Direct Energy Inputs for Milk Production in Polish Dairy Farms

Increasing milk production efficiency, along with the ongoing consolidation and mechanization of farms, is making efficient energy use more critical than ever. Therefore, this study aimed to determine the direct energy inputs for milk production at the farm level in Poland, based on diesel fuel and electricity consumption in production processes and in milk collection and storage. The study used data from 75 dairy farms in Poland, classified into three farm-size categories based on total farm area: small farms (<50 ha), medium-sized farms (50–100 ha), and large farms (≥100 ha). The average energy-intensity indicators for milk production were as follows: electricity consumption of 0.044 kWh kg–1 of milk, diesel fuel consumption of 0.272 kWh kg–1 of milk, and total direct energy consumption of 0.317 kWh kg–1 of milk. Medium-sized farms had the lowest electricity consumption (0.041 kWh kg–1 of milk). For small and large farms, this indicator was 7.3% and 19.5% higher, respectively, than for medium-sized farms. Medium-sized farms also showed the lowest diesel fuel consumption, at 0.261 kWh kg–1 of milk. For small and large farms, this indicator was 2.7% and 11.5% higher, respectively. Accordingly, medium-sized farms had the lowest direct energy consumption, at 0.302 kWh kg–1 of milk. For small farms, this indicator was 3.3% higher, whereas for large farms, it was 12.6% higher than for medium-sized farms. The unadjusted differences in direct energy intensity between farm size categories were not statistically significant. However, after accounting for milk yield per cow in the ANCOVA analysis, a significant effect of farm size category was observed (p = 0.0134), with small farms showing significantly lower adjusted total direct energy intensity than large farms (Holm-adjusted p = 0.011). Nevertheless, sensitivity analyses showed that the farm-size effect was no longer statistically significant after additionally accounting for grassland area (p = 0.097) or arable land area (p = 0.149). These results should therefore be interpreted only as a characterization and analysis of statistical relationships within the surveyed farms, rather than as evidence of causal relationships or a representative assessment of all dairy farms in Poland.