Ecological network analysis to link interactions between system components and performances in multispecies livestock farms

Organic farming relies on ecological processes to replace chemical inputs, and organic farmers have developed various strategies, including several forms of diversification, to remain viable. Herein, we hypothesized that diversified organic farming systems can enhance their performance by increasing the level of interactions between system components. We therefore performed an ecological network analysis to characterize both within-farm and farm-environment interactions. Flows were expressed on an annual basis according to the quantity of biomass exchanges multiplied by nitrogen content. Seventeen organic farms were surveyed in French grassland areas, each associating beef cattle with either sheep, pigs, or poultry. The ecological network analysis was then coupled with the assessment of farm economic, environmental, and social performances. A hierarchical clustering on principal components distinguished five groups of farms based on farm and herd size, presence of monogastrics, percentage of crops in the farm area, and system activity indicators. A large farm size, in terms of area or number of workers, can limit the implementation of a homogeneous flow network within the system. A higher level of within-system interactions did not lead to better farm economic, environmental, and social performances. Systems with large monogastric production enterprises were highly dependent on inputs, which led to less homogeneous flow networks and a poor farm nitrogen balance without gaining economic efficiency. Managing a complex system with a dense and complex flow network did not appear to increase farmers’ mental workload. To our knowledge, this study is the first to quantify farm-scale interactions using ecological network indicators in temperate livestock farms and to analyze the links between farm performance and operating processes. The ecological network analysis thus potentially provides a common framework for comparing a wide range of livestock farms. Given the variability of multispecies livestock farms, a larger database will be used to extend our conclusions.

Evaluation of Two-Phase Frictional Pressure Drop Correlations Against Experimental Data

Results are presented here from an experimental investigation on tube side two-phase characteristics that took place in four tested tubes—the 1EHT-1, 1EHT-2, 4LB, and smooth tubes. The equivalent outer diameter of the tube was 9.52 mm and the inner diameter was 8.32 mm. Condensation tests were conducted using refrigerant R410A at a saturation temperature of 318 K, over a mass flow range of 150–450 kg m−2 s−1, with inlet and outlet vapor qualities of 0.8 and 0.2, respectively. Evaporation tests were performed at a saturation temperature of 279 K, over a mass flow range of 150–380 kg m−2 s−1, with inlet and outlet vapor qualities of 0.2 and 0.8, respectively. Pressure drop data of the four tested tubes were collected to evaluate five identified prediction correlations based on the separated flow model and the homogeneous flow model. The separated flow approaches presented predictions with average MAEs of 24.9% and 16.4% for condensation and evaporation data, respectively, while the average MAEs of the homogeneous flow model were 31.6% and 43.4%, respectively. Almost all the identified correlations underestimated the frictional pressure drop of the 4LB tube with MAEs exceeding 30%. An earlier transition of different flow patterns was expected to occur in the EHT tubes while developing a new diabatic flow pattern map is needed for the 4LB tube. A new correlation was presented based on the two-phase multiplier Φ and the Martinelli parameter Xtt, which exhibited excellent predictive results for the experimental data.

Evaluation of Existing Frictional Pressure Drop Correlations During Condensation of R410A in Four Horizontal Tubes

Results are presented here from an experimental investigation on tube side condensation characteristics that took place in four tested tubes — 1EHT-1, 1EHT-2, 4LB and a smooth tube. The equivalent outer diameter of the tubes was 9.52 mm and the inner diameter was 8.32 mm. Condensation tests were conducted using refrigerant R410A at a saturation temperature of 318K, over a mass flow range of 150–450 kgm−2s−1, with inlet and outlet vapor quality of 0.8 and 0.2, respectively. Pressure drop data of the four tested tubes were collected to evaluate five identified prediction correlations based on the separated flow model and the homogeneous flow model. For 1EHT-2 and the smooth tube, all the listed correlations manage to present predictions with the Mean Absolute Relative Deviation (MARD) less than 30%, while they underestimate the frictional pressure drop of the 4LB tube with MARD exceeding 40% averagely. Regarding the experimental data, it is found that the Muller-Steinhagen and Heck correlation presents the most accurate and stable prediction for the 4 tested tubes. The listed homogeneous flow correlations can provide acceptable predictions with MARD ranging from 25% to 40% under a few conditions, but their average predictive accuracies are inferior to that of the separated flow correlations. Consequently, the separated flow approach performs better than the homogeneous flow model in the prediction of frictional pressure drop for our experimental data.