GNGTS 2015 - Atti del 34° Convegno Nazionale

GNGTS 2015 S essione 1.3 153 As SEAWAT simulates the temperature as a solute dissolved in the aquifer, is necessary to estimate the hydro-dispersive parameters: the effective porosity n e has been assumed equal to 0.20, horizontal dispersivity α L has been assumed equal to 10 (m), the ratio between horizontal and vertical dispersivity (α l ) and α v was set equal to 0.1, while the ratio between the vertical dispersivity (α v ) and α L was set equal to 0.01 for each simulation. The thermodynamic parameters required for the calculation are: the dry bluk density ρ b , the molecular diffusion coefficient K d , assimilable to the heat diffusion coefficient, and then the thermal diffusivity α. Based on the literature, for the shallow aquifer under consideration, were assumed the following parameters: ρ b was set equal to 2000 kg/m 3 , K d was set equal to 10 -7 l/mg and α was set equal to 0.20 m 2 /die. Sensitivityanalysis of the thermodynamicparameters carried in similar contexts (Piccinini et al. , 2012) have shown that as the heat transfer is a process mainly advective-diffusive, the progressive increase of two orders of magnitude of α not induce significant changes in temperature, instead an increase of n e over 60% leads to a decrease of 0.5 C for the temperature while, a variation of K d of an order of magnitude changes the velocity of rebalances of the system at the end of the activity. In conclusion, the most significant parameters for dimensioning the distance between the extraction and re-injection wells are K d and n e , while, α may be considered negligible. Materials and methods. Thermal characteristics of the aquifer. Through the introduction of a sensor installed inside the piezometer, at 25 meters of depth from ground level, the temperature data of the aquifer have been recorded. Starting from December 2011, temperature data are collected periodically, three times a day every eight hours (4:00,12:00 and 20:00) (Fig. 1C). The monitoring is still ongoing. The analysis of data temperature collected during the monitoring period December 2011- December 2014, shows that the temperature ranges around a mean value of 14.6 C. The recorded temperature are contained within the value 0.60 C. This average temperature value is in agreement with recent studies on temperature distribution in the subsoil finalized to check the “homoeothermic surface” with its relative temperature value, within the Quaternary fluvial deposits hosting a shallow aquifer. The study conducted in Turin city and its hinterland (Barbero et al ., 2015) and in the surrounding plain sector (Barbero et al. , 2014) show an average temperature value of <T> (14.56 ±0.40) C and <T> 14.00 ±0.60 C respectively. SEAWAT code. Simulations of flow and heat transport have been performed in order to evaluate the environmental thermal impact within the aquifer product by the re-injection of water used for heat exchange cycle. The simulations were performed using SEAWAT, a three-dimensional finite-difference computer code developed by the US Geological Survey for the modeling of the flow of variable density in saturated porous media. This code combines the capabilities of two existing codes: MODFLOW and MT3DMS, useful for the simulation of water flow of variable density and solute transport multi-species and heat respectively. SEAWAT (Version 4) is able to simulate the transport of heat and consider the variations in the density of the fluid as a function of the concentration of solute and temperature (Thorne et al. , 2006). For the purpose of geothermal modeling, it’s necessary to determine both hydraulic and thermal parameters of the aquifer. The values of hydraulic conductivity and storage were obtained by pumping tests, described in the preceding paragraph instead, other parameters of difficult experimental testing as conductivity, heat capacity, effective porosity and dispersivity were taken from literature. The hydrogeological and thermal parameters, have been implemented in SEAWAT for the realization, in the short and in long period, of some scenarios regarding the expansion of thermal bubble. From the practical point of view, the results of the simulations in the short period can be used to optimize the distance between the extraction and re-injection wells during the