Modelling of geothermal heat pump systems rarely integrates detailed surface and subsurface components. Simplifications such as neglecting the geothermal gradient, the different geological units intersected by the systems, and/or variable building loads, are common across current modeling approaches. These simplifications are particularly problematic for deep systems (e.g., >1 km). A new approach is presented to model a well doublet or a deep borehole heat exchanger (DBHE) with several heat pumps, which combines a comprehensive subsurface numerical model with codes capable of handling variable heat demand throughout the year. Groundwater flow and heat transfer are simulated with the subsurface model. Operating flow rates and number of activated heat pumps are adjusted during the simulation according to their efficiency and demand at a given time. Simulated heat production is constrained by technical and safety criteria to reflect realistic building conditions. The codes allow the simulation of cases in which the geothermal system is designed to partly meet demand, while maximizing its contribution. The thermal power production and electric consumption of the system are calculated. An illustrative example is provided for a sedimentary basin with a low geothermal gradient (~23.5 °C/km) using the Bécancour area in eastern Canada.

Selected runs with a physics-based model of surface water–groundwater interactions are used to examine in detail some numerical challenges and surprising behaviors that result from discretization, nested solution schemes, coupling, boundary condition, and other factors. Regardless of the spatial scale of the model domain (field, hillslope, catchment, …), the processes that are simulated by this class of integrated models can exhibit widely varying dynamics within and across the different subsystems comprising the land surface, the unsaturated zone, and deep groundwater formations. The presence of heterogeneities, nonlinearities, and complex boundary conditions can exacerbate numerical difficulties in resolving exchange fluxes across subsystems and lead to unexpected or undesired results, including localized numerical oscillations and an upper bound on adaptive time stepping. The need for accurate tracking of surface–subsurface exchanges and for better control of aspect ratio and mesh distortion can also influence and constrain spatial and temporal discretization choices. Finally, model performance assessments can be highly sensitive to the response variables of interest. We will illustrate some of these issues via test case simulations at large (13.66 km catchment transect) and small (450 m2 hillslope) spatial scales, run at time scales from 10 days to hundreds of years.

Study region : A 700 km(2) watershed near the town of Fox Creek (west-central Alberta, Canada), a region that has seen intensive oil and gas production since the 1950s, is investigated. Study focus : The overall objective is to improve our understanding of the hydrodynamics of this watershed. The hierarchical multi-model approach uses physically based numerical models and follows a stepwise progression, transitioning from saturated and variably saturated groundwater to integrated surface water-groundwater (SW-GW) models. This progression encompasses 2D and 3D configurations from a deep (similar to 1 km) hydrogeological multi-layer aquifer system to a shallow (45 m) coupled flow system. This differs from other multi-model studies where the focus is on comparing simple to complex models for a fixed domain. New hydrological insights for the region : The hierarchical modeling approach has led to better insights and estimates of processes including vertical recharge and lateral groundwater contributions. Moreover, it allowed us to corroborate hydraulic conductivity values and groundwater recharge despite the limited dataset. The hydraulic connections between the four non-marine bedrock units were shown to be negligible, allowing us to focus on the Paskapoo Formation. The unsaturated zone was shown to play a significant role in the model response, and accounting for land surface-subsurface dynamics was demonstrated to be critical to properly estimating hydraulic head profiles and land surface saturation patterns.

To investigate the operational challenges faced by groundwater heat pump systems (GWHPs), a field pilot study (two heat injection tests, HITs) was conducted to mimic operational processes involving pumping, heating and reinjection into the aquifer to simulate the cooling mode. This study aimed to acquire an understanding of the spatiotemporal evolution of groundwater chemistry as it circulates throughout different parts of the system (pumping well, heater inlet/outlet, injection well, monitoring well downgradient) and during different test phases (pre-test, HITs, post-test recovery). Hierarchical clustering analyses identified six clusters corresponding to different water types through space and time, which were categorised into two main conditions: natural (Na-Ca-Cl water) and perturbed (Ca-mixed-(HCO3-Cl-SO4) water). Pumping-induced mixing between shallow and deeper water, along with draw-in of dissolved oxygen (DO), induced oxidising conditions and precipitation of iron oxides in the injection wells, subsequently leading to well blockage and premature termination of both HITs. The chemical signatures of the injected warm water reached the monitoring well faster than the thermal plume, suggesting different heat and solute transport mechanisms at play. While principal component analyses showed that water mixing was the main driver of differences between natural and perturbed geochemical conditions, geochemical modelling confirmed that DO intrusion was the main driver of iron oxide precipitation, surpassing the effect of water mixing and increased temperatures. Results from this study underline the importance of conducting a comprehensive spatiotemporal geochemical assessment of the hydrogeological system prior to and throughout the lifecycle of a GWHP system to predict and manage operational risks.

Study region: A 706 km2 watershed located in a heavily industrialized region near Fox Creek, Alberta, Canada. Study focus: Petroleum exploration has increasingly disturbed boreal ecosystems in North America. Seismic lines are one of the major footprints of the petroleum industry: they form cleared linear corridors in forests and their regeneration is generally poor. For this study, field measurements of soil, vegetation, water and snow were conducted on 5 paired (seismic line and adjacent undisturbed area) sites located in lowland and upland ecosites, and a 1-D physically based hydrologic model, the Simultaneous Heat and Water Model (SHAW), was used to investigate the impacts of seismic lines on the water budget over a year. New hydrological insights for the region: Field measurements indicate that parameters, including soil water content and temperature, vary according to a complex combination of factors and are highly site-specific between seismic lines and undisturbed areas. Hydrological modeling of the two ecosites (lowland and upland) showed up to a 33 % reduction in evapotranspiration and a decrease in percolation (leading to none) below the root zone on seismic lines. Considering that there are 2300 km of seismic lines in the study area and that 36 % of the forest has been cleared for industrial activities, these results suggest that the water budget is impacted by anthropogenic lineaments. Climate change will likely intensify these impacts.

The Paskapoo Formation is the primary groundwater supply in the province of Alberta, Canada. Fluvial sandstone channels form effective aquifers, although lateral continuity and yield can vary significantly over short distances. Borehole cores and geophysical logs from adjacent non-cored boreholes drilled in the Fox Creek area, at the northern edge of the Paskapoo Formation domain, enabled identification of intervals of friable sandstone with intervening tightly cemented sandstone. The combination of specific geophysical logs reveals the presence of highly cemented sandstone, which acts as a barrier and focuses groundwater circulation within permeable sandstone carrier beds. Multi-stage carbonate diagenesis, involving dissolution and precipitation in the presence of groundwater has generated the main physical properties of the aquifer. The sandstone is composed of quartz with feldspars and fragments of volcanic and carbonate rocks. Two types of porosity are recognized: primary depositional and secondary dissolution. The latter results from significant local dissolution by acidic groundwater of feldspars, volcanic rock fragments and carbonates. Within the secondary pores, various volumes of calcite cement can be present. δ18OVPDB and δ13CVPDB ratios indicate the cement has precipitated from groundwater with HCO3− derived from the dissolution of Paleozoic carbonate particles. The correlation of petrophysical core data with geophysical logs indicates that low permeability calcite cemented beds are characterized by a significant increase in values of resistivity, density and velocity, and elevated reflectivity in acoustic televiewer logs across sub-meter to meter-thick beds. Porosity logs derived from density logs correlate well with measured core porosity and provide complete vertical porosity profiles.

A multidisciplinary environmental project was conducted between 2019 and 2024 to assess the potential impacts of industrial activities and, in particular, oil and gas (O&G) development, in the Fox Creek area (Alberta, Western Canada). Fox Creek was selected because this region has been intensively developed for hydrocarbons over the past 60 years, first for conventional resources, then for unconventional shale gas which involves hydraulic fracturing. The various components of the project included geological, shallow and deep geophysical, geomechanical, geochemical, hydrogeological, as well as air photos and remote sensing studies, which fed into each other. The study area covers 700 km2 and is mainly forested. Despite its small size, it includes over 800 O&G historical and producing wells and 2,300 km of seismic lines. The regional shallow aquifer is the Paleogene Paskapoo Formation, a fluvial unit comprising meter to decameter-thick sandstone-fill channels cut into a mudstone, siltstone and shale continental plain succession that also includes coal seams. By its very nature, this formation is highly heterogeneous and consequently exhibits a wide range of hydraulic conductivities and yields. The aim of this project was to study an industrial region holistically, from the surface to the gas-rich unit using both direct and indirect data, in order to obtain different lines of evidence for assessing cumulative effects.The variable hydraulic properties of the Paskapoo Formation are due to both its architecture and diagenesis. The latter has led to the development of preferential pathways in poorly consolidated sandstone-fill channels with irregularly distributed, meter-thick, highly calcite-cemented sandstone beds. The dissolution / cementation events have formed a framework which controls flow in conjunction with the fracture network mainly consisting of bedding planes. Data/results from water levels, yields, estimated vertical recharge, petrophysical analyses, borehole geophysics and geochemistry indicated that a significant portion of the water flowing through the watershed originates from the foothills of the Rocky Mountains, a mountain range located approximately 150 km to the southwest. The integrated results of geology, hydrogeology and geomechanics have shown that the aquifer's vulnerability to both surface and deep activities is low. Indeed, geochemistry revealed no impact on surface or groundwater quality. Nevertheless, the study of landscape evolution over time revealed that around 36% of forests have disappeared in the past few decades in the study area, while another study revealed that there may be significant differences between the various water budget components in undisturbed and disturbed (e.g. seismic lines) areas, including a decrease in aquifer recharge in disturbed areas. This project has shown that, although the quality of water resources does not seem to be affected, probably largely due to the favorable geological/hydrogeological context, human activities do have an impact on the water resource apportionment, affecting, among other things, vegetation and notably its recovery in disturbed areas.

Download This Paper Open PDF in Browser Add Paper to My Library Share: Permalink Using these links will ensure access to this page indefinitely Copy URL Copy DOI

A code was developed to couple numerical models of geothermal doublets and deep borehole heat exchangers (DBHE) with a simple model for heat pumps, responding to a monthly demand for heating and cooling a building. The objective was to compare the performance of systems repurposing one or two idle gas wells to install a DBHE or a doublet to heat a greenhouse. Both systems incorporate several heat pumps, activated or deactivated according to the building energy demand, with the fluid flow rate varied to adapt production to demand. The results suggest that, theoretically, the doublet could supply more heat than the DBHE. Specifically, the doublet could heat a 6950 m2 greenhouse and provide ~70% of the peak demand, while the DBHE would only heat a 2560 m2 greenhouse with the same coverage of peak demand. However, a preliminary cost analysis indicates that installing and operating a doublet over 30 years would cost almost twice as much as a DBHE, which should be considered in the decision-making process.

Unconventional hydrocarbon production has sparked public concerns for several years, especially regarding potential potable groundwater contamination by hydrocarbons, brines, and various chemicals related to hydraulic fracturing operations. One possible contamination mechanism is upward migration of deep-seated contaminants over large vertical distances, through preferential pathways such as leaky well casings or permeable geological faults. In New Brunswick (Canada), thermogenic hydrocarbons and brackish water were previously reported in shallow water wells, some of them located close to unconventional gas wells or to major faults, but the exact origin of these fluids remained uncertain. The objective of this paper is to determine whether the presence of these fluids is the result of migration from the deep (>1 km) hydrocarbon bearing units (via natural or anthropogenic migration pathways), or whether they rather originate within the shallow aquifer (<100 m) or from intermediate zone. Tracking fluid origin was achieved by fingerprinting compositional and isotopic values of three indicators: 1) water isotopic signature (including tritium (H-3), radiocarbon (C-14(DIC)), delta(OH2O)-O-18, delta H-2(H2O)), 2) salinity (including Na, Ca, K, SO4, Cl, Br,Sr- 87/Sr-86), and 3) hydrocarbons (compositional data and delta C-13(CH4)). These various analyses were conducted, when relevant, on samples of different matrices composing the hydrogeological system, namely shallow groundwater (12-90 m depth), shallow bedrock gas (8-131 m), and intermediate zone evaporitic rocks (173-332 m); they were compared with previously published values for deep basin brines and gases (1940-3168 m) from the hydrocarbon bearing Carboniferous Albert Formation. This unique suite of indicators, analytes and matrices allowed drawing the conclusion that thermogenic gas and high salinities present in the sampled wells were naturally occurring and originating from shallow and intermediate - zone bedrock units. Results obtained through this approach did not provide any evidence that hydrocarbon wells in this area have acted as preferential migration pathways for deep-seated fluids towards shallow aquifers.

Geothermal systems installed at intermediate depths (similar to 1-2 km) in sedimentary basins represent an attractive option to provide low-carbon heat in cold countries, even where the geothermal gradient is low, using either closed- or open loop systems combined with heat pumps. However, the installation of such systems is costly and risky due to the uncertainty associated with the geological, hydrogeological and thermal properties of the targeted unit(s). Additionally, the performance of different types of systems is seldom compared in the literature for a given geological context. This paper presents a numerical approach to readily assess the maximum energy that could be produced by different geothermal systems (a deep borehole heat exchanger (DBHE) and three types of doublets) and compare their performance for preselection purposes. Since deep formations are often poorly characterized, the sensitivity of these systems to the most impactful site properties is evaluated. For each scenario of site properties, the systems were simulated for different operation flow rates. The maximum usable flow rate is determined from simulation results. The flow rate maximizing net energy production is used for the DBHE, while the maximum flow rate ensuring safe injection pressure has been selected for the doublets. The heating power produced per length drilled and system COP are calculated for each simulation and interpolated for the maximum usable flow rate to fairly compare the systems' performance. An illustrative case using the Becancour area (eastern Canada) highlights that, even if the site stratigraphy and properties have been relatively well characterized previously, uncertainty regarding key properties significantly affects simulation results and, consequently, the choice of the geothermal system to install. Our numerical approach is intended as a decision - making aid in order to properly plan the installation of these expensive systems. [GRAPHICS] .

A multidisciplinary project was carried out in New Brunswick (eastern Canada), where hydrocarbon exploration was stopped in 2015 due to a moratorium on hydraulic fracturing, to investigate the vulnerability of shallow aquifers to industrial activities from both potential surface contamination and upward fluid migration. The study area included an active unconventional gas field (which also encompassed a potash mine in operation until late 2015) and a prospect hydrocarbon field. This project, a collaboration between two levels of government, universities and an oil and gas operator, created an unprecedented opportunity by allowing several monitoring wells to be drilled directly on gas well pads. Multiple datasets from geological, geophysical, geomechanical, hydrogeological and geochemical studies were collected and the interpretation of aquifer vulnerability resulted from the integration of their findings. The data coverage from reservoir depths to the surface provided no evidence for the presence of a natural connection between the deep units targeted by the industry and the shallow aquifers, or of contamination from the surface. The intermediate zone between the shallow aquifers and reservoirs appears to provide an effective barrier preventing upward migration of fluids. In particular, the locally ambiguous origin of methane in groundwater and the presence of salts in a few shallow monitoring wells were resolved through a sound understanding of the interacting elements of the geological - geophysical - geomechanical and the hydrogeological - geochemical components of the holistic system. Although this article focuses primarily on the hydrogeological component, highlights of the other components are presented here to give an overview of this multidisciplinary project's results and their contribution and to provide readers with a practical framework that can be used in other regions or other industrial contexts involving deep subsurface activities.

The shallow aquifer in the Fox Creek area is hosted by the Paleocene Paskapoo Formation. The formation consists of fluvial deposits with channel-filled high-energy sandstone cutting through fine-grained, low energy overbank sediments. Three internal members are recognized, these members define three hydrostratigraphic units (two aquifers versus one aquitard). In fall 2022, three boreholes were drilled and cored. The succession is slightly dominated by sandstone with subordinate fine-grained sediments and thin coal intervals. The calcareous to non-calcareous sandstone is either tight and well compacted or porous, friable to unconsolidated. The litharenite is composed of quartz, various types of rock fragments, chert, and feldspars. Detrital carbonates can be abundant. The post-sedimentation history of the sandstone recorded cementation and dissolution events from near surface, through shallow burial and late tectonic exhumation. The events include early clay coatings on grains, dissolution of metastable minerals, cementation from calcite, kaolinite and minor chlorite and late near surface fault-controlled freshwater circulation and dissolution. The late event resulted in friable to unconsolidated sandstone intervals.

In this study we push the limits of a physics-based detailed model of surface water/groundwater interactions, CATHY, in order to explore numerical issues related to discretization, coupling, and scale effects. Regardless of the spatial scale of the model domain (field, hillslope, catchment, ...), the processes that are simulated by integrated models such as CATHY are characterized by different dynamic time scales across subsystems and thus require appropriate time stepping schemes. Accurate tracking (in a mass balance sense) of complex exchange fluxes is also a challenge. At larger spatial scales, concerns related to aspect ratio and mesh distortion can influence and constrain grid discretization choices. Across the land surface boundary, different options for representing boundary conditions can lead to widely varying model behaviors. Finally, model performance assessments can be highly sensitive to the response variables of interest. We will illustrate some of these challenges via test case simulations of a long (13 km) transect and a small (0.3 ha) hillslope.

Non-intrusive techniques such as medical CT-Scan or micro-CT allow the definition of 3D connected pore networks in porous materials, such as sedimentary rocks or concrete. The definition of these networks is a key step towards the evaluation of fluid flow and heat transfer in energy resource (e.g., hydrocarbon and geothermal reservoirs) and CO2 sequestration research projects. As material heterogeneities play a role at all scales (from micro- to project-scale), numerical models represent a powerful tool for bridging the gap between small-scale measurements provided by X-ray imaging techniques and larger-scale transport properties. This study uses pre-existing medical CT-scan datasets of reference material, namely glass beads and conventional reservoir rocks (Berea sandstone, Boise sandstone, Indiana limestone) to extract the 3D geometry of connected pores using an open-source software (Spam). Pore networks from rock samples were generated from dry and then saturated samples. Binarized datasets were produced for these materials (generated by a thresholding technique) to obtain pore size distribution and tortuosity, as well as preferential paths for fluid flow. Average porosities were also calculated for comparison with those obtained by conventional commercial laboratory techniques. The results obtained show that this approach works well for medium and coarse-grained materials that do not contain a large percentage of fine particles. However, this approach does not allow representative networks to be obtained for fine-grained rocks, due to the fact that small pores (or pore throats) cannot be taken into account in the datasets obtained from the medical CT-Scan. A next step, using datasets produced from a micro- CT scan, is planned in order to be able to generate representative networks in this type of material as well.

Heating performance of deep borehole heat exchangers (DBHEs) is heavily impacted by its design and operating flow rate. Here, various designs of coaxial DBHEs were numerically simulated and compared for 1 year of continuous operation in a cold sedimentary basin using the FEFLOW software. Factors affecting performance, including: (1) depth, (2) repurposing an oil and gas (O&G) well or drilling a larger one, (3) repurposing options, (4) inner pipe material, made of either vacuum insulated tubing (VIT) or high-density polyethylene and (5) grout with different thermal conductivity were evaluated. Where an O&G well is available, the most cost-efficient option is repurposing, with the best performance obtained from the deepest and largest diameter wells. While VIT inner pipes performed better in some designs, their performance did not compensate for their cost.