Caracterização e Tratamento de Efluentes Vinícolas da Região Demarcada do Douro
Ficheiros
Data
2005
Autores
Pirra, António José Duque
Título da revista
ISSN da revista
Título do Volume
Editora
Resumo
Reconhecendo o valor universal e excepcional da paisagem vitícola, a Região Demarcada do Douro (RDD) foi recentemente inscrita na lista de Património Mundial da UNESCO como paisagem cultural evolutiva e viva. Nesta região domina a fileira vitivinícola, particularmente a produção de Vinho do Porto, tendo-se assistido nas últimas décadas à concentração da elaboração em grandes adegas, com o consequente aumento do impacte ambiental dos Efluentes Vinícolas (EVs) gerados. Esta agro-indústria tem actividades sazonais ao longo do ano, produzindo cada uma delas efluentes com diferentes características qualitativas e quantitativas, em função do modo de laboração, do tipo de vinho produzido, do consumo de água, da tecnologia utilizada e da dimensão da instalação.
Com este trabalho pretendeu apresentar-se a problemática da produção e gestão dos EVs, a suacaracterização, processos de tratamento utilizados, aprofundando o estudo da tratabilidade aeróbia dos EVs da RDD, através da adopção de diversas estratégias para o planeamento da adega e optimização do seu tratamento. Contemplou cinco fases principais:
i) Revisão bibliográfica e avaliação do “estado da arte” relativa ao tratamento de EVs;
ii) Caracterização das instalações produtivas, avaliação do volume diário de EVs gerados e caracterização físico-química dos efluentes produzidos ao longo do ano;
iii) Estudos laboratoriais de tratabilidade aeróbia para os diferentes EVs produzidos;
iv) Ensaios utilizando Reactores Descontínuos Sequenciais (SBRs) laboratoriais;
v) Modelação dos resultados dos ensaios experimentais realizados;
A primeira fase, mais vasta, enquadra o tema abordando os diversos aspectos da produção, gestão e tratamento de EVs. Compreende o estado da arte dos principais aspectos de Enquadramento (Parte I), Redução do volume e carga poluente (Parte II), Problemática e bases tecnológicas dos tratamentos (Parte III) e Fileiras de tratamento (Parte IV), tendo cada uma delas diversos capítulos. Na segunda fase (Parte V), dá-se início ao trabalho experimental, caracterizando-se as adegas da RDD, com informação qualitativa e quantitativa dos EVs produzidos. Esta fase é fundamental ao estabelecimento de modelos de gestão de EVs e à selecção das tecnologias de tratamento. Na terceira fase (Parte VI), e dada a grande variabilidade de concentração e tipologia de matéria oxidável, expressa como CQO, avalia-se a biodegradabilidade dos diferentes tipos de efluentes produzidos (1ª, 2ª e 3ª trasfegas) e de um efluente resultante da mistura das três trasfegas (Mix). Na quarta fase (Parte VII) avalia-se o comportamento dos Reactores Descontínuos Sequenciais (SBRs) no tratamento destes efluentes, enquanto na quinta fase (Parte VIII) se comparam os resultados experimentais obtidos com os modelos matemáticos propostos por diversos autores. Finalmente, na Parte IX apresentam-se as conclusões finais e as propostas de trabalho futuro.
O trabalho realizado mostrou que existem dois períodos distintos no que respeita a consumo de água/volume de EVs produzido nas adegas do Douro: uma época baixa (de Novembro a Agosto) com consumos reduzidos e uma época alta correspondendo ao período de vindimas/primeiras trasfegas (Setembro e Outubro), período em que os consumos são multiplicados por 2 a 5 (o consumo neste período corresponde a cerca de 35 a 50% do total anual). A produção específica de efluentes oscilou entre 1,45 e 2,58 litros de água por litro de vinho produzido, sendo inversamente proporcional ao tamanho da adega. Constatou-se também a existência de uma elevada sazonalidade e heterogeneidade dos EVs produzidos em termos de carga poluente, teor em SST e pH, com existência de maiores carga poluente e teor de sólidos em suspensão na fase da vindima e primeiras trasfegas. O estudo realizado permitiu tipificar três efluentes distintos (1ª, 2ª e 3ª trasfegas) com características físico-químicas dispares, correspondentes às operações mais poluentes do ciclo produtivo do vinho, tendo no entanto todos eles uma razão CBO5/CQO de cerca de 50 a 55%, uma razão CQOsolúvel/CQOtotal de cerca de 73% e um pH em geral ácido (inferior a 5).
Nos testes descontínuos de biodegradabilidade sem adição de nutrientes, verificou-se que os efluentes da 2ª e 3ª trasfegas apresentam globalmente uma biodegradabilidade semelhante, mas inferior à da 1ª trasfega. As percentagens de remoção mais elevadas foram obtidas para baixas concentrações de biomassa (2 a 3 g SSV L-1) e de substrato (1 a 4 g CQO L-1). Não se verificaram diferenças significativas de tratabilidade biológica entre as 3 diferentes trasfegas e o efluente Mix (obtido por mistura das 3 trasfegas). Conclui-se também que os EVs apresentam um efeito inibidor a altas concentrações de substrato, pois a máxima taxa de remoção de CQO diminuiu com o aumento da concentração de CQO inicial. A percentagem de remoção em CQO nos ensaios realizados com o efluente Mix foi superior a 70% após as primeiras 24 horas, ultrapassando os 90% decorridas 48 horas após o início da depuração. De forma geral, o teor de biomassa de 2 g SSV L-1 apresentou o melhor desempenho em termos de remoção de CQO e características de decantabilidade dos sólidos biológicos formados.
A utilização de reactores SBR no tratamento de EVs permitiu concluir que a remoção de CQO obtida sem adição de nutrientes e utilizando uma razão de enchimento de 50%, oscila entre os 70 e os 95%, alcançando-se os melhores resultados para Cargas volúmicas (Cvs) inferiores a 3,4 g CQO L-1 dia-1. A Cv removida oscila entre 1,5 e 2,8 g CQO L-1 dia-1 para Cv aplicada respectivamente de 1,5 a 6 g CQO L-1 dia-1. A maior velocidade de remoção e as melhores características de decantabilidade foram obtidas no tratamento de EVs com 12 g CQO L-1 e 2 g SSV L-1 de biomassa, a que corresponde uma Cv de 6 g CQO L-1 dia-1. Assim, pode concluir-se que o tratamento aeróbio de EVs através de reactores SBR sem adição de nutrientes mostra-se tecnicamente possível, permitindo percentagens de remoção de CQO superiores a 95%, mas para Cvs aplicadas baixas. A operação de reactores SBR após adição de nutrientes e utilizando uma razão de enchimento de 25%, permite praticamente duplicar a Cv removida face à Cv aplicada, aumentando simultaneamente a velocidade de sedimentação. Houve ainda uma melhoria das características do efluente tratado (diminuição dos teores de CQOs e de SST). Em geral a remoção de CQO foi superior a 95% em todos os ensaios, sendo a maior parte da carga orgânica degradada nas primeiras 21 horas de reacção (para valores de Cv aplicada até 3,5 g CQO L-1 dia-1 e razões F/M até 1,8 g CQO g SSV-1 dia-1). São portanto estas as condições ideais para operar um reactor SBR no tratamento de EVs da RDD.
A determinação das constantes cinéticas e a modelação dos ensaios realizados evidenciou que, de forma geral, todos os EVs seguem cinéticas de degradação típicas de Haldane/Andrews ou de Monod. O efluente da 1ª trasfega apresentou valores de Ks entre 5 e 38 g CQO L-1 e Vmax de 8 a 49 g CQO g SSV-1 dia-1, enquanto os efluentes da 2ª e 3ª trasfegas e Mix apresentaram valores de Ks entre 0,01 e 17,6 g CQO L-1 e Vmax de 0,75 a 30 g CQO g SSV-1 dia-1.
Em termos globais, a modelação do comportamento gráfico dos reactores confirma que as velocidades específicas máximas de degradação da matéria orgânica observadas são conseguidas maioritariamente para as concentração de biomassa de 2 e 3 g SSV L-1 e para teores de CQO iniciais no reactor entre 1 e 4 g L-1. As velocidades específicas máximas de remoção de substrato observadas através da modelação do comportamento gráfico dos reactores oscilaram entre 0,5 e 2,8 g CQO g SSV-1dia-1, para teores de CQO iniciais no reactor entre 1 e 4 g L-1 e teores de biomassa de 2 e 3 g SSV L-1.
Recognizing the universal and exceptional value of the wine growing landscape of the RegiãoDemarcada do Douro (RDD), this was recently enrolled in the World heritage of UNESCO asevolutionary live cultural landscape. In this area the dominant activity and in many areas the only one economically possible in the territory, is wine production, particularly Port wine. In the last decades the wine production was concentrated in great wine cellars, with the consequent increase of the environmental impact of Winery Effluents (WEs). This industry has different activities during the year, producing effluents with different qualitative and quantitative characteristics, depending on the elaboration processes, type of wine, water consumption, technology and dimension of the installation. The rationalization of resources and costs, the actuality, the urgency and the competition accomplish a systematic study of the production and treatment of WEs, centred in these region and their serious environmental problems, so that we know the conditions under which the liquid emissions of the wine cellars are generated and investigate their treatment conditions. The main goal of this PhD dissertation consists of presenting a global vision of the problem of the production and administration of the WEs, particularly with respect to the characterization, planning of the wine cellar and treatment processes widely used, the characterization and the study of the aerobic treatability of the WEs of RDD, through the adoption of several strategies for optimizing his treatment. It contemplates five sub-goals: i) Bibliographical Revision and evaluation of the "state of the art" on the treatment of WEs; ii) Characterization of the productive facilities, evaluation of the daily volumes of produced WEs, physicochemical characterization of the effluents produced during the year; iii) Laboratorial studies of aerobic treatability for the different WEs produced; iv) Laboratorial scale Sequential Discontinuous Reactors (SBRs) rehearsals; v) Mathematical modelling applied to the results obtained with the experimental rehearsals; vi) Conclusions and proposes for future work. The first sub-goal provides an approach to several aspects of the production, administration and treatment of WEs. It presents the state of the art of the main aspects of Framing of the theme (Part I), Reduction of the volume and pollutant load (Part II), Problems and technological bases of the treatments (Part III) and treatment rows (Part IV), each one of them having several chapters. The second sub-goal (Part V) starts the experimental work and the characterization of RDD, to give the wine industry base information, necessary to understand how WEs are generated, their characteristics and volumes. This information is fundamental in the establishment of models of administration of WEs and in the selection of the treatment technologies according to the capacities and typologies of the wine cellars. In the third sub-goal (Part VI), due to the great concentration variability and typology of oxidizable matter (expressed as COD), the biodegradability of the different types of effluents produced was evaluated (1st, 2nd and 3rd rackings), and also an effluent resulted from the mixture of the three rackings (Mix effluent). The fourth sub-goal (Part VII) evaluates the behaviour of Sequential Discontinuous Reactor (SBR) in the treatment of these effluents under different conditions. The fifth sub-goal (Part VIII) compares the experimental results obtained with the mathematical models proposed by several authors. Finally, in the Part IX final conclusions and some proposals for future work are presented. The work demonstrated that there are two different periods concerning water consumption/winery effluents (WEs) volume produced at the RDD wine cellars: a low season (from November to August) with reduced consumptions, and a high season corresponding to vintage/first rackings (September and October). In this period the consumptions are multiplied by a factor from 2 to 5, corresponding about 35 to 50% of the total annual consumption. The specific production of effluents oscillated between 1.45 and 2.58 litters of water by litter of produced wine, and was inversely proportional to the size of the wine cellar. Concerning pollutant load, TSS content and pH, a high seasonality and heterogeneity of the WEs production was observed. The effluents with higher pollutant load and suspension solids were produced in the vintage/first rackings season. The accomplished study allowed the identification of three typical effluents (1st, 2nd and 3rd rackings) with very different physical-chemical characteristics, corresponding to the highest pollutant operations of the wine productive cycle. However all of them had similar BOD5/COD reason (50 to 55%), a soluble COD/total COD reason about 73% and are generally acid (pH below 5). Through the accomplishment of discontinuous biodegradability tests without the addition of nutrients, it was verified that globally the effluents of the 2nd and 3rd rackings presented a similar biodegradability, but lower than the 1st racking. The higher removal percentages were obtained for low biomass concentrations (2 and 3 g TSS L-1) and 1 to 4 g COD L-1 of substratum. No great biological treatability differences were verified between the 3 different rackings and the Mix effluent (obtained by mixture of the 3 rackings). These rehearsals also lead to the conclusion that the WEs present an inhibiting effect with high substratum concentrations, because the COD removal rates decreased with the increase in the initial COD concentration. The COD removal obtained in the rehearsals accomplished with the Mix effluent was higher than 70% after the first 24 hours, rising higher than 90% 48 hours after the beginning of the depuration process. In a general way, the best performance terms of COD removal and sedimentation characteristics of the biological solids was obtained with 2 g SSV L-1 biomass content. The utilization of SBR reactors in the treatment of WEs without addition of nutrients, using a 50% filling ratio, lead to the conclusion that the COD removal rates oscillated between 70 and 95%, reaching the best results for volumetric loads (VLs) lower than 3,4 g COD L-1 day-1. The VLs removed oscillated between 1.5 and 2.8 g COD L-1 day-1 for VLs applied respectively from 1.5 to 6 g COD L-1 day-1. The higher removal rates and the best settling characteristics were obtained treating WEs with 12 g COD L-1 and 2 g TSS L-1 biomass content, that corresponds to a VL of 6 g CQO L-1 day-1. In conclusion, the aerobic treatment of WEs using SBR reactors without nutrients addition is technically possible, allowing COD removals higher than 95%, but for low VLs. After the nutrients addition and using a 25% filling ratio, the SBR reactors practically double VLs removed related to VLs applied, increasing simultaneously the settling speed. An improvement in the characteristics of the treated effluent was also verified, respecting to COD and TSS content. Generally, the COD removal was higher than 95% for all of the rehearsals, and the organic load mostly degraded in the first 21 hours of reaction (for a VL applied up to 3.5 g COD L-1 day-1 and F/M reasons up to 1.8 g COD g VSS-1 day-1). Therefore, these are the ideal conditions to operate an SBR reactor treating WEs. The kinetic constants determinations and the modelling of the accomplished rehearsals showed that, in a general way all of the effluents follow typical Haldane/Andrews or Monod degradation kinetics. The treatment of the wine-producing effluent from the 1st Racking showed Vmax values from 8 to 49 g COD g VSS-1 day-1 and Ks values from 5 to 38 g COD L-1. The treatment of 2nd and 3rd Rackings and Mix effluent showed Vmax values from 0.75 to 30 g COD g VSS-1 day-1 and Ks values from 0.01 to 17.6 g CQO L-1. Globally, modelling the reactors graphic behaviour confirms that the specific maximum organic matter degradation speeds are mainly observed for 2 and 3 g SSV L-1 biomass concentration, and for initial COD content in the reactor between 1 and 4 g L-1. Specific maximum substratum removal speed observed through the modelling of the graphic behaviour of the reactor oscillated between 0.5 and 2.8 g COD g VSS-1 day-1, for initial COD content in the reactor between 1 and 4 g L-1 and a biomass content between 2 and 3 g SSV L-1.
Recognizing the universal and exceptional value of the wine growing landscape of the RegiãoDemarcada do Douro (RDD), this was recently enrolled in the World heritage of UNESCO asevolutionary live cultural landscape. In this area the dominant activity and in many areas the only one economically possible in the territory, is wine production, particularly Port wine. In the last decades the wine production was concentrated in great wine cellars, with the consequent increase of the environmental impact of Winery Effluents (WEs). This industry has different activities during the year, producing effluents with different qualitative and quantitative characteristics, depending on the elaboration processes, type of wine, water consumption, technology and dimension of the installation. The rationalization of resources and costs, the actuality, the urgency and the competition accomplish a systematic study of the production and treatment of WEs, centred in these region and their serious environmental problems, so that we know the conditions under which the liquid emissions of the wine cellars are generated and investigate their treatment conditions. The main goal of this PhD dissertation consists of presenting a global vision of the problem of the production and administration of the WEs, particularly with respect to the characterization, planning of the wine cellar and treatment processes widely used, the characterization and the study of the aerobic treatability of the WEs of RDD, through the adoption of several strategies for optimizing his treatment. It contemplates five sub-goals: i) Bibliographical Revision and evaluation of the "state of the art" on the treatment of WEs; ii) Characterization of the productive facilities, evaluation of the daily volumes of produced WEs, physicochemical characterization of the effluents produced during the year; iii) Laboratorial studies of aerobic treatability for the different WEs produced; iv) Laboratorial scale Sequential Discontinuous Reactors (SBRs) rehearsals; v) Mathematical modelling applied to the results obtained with the experimental rehearsals; vi) Conclusions and proposes for future work. The first sub-goal provides an approach to several aspects of the production, administration and treatment of WEs. It presents the state of the art of the main aspects of Framing of the theme (Part I), Reduction of the volume and pollutant load (Part II), Problems and technological bases of the treatments (Part III) and treatment rows (Part IV), each one of them having several chapters. The second sub-goal (Part V) starts the experimental work and the characterization of RDD, to give the wine industry base information, necessary to understand how WEs are generated, their characteristics and volumes. This information is fundamental in the establishment of models of administration of WEs and in the selection of the treatment technologies according to the capacities and typologies of the wine cellars. In the third sub-goal (Part VI), due to the great concentration variability and typology of oxidizable matter (expressed as COD), the biodegradability of the different types of effluents produced was evaluated (1st, 2nd and 3rd rackings), and also an effluent resulted from the mixture of the three rackings (Mix effluent). The fourth sub-goal (Part VII) evaluates the behaviour of Sequential Discontinuous Reactor (SBR) in the treatment of these effluents under different conditions. The fifth sub-goal (Part VIII) compares the experimental results obtained with the mathematical models proposed by several authors. Finally, in the Part IX final conclusions and some proposals for future work are presented. The work demonstrated that there are two different periods concerning water consumption/winery effluents (WEs) volume produced at the RDD wine cellars: a low season (from November to August) with reduced consumptions, and a high season corresponding to vintage/first rackings (September and October). In this period the consumptions are multiplied by a factor from 2 to 5, corresponding about 35 to 50% of the total annual consumption. The specific production of effluents oscillated between 1.45 and 2.58 litters of water by litter of produced wine, and was inversely proportional to the size of the wine cellar. Concerning pollutant load, TSS content and pH, a high seasonality and heterogeneity of the WEs production was observed. The effluents with higher pollutant load and suspension solids were produced in the vintage/first rackings season. The accomplished study allowed the identification of three typical effluents (1st, 2nd and 3rd rackings) with very different physical-chemical characteristics, corresponding to the highest pollutant operations of the wine productive cycle. However all of them had similar BOD5/COD reason (50 to 55%), a soluble COD/total COD reason about 73% and are generally acid (pH below 5). Through the accomplishment of discontinuous biodegradability tests without the addition of nutrients, it was verified that globally the effluents of the 2nd and 3rd rackings presented a similar biodegradability, but lower than the 1st racking. The higher removal percentages were obtained for low biomass concentrations (2 and 3 g TSS L-1) and 1 to 4 g COD L-1 of substratum. No great biological treatability differences were verified between the 3 different rackings and the Mix effluent (obtained by mixture of the 3 rackings). These rehearsals also lead to the conclusion that the WEs present an inhibiting effect with high substratum concentrations, because the COD removal rates decreased with the increase in the initial COD concentration. The COD removal obtained in the rehearsals accomplished with the Mix effluent was higher than 70% after the first 24 hours, rising higher than 90% 48 hours after the beginning of the depuration process. In a general way, the best performance terms of COD removal and sedimentation characteristics of the biological solids was obtained with 2 g SSV L-1 biomass content. The utilization of SBR reactors in the treatment of WEs without addition of nutrients, using a 50% filling ratio, lead to the conclusion that the COD removal rates oscillated between 70 and 95%, reaching the best results for volumetric loads (VLs) lower than 3,4 g COD L-1 day-1. The VLs removed oscillated between 1.5 and 2.8 g COD L-1 day-1 for VLs applied respectively from 1.5 to 6 g COD L-1 day-1. The higher removal rates and the best settling characteristics were obtained treating WEs with 12 g COD L-1 and 2 g TSS L-1 biomass content, that corresponds to a VL of 6 g CQO L-1 day-1. In conclusion, the aerobic treatment of WEs using SBR reactors without nutrients addition is technically possible, allowing COD removals higher than 95%, but for low VLs. After the nutrients addition and using a 25% filling ratio, the SBR reactors practically double VLs removed related to VLs applied, increasing simultaneously the settling speed. An improvement in the characteristics of the treated effluent was also verified, respecting to COD and TSS content. Generally, the COD removal was higher than 95% for all of the rehearsals, and the organic load mostly degraded in the first 21 hours of reaction (for a VL applied up to 3.5 g COD L-1 day-1 and F/M reasons up to 1.8 g COD g VSS-1 day-1). Therefore, these are the ideal conditions to operate an SBR reactor treating WEs. The kinetic constants determinations and the modelling of the accomplished rehearsals showed that, in a general way all of the effluents follow typical Haldane/Andrews or Monod degradation kinetics. The treatment of the wine-producing effluent from the 1st Racking showed Vmax values from 8 to 49 g COD g VSS-1 day-1 and Ks values from 5 to 38 g COD L-1. The treatment of 2nd and 3rd Rackings and Mix effluent showed Vmax values from 0.75 to 30 g COD g VSS-1 day-1 and Ks values from 0.01 to 17.6 g CQO L-1. Globally, modelling the reactors graphic behaviour confirms that the specific maximum organic matter degradation speeds are mainly observed for 2 and 3 g SSV L-1 biomass concentration, and for initial COD content in the reactor between 1 and 4 g L-1. Specific maximum substratum removal speed observed through the modelling of the graphic behaviour of the reactor oscillated between 0.5 and 2.8 g COD g VSS-1 day-1, for initial COD content in the reactor between 1 and 4 g L-1 and a biomass content between 2 and 3 g SSV L-1.
Descrição
Tese de Doutoramento em Engenharia Agrícola, apresentada à Universidade de Trás-os-Montes e Alto Douro
Palavras-chave
Efluentes vinícolas, Região Demarcada do Douro , Tratamento aeróbio , Reactor Descontínuo Sequencial (SBR) , Modelação