Modelagem matemática do processo de oxidação química na remediação de solo tropical contaminado por tolueno
Data
2020-10-15
Autores
Ribeiro, Gabriel Ferreira 
Leme, Isabela Pinheiro
Orientadores
Freitas, Juliana Gardenalli de
Tipo
Trabalho de conclusão de curso
Título da Revista
ISSN da Revista
Título de Volume
Resumo
A utilização da técnica de Oxidação Química In Situ (ISCO) para remediação de áreas
contaminadas vem ganhando cada vez mais importância, contudo a obtenção de dados para
a otimização do emprego desta técnica pode ser muito trabalhosa em estudos experimentais.
Uma forma de auxiliar os estudos de bancada é por meio simulações computacionais. Este
trabalho teve por objetivo entender o funcionamento da ISCO em solos tropicais contaminados
pelo LNAPL utilizando o persulfato como oxidante, por meio de um modelo matemático
numérico (BIONAPL/3D) a partir de estudos realizados em bancada. Nestes estudos
experimentais Latossolo Vermelho foi compactado buscando manter as caracteríscas
próximas do natural encontradas na área de coleta, após isso foi simulado derramamento do
LNAPL (tolueno + heptano) e posteriormente realizada a injeção do oxidante. Obteve-se uma
boa aproximação para simulação da interação do oxidante com o solo na ausência do
contaminante. Com relação as características do meio, o modelo mostra-se pouco responsivo
à algumas características do solo como densidade aparente e dispersividade transversal
horizontal e vertical. Por outro lado, o modelo mostra-se muito responsivo à condutividade
hidráulica e dispersividade longitudinal. Estimou-se pelo modelo simulado um valor um pouco
superior de dispersividade longitudinal, isso pode ocorrer pois a dispersividade é um
parâmetro diretamente ligado ao grau de heterogeneidade e anisotropia do solo, assim o fato
do modelo considerar um solo perfeitamente homogêneo pode influenciar nesse resultado.
Com relação aos parâmetros de interação do persulfato com o solo, as plumas que mais se
adequavam ao modelo experimental indicavam uma alta interação com o solo por meio dos
coeficientes de decaimento e retardamento. Em relação ao contaminante, como o modelo não
considera a migração em fase livre, optou-se por posicionar e dimensionar a área fonte de
LNAPL no local onde foi verificada a ocorrência de concentrações acima da solubilidade
efetiva de contaminante, posterior a infiltração. A pluma de contaminante do modelo simulado
apresenta um formato uniforme, diferente do experimental, ocorrendo provavelmente devido
à homogeneidade do solo na simulação. Quanto aos parâmetros que exercem grande
influência no contaminante, o modelo apresentou alta sensibilidade em relação ao coeficiente
de distribuição. Para a reação de degradação do tolueno, não foi possível atingir uma boa
calibração, porém, acredita-se que melhores aproximações pudessem ser estimadas com um
maior número de simulações. O modelo possui um potencial para aplicações práticas, para
otimizar o processo da ISCO, no entanto é fundamental a determinação das propriedades do
campo.
The use of the In Situ Chemical Oxidation (ISCO) technique for the remediation of contaminated areas is gaining more and more importance, however, obtaining data for the use of this technique can be very laborious in experimental studies. One way to assist bench studies is through computer simulations, this work aimed to understand the functioning of ISCO in tropical soils contaminated by LNAPL using persulfate as an oxidizer, through a numerical mathematical model (BIONAPL/3D) based on bench studies. In these experimental studies Red Latosol was compacted in order to keep the characteristics close to the natural ones found in the collection area, after which LNAPL (toluene + heptane) spillage was simulated and the oxidant injection was subsequently performed. A good approximation was obtained when simulating the interaction of the oxidant with the soil in the absence of the contaminant. Regarding the characteristics of the environment, the model shows little responsiveness to some characteristics of the soil and environment such as apparent density, horizontal and vertical dispersiveness, which is consistent with reality. On the other hand, the model is very responsive to hydraulic conductivity and longitudinal dispersivity. A slightly higher value of longitudinal dispersivity was estimated by the simulated model compared to the experimental model, this may occur because the dispersivity is a parameter directly linked to the degree of heterogeneity and anisotropy of the soil, thus the fact that the model considers a perfectly homogeneous soil can influence that outcome. Regarding the parameters of interaction of persulfate with the soil, the plumes that best suited the experimental model indicated a high interaction with the soil through the decay and retardation coefficients, showing good sensitivity of the model in relation to the decay and delay parameters. Regarding the contaminant, it was observed that the model does not consider migration in the free phase, so it was decided to position and dimension the LNAPL source area in the place where the occurrence of concentrations above the effective contaminant solubility was verified, after infiltration. In addition, it is also noted that the contaminant plume of the simulated model has a uniform shape, probably occurring due to the homogeneity of the soil. As for the parameters that exert a great influence on the contaminant, the model presented a good approximation in relation to the distribution coefficient. For the degradation reaction, the Monod parameters were estimated taking into account the concentration of the LNAPL, but values that better represented the simulated to the experimental model were not reached, it is believed that better approximations could be estimated with a greater number of simulations. The model has a potential for practical applications, to optimize the ISCO process, however it is essential to determine the properties of the soil.
The use of the In Situ Chemical Oxidation (ISCO) technique for the remediation of contaminated areas is gaining more and more importance, however, obtaining data for the use of this technique can be very laborious in experimental studies. One way to assist bench studies is through computer simulations, this work aimed to understand the functioning of ISCO in tropical soils contaminated by LNAPL using persulfate as an oxidizer, through a numerical mathematical model (BIONAPL/3D) based on bench studies. In these experimental studies Red Latosol was compacted in order to keep the characteristics close to the natural ones found in the collection area, after which LNAPL (toluene + heptane) spillage was simulated and the oxidant injection was subsequently performed. A good approximation was obtained when simulating the interaction of the oxidant with the soil in the absence of the contaminant. Regarding the characteristics of the environment, the model shows little responsiveness to some characteristics of the soil and environment such as apparent density, horizontal and vertical dispersiveness, which is consistent with reality. On the other hand, the model is very responsive to hydraulic conductivity and longitudinal dispersivity. A slightly higher value of longitudinal dispersivity was estimated by the simulated model compared to the experimental model, this may occur because the dispersivity is a parameter directly linked to the degree of heterogeneity and anisotropy of the soil, thus the fact that the model considers a perfectly homogeneous soil can influence that outcome. Regarding the parameters of interaction of persulfate with the soil, the plumes that best suited the experimental model indicated a high interaction with the soil through the decay and retardation coefficients, showing good sensitivity of the model in relation to the decay and delay parameters. Regarding the contaminant, it was observed that the model does not consider migration in the free phase, so it was decided to position and dimension the LNAPL source area in the place where the occurrence of concentrations above the effective contaminant solubility was verified, after infiltration. In addition, it is also noted that the contaminant plume of the simulated model has a uniform shape, probably occurring due to the homogeneity of the soil. As for the parameters that exert a great influence on the contaminant, the model presented a good approximation in relation to the distribution coefficient. For the degradation reaction, the Monod parameters were estimated taking into account the concentration of the LNAPL, but values that better represented the simulated to the experimental model were not reached, it is believed that better approximations could be estimated with a greater number of simulations. The model has a potential for practical applications, to optimize the ISCO process, however it is essential to determine the properties of the soil.