Thesis submitted in the requirements for the degree
Doctor Technologiae: Chemical Engineering
in the Faculty of
Engineering
at the
CAPE PENINSULA UNIVERSITY OF TECHNOLOGY, 2013 / To date, limited information has been published on textile wastewater treatment, for re-use, in South Africa (SA), with treatment processes focusing on conventional wastewater treatment methods. A large contributor to the contamination of water within textile industries is from dyehouse processes. A major concern in textile wastewater treatment is the release of azo dyes and their metabolites, some of which are carcinogenic and mutanogenic, into the environment since they are xenobiotic and aerobically recalcitrant to biodegradation. A necessity therefore exists to find an effective treatment method capable of removing both the strong colour and the toxic organic compounds from textile wastewater. Membrane bioreactors (MBRs) are favoured when treating high-strength wastewater, since the membrane area is determined by the hydraulic throughput and not the biological load; no sludge is wasted and all bacteria are retained within the reactor, including specific bacteria capable of degrading the toxic, non-biodegradable constituents present in textile wastewater. MBR systems, using various configurations have been utilised extensively in the rest of the world to treat textile wastewater at both lab and pilot-scale.
This DTech project formed part of a collaborative Water Research Commission (WRC) funded project K5/1900 - Pilot application of a dual-stage membrane bioreactor (dsMBR) for industrial wastewater treatment. The main purpose of this study was the on-site evaluation of a pilot-scale dsMBR incorporating two ultrafiltration (UF) sidestream membrane modules for the treatment, recovery and re-use of textile wastewater.
The objectives of this project were to determine the treatment efficiency of the system; to evaluate the degree of colour removal from the textile wastewater; to improve residual colour removal within the system using treatment processes, such as NF and RO, as well as to propose a design and cost for a full-scale plant.
A textile industry located in Bellville, Western Cape, was chosen as the industrial partner for the on-site evaluation of a semi-automated pilot wastewater treatment MBR plant using two 5.1 m2 Norit X-flow AirliftTM membrane modules. Since the wastewater treatment system was located on the premises, real continuously changing industrial wastewater was being treated. The industrial textile wastewater was treated in a series of tanks: 1) an anaerobic tank, which cleaved the azo bonds of the reactive dyes; 2) an anoxic tank containing reduced amounts of dissolved oxygen, in which denitrification occurred; and 3) an aerobic tank, in which i) nitrification, as well as ii) mineralisation of the aromatic amines occurred. The UF-membrane modules would account for the removal of any organic material.
The wastewater stream was characterised by a chemical oxygen demand (COD) range of between 45 to 2,820 mg/L and an average biological oxygen demand (BOD) of 192.5 mg/L. The dsMBR achieved an average COD reduction of 75% with a maximum of 97% over the 220 day test period. The COD concentration obtained after dsMBR treatment averaged at 191 mg/L, which was well within the City of Cape Town industrial wastewater discharge standard. The average reduction in turbidity and TSS was 94% and 19.6%, respectively, during the UF-MBR stage of the system. Subsequent treatment of the UF permeate with nanofiltration (NF) for 4 days, alternated with reverse osmosis (RO) for 14 days removed both the residual colour and salt present in the UF permeate. A consistent reduction in the colour of the incoming wastewater was evident. The colour in the wastewater was reduced from an average of 659 ADMI units to ~12 ADMI units in the NF permeate, a lower American dye manufacturing index (ADMI) (i.e. method of colour representation) compared to the potable water (~17 ADMI units) utilised by the industrial partner in their dyeing processes. The colour was reduced from an average of 659 to ~20 ADMI units in the RO permeate, a lower ADMI and therefore colour when compared to the potable water. An average conductivity rejection of 91% was achieved with conductivity being reduced from an average of 7,700 to 693 μS/cm and the TDS reduced from an average of 5,700 to 473 mg/L, which facilitated an average TDS rejection of 92%. Based on the composition of the UF permeate fed to the RO membrane a maximum removal of 98.7% was achieved for both conductivity and total dissolved solids (TDS).
The proposed full-scale plant would incorporate a UF-MBR system, followed by NF, RO, flocculation and a filter press. Therefore, the two waste products produced during operation of the proposed full-scale plant, would be the solid filter cakes and the liquid filtrate from the filter press. Implementing the proposed full-scale plant it would cost the industrial partner an operating cost of ZAR 113.85 and ZAR 3,415.49 to treat 97.1 m3 and 2,913 m3 of textile wastewater, respectively, per day and per month. This results in an annual saving of ZAR 845,848 on potable water expenses.
This research, would provide SA textile industries, with an option to: 1) reduce their water consumption, thereby utilising less of a valuable decreasing commodity; 2) meet the SA government discharge standards and reduce their discharge costs; 3) reduce their carbon footprint (i.e. reduce their impact on the environment) by re-using their treated wastewater and therefore using less water from the municipality; and 4) decrease their annual expenditure on water, since the treated wastewater would be available for re-use.
| Identifer | oai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:cput/oai:localhost:20.500.11838/926 |
| Date | January 2013 |
| Creators | De Jager, Debbie |
| Publisher | Cape Peninsula University of Technology |
| Source Sets | South African National ETD Portal |
| Language | English |
| Detected Language | English |
| Type | Thesis |
| Rights | http://creativecommons.org/licenses/by-nc-sa/3.0/za/ |
Page generated in 0.0032 seconds