Investigating the performance of a novel Anaerobic Sequencing Batch Reactor (AnSBR) and optimisation of operational parameters to treat synthetic winery wastewater

Date
2016-03
Journal Title
Journal ISSN
Volume Title
Publisher
Stellenbosch : Stellenbosch University
Abstract
ENGLISH ABSTRACT: Currently the South African agricultural sector withdraws approximately 70% of the freshwater resources, which is of great concern considering the forecast increase in water withdrawals due to the expected increase in required food production. The wine industry, in particular, produces large amounts of wastewater to produce 1 L of wine. This winery wastewater legally requires treatment before it can be disposed of into the environment or be sent to municipal treatment plants. Treatment is required for winery wastewater due to the high volumes of wastewater (seasonal production) and due to the winery wastewater’s chemical characteristics (high COD concentrations, high salinity and sodicity, high nutrient content, acidic pH and polyphenol concentrations) which are detrimental to the environment if discharged. Treated wastewater should rather be used for irrigation to partially relieve the strain on freshwater withdrawals by the agricultural industry. Re-using treated wastewater is an advantageous alternative to ensure freshwater sustainability. Research and development are still ongoing to develop more cost effective winery wastewater (WW) treatment options. Anaerobic digestion (AD) is a compatible and successful treatment option to treat this highly biodegradable winery wastewater. Only some AD processes have, however, been successfully applied in the wine industry. One specific process, the Anaerobic Sequencing Batch Reactor (AnSBR) is still under development, but has potential due to it’s kinetic advantage (alternating F:M ratios (food to microorganisms)), flexible mode of operation which is advantageous for varying wastewater production volumes (batch system (shorter feeding time) or fed-batch system (semi-continuous, longer feeding time) and due to it having good control systems, the effluent can be withdrawn from the process when it complies with legislation parameters. This process therefore requires optimisation at a laboratory scale as the limited number of operational parameters in literature vary greatly. Therefore, the main objective of this study was to investigate the feasibility of the AnSBR to specifically treat synthetic winery wastewater and to optimise the operational parameters that affect the efficiency of the process. The study was sub-divided into different Experimental Phases (A – D). The aim of Phase A and C was to investigate the feasibility of the novel AnSBR (14.7 L) to treat a synthetic winery wastewater substrate at COD’s ranging from 1 000 – 7 000 mg.L-1. At a COD of 4 000 mg.L-1 (Phase A) a 88% COD reduction occurred, the alkalinity was approximately at 3 000 mgCaCO3.L-1, the total Volatile Fatty Acid (VFA) concentration at 125 mg.L-1 and the pH was effectively controlled at ca. 7.2 by the 2 M KOH chemical dosing system. At a COD of 7 000 mg.L-1 (Phase C) an 80% COD reduction occurred, the alkalinity was approximately at 3 000 mgCaCO.L-1 3, the total VFA’s = 500 mg.L-1 and the pH was effectively controlled at ca. 7.2. A central composite experimental design (CCD) was used to optimise three independent parameters at a fivelevel design at COD = 4 000 mg.L-1 (Phase B) and COD = 7 000 mg.L-1 (Phase D). The parameters investigated were: �􀬵 = pH; �􀬶 = feeding time (shorter feeding time (batch process) vs. longer feeding time (fed-batch process)) and �􀬷 = mixing frequency (mixing set for 10 s every �􀬷 min). The data was used to fit a regression coefficient model from which response surface methodology (RSM) plots were drawn to indicate an interaction effect with regard to a specific efficiency measurement. The optimisation of the three parameters was measured according to these efficiency measurements: 1) COD reduction; 2) Total Suspended Solids (TSS) content of the effluent; 3) VFA:Alkalinity ratio; 4) Methane percentage; and 5) Polyphenol reduction percentage. The results of Phases A and C showed that the AnSBR was capable of efficiently treating SWW at COD’s up to 7 000 mg.L-1. The optimisation studies (Phase B and D) showed that the optimal operational conditions at an influent COD = 4 000 mg.L-1 and 7 000 mg.L-1, would be an influent pH of ca. 6.7 – 7.3 (pH = 7.3), a longer feeding time (possibly 240 min) and a less frequent mixing (10 s every ca. 110 min).
AFRIKAANSE OPSOMMING: Tans onttrek die landbou sektor ongeveer 70% van die vars waterbronne, wat ‘n groot bekommernis is aangesien ‘n verhoging in water onttrekking voorspel word vir die toekoms weens die verwagte verhoging in voedsel produksie. Die wynindustrie spesifiek produseer groot hoeveelhede wynkelderuitvloeisels om 1 L wyn te produseer. Hierdie wynkelderuitvloeisels vereis behandeling voordat dit in die omgewing vrygelaat word of gestuur word na munisipale behandelingsfasiliteite. Behandeling is ‘n vereiste vir die wynkelderuitvloeisel weens die groot volumes wat geproduseer word (seisoenale produksie) en weens die chemiese eienskappe van die wynkelderuitvloeisel (hoë konsentrasies chemiese suurstof vereiste (CSV), hoë sout en natrium gehaltes, hoë nutriënt inhoud, suur pH en die konsentrasie polifenole) wat nadelig is vir die omgewing as dit ontslaan word. Behandelde afvalwater moet eerder gebruik word vir besproeiingswater om die landboubedryf se druk op die varswaterbronne gedeeltelik te verlig. Die gebruik van behandelde afvalwater is ‘n alternatiewe metode om die varswater volhoubaarheid vir die toekoms te verseker. Navorsing en ontwikkeling gaan steeds voort om meer koste-effektiewe afvalwater behandelingsopsies te ontwikkel. Anaerobiese vertering (AV) is ‘n aanvaarbare en suksesvolle behandelingsopsie om hierdie hoogs bioafbreekbare wynkelderuitvloeisel te behandel. Slegs sekere AV prosesse is suksesvol tot die industrie toegepas. Een spesifieke proses, die Anaerobiese Opeenvolgende Lot Reaktor (AOLR) is steeds in ontwikkeling, maar toon die potensiaal deur sy kinetiese voordeel (afwisselende K:M verhouding (kos tot mikro-organismes)), buigsame wyse van werking wat ‘n voordeel is vir die varieërde volumes afvalwater produksie (lot proses (korter voer tyd) vs. semi-lot proses (semi-deurlopend, langer voer tyd) en deurdat dit goeie kontrole sisteme in plek het deurdat die behandelde uitvloeisel eers van die AOLR onttrek kan word wanneer dit aan die regulasie parameters voldoen. Die proses op ‘n laboratorium skaal vereis optimisering aangesien die beperkte hoeveelheid operasionele parameters grootliks in die literatuur varieër. Die hoofdoel van hierdie studie was om die herhaalbaarheid van die AOLR om spesifiek sintetiese wynkelderuitvloeisel te behandel te ondersoek, en ook die operasionele parameters wat die effektiwiteit van die proses affekteer te optimiseer. Die studie is onderverdeel in verskillende Eksperimentele Fases (A – D). Die doel van Fase A en C was om die herhaalbaarheid van die nuut-ontwerpte AOLR (14.7 L) te ondersoek om ‘n sintetiese wynkelderuitvloeisel substraat te behandel by ‘n CSV reeks van 1 000 – 7 000 mg.L-1. By ‘n CSV van 4 000 mg.L-1 (Fase A) het ‘n 88% CSV reduksie plaasgevind, die alkaliniteit was ongeveer by 3 000 mgCaCO3.L-1, die vlugtige vetsure konsentrasie was by 125 mg.L-1 en die pH is effektief beheer by ongeveer 7.2 deur die 2 M KOH chemiese dosering stelsel. Fase C (CSV = 7 000 mg.L-1) het ‘n resultaat van ‘n 80% COD reduksie aangetoon, die alkaliteit was ongeveer by 3 000 mgCaCO3.L-1, die totale vlugtige vetsure was by 500 mg.L-1 en die pH was effektief beheer by ongeveer 7.2. ‘n Sentrale saamgestelde eksperimentele ontwerp (SSO) is gebruik om die drie onafhanklike parameters te optimiseer op ‘n vyf-vlak ontwerp by CSV = 4 000 mg.L-1 (Fase B) en CSV = 7 000 mg.L-1 (Fase D). Die drie parameters wat ondersoek is: �􀬵 = pH; �􀬶 = voer tyd (korter voer tyd (lot proses) vs. langer voer tyd (semi-lot proses)) en �􀬷 = vermenging frekwensie (vermenging elke 10 s elke �􀬷 min). Die data is gebruik om in die regressie koëffisiënt model te pas waaruit reaksie oppervlak metodologie (ROM) grafieke getrek is om die interaksie effek waar te neem met betrekking tot ‘n spesifieke effektiwiteitsmeting. Die optimisering van die drie parameters is gemeet volgens hierdie effektiwiteitsmetings: 1) CSV reduksie; 2) Totale gesuspendeerde partikels inhoud van die behandelde water; 3) Vlugtige vetsure:Alkaliniteit verhouding; 4) Metaan persentasie; en 5) Polifenole vermindering persentasie. Die resultate van Fases A en C toon dat dit moontlik was vir die AOLR om effektief sintetiese wynkelderuitvloeisel by CSV’s tot by 7 000 mg.L-1 te behandel. Die optimiserings studies (Fase B en D) toon dat die optimale operasionele kondisies by ‘n CSV = 4 000 mg.L-1 en 7 000 mg.L-1 sal soos volg wees: ‘n pH van omtrent 6.7 – 7.3 (pH = 7.3), ‘n langer voer tyd (vermoedelik 240 min) en ‘n minder vermening frekwensie (10 s omtrent elke 110 min).
Description
Thesis (MSc Food Sc)--Stellenbosch University, 2016.
Keywords
Water -- Purification -- South Africa, UCTD, Wine-industry -- Waste disposal -- South Africa, Sewage -- Purification -- Anaerobic treatment -- South Africa
Citation