| For protection of groundwater resources,remediation of groundwater contamination,safety assessment of geological nuclear waste storage,and so on,Those fields often need to highly accuracy prediction simulation to groundwater and contaminant migration process in large space and scale.The main problem is how to characterize aquifer heterogeneity,which is also a important problem in hydrogeological study.To efficiently gain information of the spatial distribution of hydraulic parameters,theX-ray Computed tomography technology(CT) and the geophysical tomography concept has recently been employed Hydraulic tomography(HT).HT is a method that fully screened wells are segregated into many vertical intervals using packers. Water is pumped from or injected into an aquifer at one of the intervals to create a steady flow condition.Hydraulic head responses of the aquifer at other intervals are then monitored, yielding a set of head/discharge (or recharge) data. By sequentially pumping (or injecting) water at one interval and monitoring the steady state head response at others,many head/discharge (or recharge) data sets are obtained.The spatial distribution of hydraulic parameters can be identified by using Successive Linear Estimator (SLE) and head/discharge (or recharge) data sets.HT basing on SLE can highly accuracy characterize aquifer heterogeneity in numerical pumping test and sandbox pumping test,which has been confirmed.However.it has not widely used in practical problems.The reasons are that the number of pumping test are greatly and design of pumping test has a great influence on HT in the practical application.How to obtain pumping test optimal design and highly accuracy characterize aquifer heterogeneity is an urgent issue.It is lack of effective quantitative or qualitative criterion to guide HT pumping test optimal design.In this paper,we develop a series of numerical pumping test and sandbox pumping test to study HT characterizing aquifer heterogeneity pumping test optimal design,which are composed of observation well interval,pumping well interval,pumping flow,number of pumping tests and samping times of selecting water data.ln this paper we develop the following research:1.Optimal design of pumping test basing on numerical pumping test of heterogeneous aquifer.Through the computer to build a long,wide and high as160cm,10cm,80cm two-dimensional cross-section of heterogeneous aquifer,the initial head as100cm,bottom and upper boundary is impermeable boundary.left and right boundary is constant head boundary(head is100cm).Different of observation well interval,pumping well interval, pumping flow,number of pumping tests and samping times of selecting water level data are designed to complete pumping test optimal design,and guide sandbox pumping test optimal design.2. Optimal design of pumping test basing on sandbox pumping test of heterogeneous aquifer.(1)Build the laboratory sandbox heterogeneous aquifer, the aquifer model and numerical simulation of the same size, the initial head as48cm,bottom boundary is impermeable boundary, left and right boundary,and the upper boundary is constant head boundary(head is48cm), arrangement of nine wells,each well has four pumping or observation period,a total of36pumping or observation period.(2)Completing of36times constant flow pumping test (Q1=4.89cm3/s),and choosing four pumping well increase four different flow pumping test.All the observation wells are stable in the10s.(3)Choosing different of observation well interval,pumping well interval, pumping flow,number of pumping tests and using steady data in the water level curve and model of steady flow study different pumping test factors how to impact on the estimated hydraulic conductivity (K).(4)Choosing different samping times in the water level curve and using model of transient flow study different samping times how to impact on the estimated K and storage coefficient(S).We come to the following conclusion after above studies:1.The observation well interval,pumping well interval optimal design associated with horizontal (X)and vertical(Z) correlation scales of K.(1) With the decreasing of observation well interval and the increase of the number of observation well,the accuracy of characterizing K are improved.(2)As observation well interval do not change,with the increase of pumping well interval,the accuracy of characterizing K are improved.2.Pumping rate optimal design associated with the signal-to-noise ratio(SNR).The SNR is defined as the ratio between the maximum drawdown of the de-noised hydrograph and the standard deviation of noise.When the number and location of pumping well and observation well are fixed,pumping rate have not influence on estimation of spatial variances of K in numerial pumping test because of observation water level have not noise. In sandbox pumping test,with the increase of pumping rate results to SNR of observation well become big,the results of characterizing K become better and better in sandbox pumping test.However pumping rate are greater than a certain extent,precision of estimation of spatial variances of K with increase of pumping rate.3.Optimal design of the number of pumping test associated with horizontal(X) and vertical(Z) correlation scales of K,and reciprocity principle in pumping test.The existence of the reciprocity principle shows that with the increase of pumping test groups,the new water level information are less and less.Thus accuracy of parameters also have influence.The analysis results show that increase the number of pumping test can effectively improve the accuracy of estimated K for a given well field.However,reciprocity principle reduces the improve efficiency of estimated K filed through increase the number of pumping test.For different well field,the mumber of pumping test have same number when the accuracy of estimated K have same.For two-dimensional heterogeneous aquifer numerical pumping test, as long as two of three lines between pumping wells are not parallel to the direction of correlation scales of K, three pumping tests are enough to characterize the heterogeneity of hydraulic conductivity for a given well field. Increasing the number of pumping test cannot significantly increase the accuracy of estimates of K after three pumping tests.For two-dimensional heterogeneous aquifer sandbox pumping test, the optimal number pumping tests are three to six,and increasing the number of pumping test cannot significantly increase the accuracy of estimates of K after three pumping tests..The reason is that the influence of observation water level noise and constant head boundary reduce the SNR of observation water level data,which affect of accuracy of estimates of K in the near of constant head boundary.4.Using different samping times observation drawdown data estimate spatial variances of K and storage coefficient(S).The results show pumping times optimal design of selecting drawdown data.(1)For numerical pumping test,the accuracy of estimated K was improved with increase of samping times in the water level drawdown curve.What’s more estimated K using steady flow time water level is best.The accuracy of estimated K can not be improved when pumping time are greater than a certain extent time in numerial pumping test. If samping times are small than to(when the straight line portion of the plot is extrapolated to the point where the drawdown is zero according Cooper-Jacob’s approach),the accuracy of estimated S was improved with increase of samping times in pumping test. If the time of selecting drawdown data are greater than to,the accuracy of estimated S was reduced with increase of samping times of selecting drawdown data.(2)For sandbox pumping test,using different samping times data simultaneously estimate K and S.For far away from the boundary conditions fields of K,with the increase of samping times,the accuracy of estimated K are improved. For neighbouring from the boundary conditions fields of K,the accuracy of estimated K are poor.If fields of K know and use steady flow observation water level estimated K,using different samping times data estimate S and samping times are greater than to,the accuracy of estimated S was reduced with increase of samping times of selecting water level data.Through HT numerical pumping test and sandbox pumping test optimal design,get the following innovation:(1)For two-dimensional heterogeneous aquifers,the optimal number pumping tests are three to six groups because of the influence of that reciprocity principle reduces the improve efficiency of estimated K filed through increase the number of pumping test.(2)Using steady flow observation water level estimated K is best,and using to observation water level estimated S is best through numerical and sandbox pumping test. |
PDF Full Text Request