisfactory cost benefit-1 ratios (greater than 1). The EOC of one cubic meter of recycled sand was estimated at R$5.04, R$4.51 and R$5.72 for scenarios 1, 2 and 3, respectively, whereas the average TC, considering all scenarios, was R$6.84 (+0.81), which is less than the acquisition price of R$28.57 at the sand extraction site. The TC was R$37,219.51 and R$34,637.74 for the scenarios with HR of 8.50 and 6.99%, respectively, whereas TOC was R$22,572.08 in all analyzed scenarios. The estimated total annual sand consumption by the free-stall system was 526.44 m³, with an estimated average of 1.23 m³ (+0.28) bed-1 year-1. All Monte Carlo simulation models showed positive NPV as well as HR of up to 90%, which reflect a high probability of positive NPV.