| Active heat tracing technology is an effective method to measure the magnitude and direction of groundwater velocity.At present,there are many design schemes about active heat tracer velocimetry devices,which had often been used in shallow saturated sediments and deep aquifers.The principle of active heat tracing technology is to change the temperature at a certain position by electric heating or cold/warm water injection.Due to the groundwater flow,the heat transfer process obviously deviates from the temperature diffusion under heat conductivity condition,so the temperature changes with time at the location of heat injection or nearby in a predictive way.At present,distributed temperature sensing systems(DTS)with active heating devices are usually used to measure the temperature in deep aquifers.However,DTS cost relatively high and often ignore the interference on the flow field due to cable installation.In addition,the number of optical fibers surrounding the heating cable is relatively small,which brings great uncertainty to cable installation.For groundwater flow in shallow saturated sediments,the heat pulse probe has been widely concerned,which has relatively high accuracy and short measurement period However,during installation,the probe is buried in sediment,which limits its application in deep aquifers.Considering the drawbacks stated above for the existing active heat tracer methods in deep aquifer,a novel recyclable and packaged porous probe for measuring the velocity distribution in deep aquifers is proposed.The configuration of the designed probe is that a solid cylindrical heater locates at the center vertically,and a few temperature sensors are evenly distributed around the heater.The whole probe is filled with glass beads,which were wrapped with fine wire mesh.The probe can be placed in the observation well coaxially at a measured depth.The main content of this thesis can be summarized as follows:First,a fluid flow model with such a configuration is proposed.In the fluid flow model,Brinkman models are applied in all of the porous media,and Stokes model in the pure water region.By coupling the velocity and stress at the interfaces between different regions,an analytical solution for the fluid flow in the whole domain can be derived,which is then substituted into the energy governing equation for solving the temperature distribution numerically in the saturated porous media.The finite difference method is used for the numerical solution.The sensitivity analysis is performed for several physical properties and the geometric parameter.Finally,a experimental setup for measuring the velocity in saturated porous media was established in the laboratory.According to the experimental and simulated data obtained,several conclusions following can be drawn:A sensitivity analysis was conducted theoretically,the results show that the water ring gap,internal permeability and heater volume have a greater influence on the velocity distribution than the aquifer permeability for the probe.However,in a heat transfer point of view,the heating duration and largeness,infinite aquifer velocity and the thermal conductivity of filled glass beads have a greater influence on the temperature distribution in the probe than the effective thermal conductivity of aquifer.It also shows that in general the physical properties,such as the permeability and thermal conductivity of the aquifer,will not result in a large error.While the probe geometry such as the heater and the assembled probe sizes,however,may significantly affect the measured temperature distribution.But,these parameters can be designed or calibrated in the laboratory.In addition,the accurate measurement of the borehole diameter is also important for the designed probe.A comparison was made between the measured temperatures with those numerical simulated data at the same fluid flow and heat transfer conditions.The results show that and simulated thermal response curves fit very well with the measured data under different conditions.Furthermore,four evaluation indicators are introduced to quantitatively analyze the fitting degree between the simulated thermal response curves and the measured data.The four evaluation indicators used include the root mean square error(RMSE),mean absolute error(MAE),determination coefficient(R2)and energy relative error(ERE),which give 0.013～0.38 ℃,0.01～0.38 ℃,-1.94～0.999 and0.22 %～31.3 %,respectively,in the present analysis.It can be concluded that the theoretical model has a good performance to analyze and interpret the measured data. |
