| With the increasing demand for groundwater resources development,artificial groundwater recharge,and groundwater-surface water source replacement,stricter requirements have been put forward for water resources and groundwater protection in aquifers,especially in confined aquifers.Currently,scientific evaluation and management of groundwater resources is an effective way to solve the above problems,and accurate acquisition of hydrogeological parameters of aquifers is an important prerequisite for groundwater resources and environmental impact assessment.Pumping test is one of the important ways to accurately obtain hydrogeological parameters and conduct quantitative research on groundwater.It is widely used in geology,environment,water conservancy and other related industries because of its simple operation and clear specifications.Therefore,the research on the theoretical model of the pumping test and the parameter estimation method is of great significance for perfecting and establishing the theory of well flow dynamics.The literature review found that although the mechanisms of how the drawdown changes during a pumping test have been widely investigated theoretically and experimentally,the existing classical models and test specifications are based on the common assumptions.These assumptions include the pumping rate being constant.However,in the actual field operations,variability has been observed due the complex characteristics of aquifers,the difference of test equipment and the influence of human operation.These factors create inconsistences that do not agree with the classical models.Hence,extensive scientific problems still exist that need further study:1)Construction and solution of the theoretical model of variable rate pumping test considering the influence of aquifer characteristics.The existing variable rate model mainly discusses the pumping rate form,while the influence on the aquifer characteristics that may cause the variable rate phenomenon,such as well storage,skin effect,leakage and other factors,remains to be considered;2)Improvement of the method for estimation of aquifer and well loss parameters in step-drawdown pumping test.The traditional step-drawdown pumping test is forced to test in a penetrating well and requires the pumping rate increases step by step.The above two points limit its field application,so it is urgent to improve the test method;3)A new method for interpretation of aquifer parameters.The variable rate pumping test model is more complex than the classical model,and the traditional parameter inversion method is no longer applicable.It is necessary to propose a new method of parameter inversion based on the variable rate pumping process.Also,the classical theories mainly aimed at interpreting the aquifer permeability coefficient and storage coefficient,whereas high-precision aquifer characterization should also derive other characteristic parameters of the aquifer.Considering the foregoing,this work relies on the existing theoretical models of variable-rate pumping experiments,along with actual variable rate pumping test cases in the field.In addition,we constructed three forms of variable-rate pumping experimental theoretical models and proposes a new parameter inversion method to reveal the regularities of how variable rate and its inductions affect the drawdowns.Main research contents and conclusions are as follows:（1）Established the linearly decayed and exponentially decayed rate pumping test mod,taking the well storge effect,skin effect and leakage into consideration,respectively.The quantitative description index of the effects of the variable rate process was proposed to reveal the mechanism of drawdown change during a variable rate pumping test.The linear decayed and exponentially decayed rate pumping tests will cause a"three-stage"fluctuation phenomenon.The drawdown increases first,then decreases for a while,and then increases again at the early stage of the pumping test,which shows significant differences with those in the constant rate pumping case.Apart from this,the drawdown at the initial stage of the pumping test considering the well storage effect is smaller than that without considering the well storage;different skin characteristics will increase or decrease the drawdown near the pumping well while showing little effect on drawdown far away from the well;The existing leakage shows little influence on drawdown at the beginning of the pumping test,but it will reduce the drawdown and make the water level tend to be stable in the later stage of the pumping test.（2）Considering the strict regulation of flow form and constraints of testing in a penetrating well for a step-drawdown test.This thesis proposes a nonlinear growth step-drawdown test in a partially penetrating well based on the traditional step-drawdown test.The four consecutive incremental pumping rates and a penetrating well are required by the traditional step-drawdown test,which significantly limits the application in field work.Therefore,this thesis expands the flow transformation form of each step and extends the application conditions from a penetrating well to a partially penetrating well.The influence of pumping rate variation and the well structure is clarified.The results show that:the parameters accuracy would effectively improve when introducing the pumping rate difference coefficient and well structure coefficient to quantitatively characterize the well loss difference caused by irregular pumping rate changes and well screen length.For example,the new method was applied in XY001 borehole,and the calculated drawdown can match the observed drawdown splendidly,in which RMSE was 0.12.It could effectively decrease the well loss by reducing the length of the well screen appropriately in a well with a good water yield property.In a word,the improvement of the form of the step-drawdown test and the extension to a partially penetrating well can provide quantitative theoretical support for the design and test process of the pumping well in the future.（3）In view of multiple parameters inversion problem associated with theoretical model of variable-rate pumping test,a new parameter estimation method using Particle swarm optimization（PSO）was proposed.Similarly,the influence of different particle swarm numbers on the accuracy of results was also evaluated.Since the variable-rate pumping test theoretical model considers many hydrogeological parameters such as skin-related parameters,leakage-related parameters,well loss parameters,etc.,it is difficult to inverse all parameters at the same time by traditional analytical or numerical methods.Hence,we combine the drawdown semi-analytical solution in the Laplace domain caused by a variable rate pumping test with the modified PSO method to estimate the aquifer parameters.The Root Mean Square Error（RMSE）was selected as the objective function to indicate the matching degree of the calculated drawdown and observed drawdown.For instance,the exponentially decayed rate pumping test observed in YLW01 was applied for parameters estimation by PSO method based on the considering multi-factor exponential decayed rate pumping test theoretical model.The results show that:the conductivity of the aquifer is 0.33m/d,storage coefficient is 9.69×10-3 and the RMSE was reduced from 0.14 to 0.032 compared to previous work which indicate that the PSO method shows great advantages.Besides,the initial particle numbers between 40 to 80could all obtain a high precision result.In summary,this thesis mainly focuses on constructing three different variable rate forms pumping test theoretical models and the new parameter inversion method.Including single drawdown and step-drawdown tests,linear decayed rates,exponential decayed rates,and step-drawdown in a partially penetrating well,respectively.Analytical calculation and field test was used to deeply investigate the influence on drawdown by a variable pumping rate and the feasibility of the new parameters estimation method.The research results in this thesis enrich the theoretical model on the variable rate pumping tests and provide a new insight for the research of parameter inversion method. |
PDF Full Text Request