Fernando Dornelles | Universidade Federal do Rio Grande do Sul (original) (raw)
Uploads
Papers by Fernando Dornelles
RBRH, Sep 25, 2017
Implicações da utilização de equações IDF descontínuas na geração de hidrogramas de escoamento su... more Implicações da utilização de equações IDF descontínuas na geração de hidrogramas de escoamento superficial. Estudo de caso: IDF-Porto Alegre (8º DISME
Revista Brasileira de Recursos Hídricos, 2013
The use of data-driven models may be an important alternative in several scientific fields, espec... more The use of data-driven models may be an important alternative in several scientific fields, especially when the available data do not allow utilizing physical hydrologic models because these data must be measured in the basin.. This paper explores important aspects of ANN use: initial training conditions, performance assessment, partitioning of the strong seasonal component in short-term samples and ranking results by a weighted score. Sequential partitioning of the sample was shown to be adequate for cases where the data series has a strong seasonal component and short time response. The nonexceeded error was associated with its frequency, giving a measure of performance that is easily understood and which does not depend on the long familiarity required by traditional methods to evaluate results. A weighted score calculated from several indices removed the difficulty of how to reconcile several statistical measures of performance. The need for repeated artificial neural network training using random starting conditions is established, and the ideal number of repetitions to ensure good training was investigated. A straightforward approach to visualization of forecasting errors is presented, and a pseudoextrapolation region at the domain extremes is identified. The methods were explored using the Quaraí river basin, whose specific characteristics include a rapid response to precipitation events. It therefore provides a good test of artificial neural network methods, including the use of rainfall forecasts which, to be combined with existing data resources, required novel methodological approaches.
Revista Brasileira de Recursos Hídricos, 2015
O objetivo deste artigo é a avaliação da capacidade de infiltração de pavimentos permeáveis por m... more O objetivo deste artigo é a avaliação da capacidade de infiltração de pavimentos permeáveis por meio do uso da Norma ASTM (American Society for Testing and Materials) C1701-Standard Test Methods for Infiltration Rate of in Place Pervious Concrete e da metodologia do infiltrômetro de anéis concêntricos. Os pavimentos avaliados foram o asfalto poroso e o bloco de concreto intertravado vazado (modelo "S"), ambos executados em um estacionamento do Instituto de Pesquisas Hidráulicas (IPH), em 2004, localizado na cidade de Porto Alegre. Os resultados obtidos revelaram que os pavimentos apresentam colmatação, devido ao uso e a falta de manutenção, ao longo do tempo.Os resultados obtidos apresentaram valores de taxa de infiltração entre 8,9x10-6 m/s e 19x10-6 m/s para o asfalto poroso e superiores a 10-5 m/s no bloco de concreto intertravado vazado, valores estes que equivalem ao grau de permeabilidade de médio a baixo. Avaliando os valores obtidos no asfalto poroso, o bloco de concreto intertravado vazado apontou uma capacidade de infiltração superior, com grau de permeabilidade média. Evidenciou-se que o asfalto poroso no módulo experimental é o menos resistente em relação ao uso, pois, após 9 anos, este perdeu de modo significativo a sua capacidade de infiltração, diferenciando do bloco de concreto intertravados vazado. Esta redução ocorreu devido à falta de manutenção dos pavimentos, que foi propositalmente definida durante a instalação do experimento, de modo a avaliar as condições reais de uso.
Journal of Cleaner Production
Abstract Urban water systems throughout the world are under recurring and increasing water scarci... more Abstract Urban water systems throughout the world are under recurring and increasing water scarcity, given demand growth, aging infrastructure, variability and uncertainty imposed by climate change. Rainwater harvesting systems (RWHS) represent a promising alternative to increase flexibility and robustness of water supply systems. Given the importance of the tank size in establishing a reliable RWHS, we propose an implicit stochastic approach to assess performance and aid designers. The main objective of this study is to simulate common non-potable water demand and roof area typologies and verify how the demand and roof area affect the efficiency of the RWHS. To highlight the usefulness of the model to aid in the definition of tank sizes, we have also performed an economic assessment. Results are specific to the study region’s climate and might vary under different climate. However, we have chosen a rather unfavourable climate, with poor rainfall distribution along the year. Under such conditions, results are likely closer to lower bound benefits. This indicates potential water savings benefits of a RWHS even for unfavourable scenarios. The RWHS indicated to be more efficient at meeting demands with smaller Demand-Roof Area ratios. Furthermore, it was possible to determine the efficiency of the RWHS under various scenarios, identifying the minimum and maximum likely benefits. Several scenarios were compared directly allowing for the establishment of a general maximum tank size per unit roof area in a given region of interest, beyond which no significant benefits are likely. The economic assessment indicates positive net present values for investing in RWHS, especially for large rooftop areas and small Demand-Roof Area ratios. Under such conditions, investing in a RWHS has very low risk and will most likely have short payback times.
Revista Brasileira de Recursos Hídricos, 2012
Rainwater harvesting generates sewage volumes released to the public network, whose destination a... more Rainwater harvesting generates sewage volumes released to the public network, whose destination and treatment costs are not accounted for by the sewage charging scheme generally adopted in Brazil, since this collection is carried out based on the volume of potable water supplied, although the advantages of its use are known. This paper proposes a methodology for estimating the volume of sewage generated, based on the characteristics of the rainwater harvesting system and indexed by the volume of treated water measured in the meter of the consumer unit. Two examples with distinct features (efficiency and service capacity) are presented and discussed to elucidate the use of the new methodology. The conclusions indicate that the method to estimate the sewage generated volume by rainwater harvesting is coherent to estimate the volume of sewage generated by the use of rainwater and has immediate practical application since it requires no investment in any additional equipment
RESUMO --- Bacias hidrográficas com pequenas dimensões, com características de coeficiente de esc... more RESUMO --- Bacias hidrográficas com pequenas dimensões, com características de coeficiente de escoamento alto e tempo de pico curto formam cenários que dificultam a utilização de técnicas tradicionais de previsão de nível. A utilização de previsões hidrometeorologicas permite uma ampliação do tempo disponível entre a previsão e a ocorrência do nível previsto, porém acrescentam muita incerteza nos dados de entrada para os modelos. Ferramentas como Redes Neurais (RNs) e Sistemas de Inferência Difusa (SIDs) possuem capacidades de assimilação de incertezas e generalização de resultados, sendo portanto modelos indicados para estes casos. A análise, para o caso do Rio Quarai, mostrou vantagens do uso de RNs para antecedências pequenas (1 e 2 dias) e vantagens dos SIDs para as maiores antecedências (3 e 4 dias). As previsões hidrometeorológicas foram fornecidas pelo modelo ETA-CPTEC de forma numérica para uma malha de 40 x 40km, enquanto que os níveis observados provêm de um linígrafo oper...
The use of data-driven models may be an important alternative in several scientific fields, espec... more The use of data-driven models may be an important alternative in several scientific fields, especially when the available data do not allow utilizing physical hydrologic models because these data must be measured in the basin. . This paper explores important aspects of ANN use: initial training conditions, performance assessment, partitioning of the strong seasonal component in short-term samples and ranking results by a weighted score. Sequential partitioning of the sample was shown to be adequate for cases where the data series has a strong seasonal component and short time response. The nonexceeded error was associated with its frequency, giving a measure of performance that is easily understood and which does not depend on the long familiarity required by traditional methods to evaluate results. A weighted score calculated from several indices removed the difficulty of how to reconcile several statistical measures of performance. The need for repeated artificial neural network training using random starting conditions is established, and the ideal number of repetitions to ensure good training was investigated. A straightforward approach to visualization of forecasting errors is presented, and a pseudoextrapolation region at the domain extremes is identified. The methods were explored using the Quaraí river basin, whose specific characteristics include a rapid response to precipitation events. It therefore provides a good test of artificial neural network methods, including the use of rainfall forecasts which, to be combined with existing data resources, required novel methodological approaches.
RBRH, Sep 25, 2017
Implicações da utilização de equações IDF descontínuas na geração de hidrogramas de escoamento su... more Implicações da utilização de equações IDF descontínuas na geração de hidrogramas de escoamento superficial. Estudo de caso: IDF-Porto Alegre (8º DISME
Revista Brasileira de Recursos Hídricos, 2013
The use of data-driven models may be an important alternative in several scientific fields, espec... more The use of data-driven models may be an important alternative in several scientific fields, especially when the available data do not allow utilizing physical hydrologic models because these data must be measured in the basin.. This paper explores important aspects of ANN use: initial training conditions, performance assessment, partitioning of the strong seasonal component in short-term samples and ranking results by a weighted score. Sequential partitioning of the sample was shown to be adequate for cases where the data series has a strong seasonal component and short time response. The nonexceeded error was associated with its frequency, giving a measure of performance that is easily understood and which does not depend on the long familiarity required by traditional methods to evaluate results. A weighted score calculated from several indices removed the difficulty of how to reconcile several statistical measures of performance. The need for repeated artificial neural network training using random starting conditions is established, and the ideal number of repetitions to ensure good training was investigated. A straightforward approach to visualization of forecasting errors is presented, and a pseudoextrapolation region at the domain extremes is identified. The methods were explored using the Quaraí river basin, whose specific characteristics include a rapid response to precipitation events. It therefore provides a good test of artificial neural network methods, including the use of rainfall forecasts which, to be combined with existing data resources, required novel methodological approaches.
Revista Brasileira de Recursos Hídricos, 2015
O objetivo deste artigo é a avaliação da capacidade de infiltração de pavimentos permeáveis por m... more O objetivo deste artigo é a avaliação da capacidade de infiltração de pavimentos permeáveis por meio do uso da Norma ASTM (American Society for Testing and Materials) C1701-Standard Test Methods for Infiltration Rate of in Place Pervious Concrete e da metodologia do infiltrômetro de anéis concêntricos. Os pavimentos avaliados foram o asfalto poroso e o bloco de concreto intertravado vazado (modelo "S"), ambos executados em um estacionamento do Instituto de Pesquisas Hidráulicas (IPH), em 2004, localizado na cidade de Porto Alegre. Os resultados obtidos revelaram que os pavimentos apresentam colmatação, devido ao uso e a falta de manutenção, ao longo do tempo.Os resultados obtidos apresentaram valores de taxa de infiltração entre 8,9x10-6 m/s e 19x10-6 m/s para o asfalto poroso e superiores a 10-5 m/s no bloco de concreto intertravado vazado, valores estes que equivalem ao grau de permeabilidade de médio a baixo. Avaliando os valores obtidos no asfalto poroso, o bloco de concreto intertravado vazado apontou uma capacidade de infiltração superior, com grau de permeabilidade média. Evidenciou-se que o asfalto poroso no módulo experimental é o menos resistente em relação ao uso, pois, após 9 anos, este perdeu de modo significativo a sua capacidade de infiltração, diferenciando do bloco de concreto intertravados vazado. Esta redução ocorreu devido à falta de manutenção dos pavimentos, que foi propositalmente definida durante a instalação do experimento, de modo a avaliar as condições reais de uso.
Journal of Cleaner Production
Abstract Urban water systems throughout the world are under recurring and increasing water scarci... more Abstract Urban water systems throughout the world are under recurring and increasing water scarcity, given demand growth, aging infrastructure, variability and uncertainty imposed by climate change. Rainwater harvesting systems (RWHS) represent a promising alternative to increase flexibility and robustness of water supply systems. Given the importance of the tank size in establishing a reliable RWHS, we propose an implicit stochastic approach to assess performance and aid designers. The main objective of this study is to simulate common non-potable water demand and roof area typologies and verify how the demand and roof area affect the efficiency of the RWHS. To highlight the usefulness of the model to aid in the definition of tank sizes, we have also performed an economic assessment. Results are specific to the study region’s climate and might vary under different climate. However, we have chosen a rather unfavourable climate, with poor rainfall distribution along the year. Under such conditions, results are likely closer to lower bound benefits. This indicates potential water savings benefits of a RWHS even for unfavourable scenarios. The RWHS indicated to be more efficient at meeting demands with smaller Demand-Roof Area ratios. Furthermore, it was possible to determine the efficiency of the RWHS under various scenarios, identifying the minimum and maximum likely benefits. Several scenarios were compared directly allowing for the establishment of a general maximum tank size per unit roof area in a given region of interest, beyond which no significant benefits are likely. The economic assessment indicates positive net present values for investing in RWHS, especially for large rooftop areas and small Demand-Roof Area ratios. Under such conditions, investing in a RWHS has very low risk and will most likely have short payback times.
Revista Brasileira de Recursos Hídricos, 2012
Rainwater harvesting generates sewage volumes released to the public network, whose destination a... more Rainwater harvesting generates sewage volumes released to the public network, whose destination and treatment costs are not accounted for by the sewage charging scheme generally adopted in Brazil, since this collection is carried out based on the volume of potable water supplied, although the advantages of its use are known. This paper proposes a methodology for estimating the volume of sewage generated, based on the characteristics of the rainwater harvesting system and indexed by the volume of treated water measured in the meter of the consumer unit. Two examples with distinct features (efficiency and service capacity) are presented and discussed to elucidate the use of the new methodology. The conclusions indicate that the method to estimate the sewage generated volume by rainwater harvesting is coherent to estimate the volume of sewage generated by the use of rainwater and has immediate practical application since it requires no investment in any additional equipment
RESUMO --- Bacias hidrográficas com pequenas dimensões, com características de coeficiente de esc... more RESUMO --- Bacias hidrográficas com pequenas dimensões, com características de coeficiente de escoamento alto e tempo de pico curto formam cenários que dificultam a utilização de técnicas tradicionais de previsão de nível. A utilização de previsões hidrometeorologicas permite uma ampliação do tempo disponível entre a previsão e a ocorrência do nível previsto, porém acrescentam muita incerteza nos dados de entrada para os modelos. Ferramentas como Redes Neurais (RNs) e Sistemas de Inferência Difusa (SIDs) possuem capacidades de assimilação de incertezas e generalização de resultados, sendo portanto modelos indicados para estes casos. A análise, para o caso do Rio Quarai, mostrou vantagens do uso de RNs para antecedências pequenas (1 e 2 dias) e vantagens dos SIDs para as maiores antecedências (3 e 4 dias). As previsões hidrometeorológicas foram fornecidas pelo modelo ETA-CPTEC de forma numérica para uma malha de 40 x 40km, enquanto que os níveis observados provêm de um linígrafo oper...
The use of data-driven models may be an important alternative in several scientific fields, espec... more The use of data-driven models may be an important alternative in several scientific fields, especially when the available data do not allow utilizing physical hydrologic models because these data must be measured in the basin. . This paper explores important aspects of ANN use: initial training conditions, performance assessment, partitioning of the strong seasonal component in short-term samples and ranking results by a weighted score. Sequential partitioning of the sample was shown to be adequate for cases where the data series has a strong seasonal component and short time response. The nonexceeded error was associated with its frequency, giving a measure of performance that is easily understood and which does not depend on the long familiarity required by traditional methods to evaluate results. A weighted score calculated from several indices removed the difficulty of how to reconcile several statistical measures of performance. The need for repeated artificial neural network training using random starting conditions is established, and the ideal number of repetitions to ensure good training was investigated. A straightforward approach to visualization of forecasting errors is presented, and a pseudoextrapolation region at the domain extremes is identified. The methods were explored using the Quaraí river basin, whose specific characteristics include a rapid response to precipitation events. It therefore provides a good test of artificial neural network methods, including the use of rainfall forecasts which, to be combined with existing data resources, required novel methodological approaches.