SUNScholar will be unavailable on 25 April 2018 from 08:00 to 10:00 South African Time for routine maintenance. Please contact scholar@sun.ac.za for queries.

Predicting the effect of biofilm growth on the pressure drop over a biofilter

van Jaarsveld, Jacoline (2017-12)

Thesis (MSc)--Stellenbosch University, 2017

Thesis

ENGLISH ABSTRACT : Hydrogen sulphide is a poisonous gas produced by several industries. It is therefore crucial to investigate and mathematically model the environmentally friendly biofiltration process, which has the ability to remove poisonous gasses, such as hydrogen sulphide, from an air stream. In this study two approaches are used for modelling biofilm growth in three different biofilters, i.e. an empirical and analytical approach. In the empirical modelling approach, the pressure drop prediction of the Modified-Macdonald equation, the existing granular rectangular Representative Unit Cell (RUC) model and the model of Comiti and Renaud are used to determine the changes in biofilm affected porosity, specific surface area, tortuosity and biofilm thickness. The results are obtained by using Excel R Solver, which is based on an optimization method. Thereafter a sensitivity analysis is performed in order to analyze the effect of the sphericity. The analytical modelling approach involves only the RUC model. The first step is to predict the biofilm thickness. Thereafter, two methods are suggested for predicting the biofilm affected specific surface area. The first method is based on an approach suggested in the literature. In the second method, the RUC model is used to express the pressure drop in terms of the biofilm affected specific surface area, which yields the adapted RUC model. The biofilm affected specific surface area values are then obtained by making use of experimental pressure drop data and superficial velocity values. After incorporating both the analytically and empirically determined sphericity values into the adapted RUC model, the pressure drop results show the significant effect that the sphericity value has on the model predictions. Finally, a sensitivity analysis is performed on the input parameters to the model.

AFRIKAANSE OPSOMMING : Verskeie nywerhede stel giftige waterstofsulfiedgas vry in die atmosfeer. Dit is dus noodsaaklik om die omgewingsvriendelike biofiltreringsproses, wat die vermoë het om giftige gasse, soos waterstofsulfied, vanuit die lugstroom te verwyder, te ondersoek en wiskundig te modelleer. Twee benaderings word gebruik in hierdie studie om die groei van biofilm in drie verskillende biofilters wiskundig te modelleer, d.i. ’n empiriese en ’n analitiese benadering. Die empiriese benadering behels die gebruik van die drukvalvoorspellings van die Gewysigde-Macdonald vergelyking, die bestaande korrelagtige reghoekige Representative Unit Cell (RUC) model en die model van Comiti en Renaud om die verandering in biofilm-geaffekteerde porositeit, spesifieke oppervlak-area, tortuositeit en biofilm dikte te bepaal. Die resultate is verkry deur gebruik te maak van Excel R Solver, wat gebaseer is op ’n optimeringsmetode. Daarna volg ’n sensitiwiteitsanalise op die effek van die bolvormigheid. Die analities gemodelleerde benadering is op slegs die RUC model gebaseer. Die eerste stap is om die dikte van die biofilm te voorspel. Daarna word twee metodes voorgestel vir die voorspelling van die biofilm-geaffekteerde spesifieke oppervlak-area. Die eerste metode is gebaseer op ’n benadering wat in die literatuur voorgestel word. Die tweede metode behels die gebruik van die RUC model om die drukval in terme van die biofilm-geaffekteerde spesifieke oppervlak-area uit te druk, wat die aangepaste RUC model produseer. Die biofilm-geaffekteerde spesifieke oppervlak-area waardes word dan bepaal deur gebruik te maak van die eksperimentele drukval data, asook die snelheidswaardes. Na die inkorporering van beide die analities- en empiries-bepaalde bolvormigheidswaardes in die aangepaste RUC model opgeneem is, wys die drukval resultate die aansienlike effek wat die bolvormigheidswaarde op die modelvoorspellings het. Ten slotte, word ’n sensitiwiteitsanalise op die invoer-parameters tot die model uitgevoer.

Please refer to this item in SUNScholar by using the following persistent URL: http://hdl.handle.net/10019.1/102693
This item appears in the following collections: