數值模擬探討井際分溶示蹤試驗估計非水相液體飽和度之誤差
Date Issued
2004
Date
2004
Author(s)
李基毓
DOI
zh-TW
Abstract
According to the basic assumptions brought up by Jin et al. (1995), which use the breakthrough curves of each tracer got by the partitioning interwell tracer tests (PITT) while going with moment method to estimate residual NAPL saturation, building 2-D vertical sandbox simulation with finite difference method. The main purpose is to search the restriction of estimating residual NAPL saturation by using the method of direct problem in a heterogenenous medium.
The construction of this model can be mainly divided into three parts: the steady flow field, transportation model, and transportation model with partitioning effect. Then, compare and examine the analytical solutions with the numerical solutions by the use of simple 1-D and 2-D cases under the homogeneous medium condition. From the results, when the , the numerical solutions of the model is very close to the analytical solutions. Although the model is not sensitive to , from the results of comparing numerical solutions and analytical solutions, is a better choice. From the breakthrough curve of each tracer at the pumping well which derived from tracer transportation simulation, and take the 0th and 1st moment, then we will get the average NAPL residual saturation in different heterogenenous sandbox, while the tracer recovery criteria is also different. Moreover, comparing the residual NAPL saturation got by simulating with the real one, we can finally evaluate the application of this method.
The study shows that, when the aquifer can satisfy the basic assumptions: (1) the retardation effect is totally caused by NAPL, and the adsorption/desorption effect of soil can be neglected; (2) the partition coefficient and tracer concentration has no relation with the composition of NAPL; (3) the assumption of 1st order equilibrium has established; (4) groundwater flow can have effective contact with all the residual NAPL oil globs; and the full recovery of all tracers; we can get exact estimation under any heterogeneous situation by using this method. The worse recovery rate of each tracer, the less response from the information of tracers passing by residual NAPL oil globs to breakthrough curves, and the worse result of residual NAPL saturation estimation will have. In addition, if the tracers can’t be completely recovered, the larger heterogeneity is, and the closer distance between the location of heterogeneity variation and injection well is, the estimation of residual NAPL saturation will become worse.
Subjects
延滯現象
地下水
含水層
殘餘量
數值模擬
動差法
分溶
異質性
PITT
NAPL
Type
thesis