Pilotforsøk med fargefjerning i nedstrøms tomedia-filter og kartlegging av optimale koaguleringsbetingelser og filterhastighet

Abstract

Masteroppgaven er initiert av ABV og Oslo VAV, med et ønske om å optimalisere et fremtidig koagulering/kontaktfiltreringsanlegg for fargefjerning på Kattås vannbehandlingsanlegg i Bærum. Hovedmålet med oppgaven var å kartlegge optimale koaguleringsbetingelser (pH og koagulantdose) og filterhastighet ved å kjøre forsøk på et pilotanlegg med råvann fra Holsfjorden. Pilotanlegget består av to filterkolonner, begge nedstrøms to-lags filtre. Filtermediet til filter 1 bestod av Filtralite® NC 0,8 – 1,6 mm (topp) og Filtralite® HC 0,5 – 1,0 mm (bunn). Filtermediet til filter 2 bestod Filtralite® NC 0,8 – 1,6 mm (topp) og Rådasand® 0,8 – 1,0 mm (bunn). Filterkolonnene ble kjørt parallelt. Det ble tilsatt en fellingspolymer, SUPERFLOC® A-100 PWG (0,05 mg/L). Denne var konstant gjennom hele pilotkjøringen. Piloten ble systematisk kjørt med ulike PAX-18 doseringer (1,0 – 2,5 mg Al/L) ved ulike pHnivåer. Vannprøver ble tatt for analysering av farge, turbiditet, pH og Al-rest. Filter 1 ga generelt bedre kvalitet enn filter 2, trolig på grunn av mer finkornet filtermasse. Optimal dose og pH ble funnet til å være 1,2 mg Al/L for begge filtrene, for mer robust drift og lavest driftskostnader. Videre ble det sett på økning i filterhastighet, for de optimale koaguleringsbetingelsene som ble funnet. Hastighet utprøvd var 4,5, 7,0 og 9,0 m/h. For filterhastigheter over 4,5 m/h ble filtersyklusen halvert for samtlige doser. Optimal hastighet ble funnet til være 4,5 m/h. På grunn av dårlige filtersykluser, må ny type og/eller dose polymer vurderes og testes ut i piloten. Dette lå utenfor oppgaven.

This master thesis was initiated by Asker, Bærum and Oslo municipality, in order to optimize a future coagulation/contact filtration treatment plant for removing color in raw water from Holsfjorden. The main objective of this thesis was to identify optimal coagualtion conditions (pH and dosage) and filtration rate for raw water by pilot scale investigations. The pilot plant consists of two filter columns, both of them downstream two-layered filters. Filtermedia 1 consisted of Filtralite® NC 0,8 – 1,6 mm (top) and Filtralite® HC 0,5 – 1,0 mm (bottom). Filtermedia 2 consisted of Filtralite® NC 0,8 – 1,0 mm (top) and Rådasand® 0,8 – 1,0 mm (bottom). The filter columns were run parallelly. In order to help the coagulation process, polymer SUPERFLOC® A-100 PWG (0,05 mg/L) was added. The dosage was constant throughout the pilot scale investigation. The coagulant tested was PAX-18, which is a prepolymerized aluminium based coagulant. The raw water was systematically dosed with different dosages (1,0 – 2,5 mg Al/L), at different pHlevels. The water samples were taken and analyzed for color, turbidity, pH and Al-residuals. Filter 1 generally showed better water quality, probably due to finer filter grains. Optimal dosage and pH was found to be 1,2 mg Al/L, for both filters. This dosage will enable a more robust operation and lower operating costs. The pilot was also tested with higher filter rates for the optimal coagulation conditions that were identified. Increased filter rates up to 4,5 7,0 and 9,0 m/h was tested. When the filter rate was increased to 7,0 m/h, the filter run was halved, for all dosages. Optimal filter rate was identified to be 4,5 m/h. Since the filter runs were shorter than expected, a new type/dosage of polymer has to be tested/optimized. This was not included in this thesis, but should be addressed in the future.