Experiments were carried out to monitor behavior of odour intensity, by means of odour sensor, in a closed breeding facility with internal dimension of 850°1550°1850 (Height) for a Shiba goat equipped with a catalyst filter system for the purpose of biological contaminant elimination for CELSS. A response value of odour sensor at 23 hours in a blank test without the filter system was 322 point and that in an elimination test with the catalyst filter system was 116 point. Author's subjective evaluation of this difference in these response values concluded that the catalyst filter system performed significant reduction of odour intensity. No correlation was observed between behavior of response values and excretion pattern of Shiba goat, and also between behavior of response values and methane production. Drastic fermentation of excreta remained in excreta separator was not observed at the end of the each test. Steep increasing of the response values was recorded during 30 minutes period after feeding in the tests. Shiba goat was particularly busy in taking bait, and a quite few excreta and lively metabolism were observed and recorded in this period. It was, hence, suggested that contaminants via expiration and body odour were rather dominant than excrete in terms of odour intensity in the tests. GC Mass analysis showed that any contaminant desorbed from the catalyst filter operated in the elimination test was not detected at 25°C, which meant no adsorption equilibrium on the catalyst filter occurred during the elimination test. A reason why the response values was kept around 100 point since 2 hours after feeding in the elimination test was due to that the odour sensor was located in downstream between Shiba goat and the catalyst filter, and detecting fresh odour coming from Shiba goat. Though chloric compounds as a typical contaminant disturbing responsibility of the odour sensor were not detected in this study, an effect of contaminants on responsibility of the odour sensor might be a subject for further research.

View full abstract

A mathematical model is set up for the batch recirculation reactor system (consisting of a perfectly-mixed flow room, plug-flow reactor, and plug-flow pipes situated in upstream and downstream regions of the reactor), which may be constructed to treat harmful substances in a CELSS, and the performances of two numerical calculation methods (Method 1; a combination of Runge-Kutta and modified Runge-Kutta methods, Method 2; a combination of Runge-Kutta and Euler methods) are investigated. First, these methods are applied to a simplified batch recirculation reactor system with first-order reaction kinetics and negligible reactor volume and the numerical solutions are compared with the analytical solutions. As a result, it is found that the accuracies of the numerical solutions by Method 1 and 2 consistently increase with the increase of step number and Method 1 gives numerical solutions whose significant figures are always larger by at least two-digits than those of Method 2. Method 1 is further applied to a practical reactor system with first-order or Langmuir-Hinshelwood type reaction kinetics and finite reactor volume to elucidate the effects of initial conditions and systemic parameters on the characteristics of the practical batch recirculation reactor system. The results of the present work clearly show that Method 1 promises highly stable and reliable numerical calculation because of the introduction of modified Runge-Kutta method and the devised calculation algorithm.

View full abstract