[en] The objective of the work was to investigate hydraulic effects of open and poorly sealed boreholes on groundwater flow conditions through simulations using a numerical groundwater model. Specifically, the boreholes KFM07A, KFM09A, and KFM09B in Forsmark and the boreholes KLX04, KLX06, and KLX10 in Laxemar were studied. The criteria for the selection of these boreholes were that the boreholes should represent typical conditions of the site, the borehole length should exceed 500 m, and that several major fractured zones should be intersected. The boreholes KFM07A and KLX06, respectively, were selected as reference boreholes for more detailed studies of different sealing schemes. The model setup of the Forsmark model followed the Forsmark 2.2 regional-scale conceptual hydrogeological model. The model domain was approximately 15 km (north-south) x 10 km (west- east) x 1.2 km (depth). The 131 deformation zones and three layers of superficial horizontal sheet joint were modelled deterministically. A stochastic discrete fracture network (DFN) representing fractures and minor deformation zones were also generated between the deterministic deformation zones inside central model volume. The side lengths of the square fractures were from 1,000 m down to 10 m. In order to resolve the details of flow in to and out from the borehole, a more detailed DFN was generated in a zone around the borehole KFM07A, where fractures down to a side length of 0.5 m were considered. The model setup of the Laxemar model followed the SDM-Site Laxemar (Laxemar 2.3) regional scale conceptual hydrogeological model. The model domain was approximately 12 km (north-south) x 20 km (west-east) x 2.1 km (depth). A number of 71 deformation zones were modelled deterministically, and one realization of a stochastic DFN, the so-called hydrogeological DFN model base case, was imported to the model. Similar to the Forsmark case, a more detailed DFN was also generated around the reference borehole KLX06. The numerical groundwater modelling code DarcyTools was used for the simulations. Continuum hydraulic property fields for the flow simulations were generated from the deterministic deformation zones and the modelled DFN. DarcyTools has a special routine for simulation of open boreholes. A reference borehole plugging scheme and a simplified version were applied for the reference boreholes. The concept for borehole sealing included alternating sections of silica-concrete and bentonite along the borehole. In the models, appropriate values of hydraulic conductivity were assigned to the grid cells representing the studied boreholes to accommodate simulation of the borehole sealing. The hydraulic impacts on the groundwater flow conditions of the open (unsealed) and poorly sealed boreholes were investigated by steady-state simulations. No salinity and no density effects were included in the simulations. The variables that were investigated were changes in the hydraulic head and flow fields around the boreholes at repository depth, the total flow through a defined rock volume surrounding the boreholes, and the flow along the boreholes. Also, in order to study the impact on advective travel time for water and solutes between repository depth and surface, particle tracking was performed between a horizontal plane at -600 m and the -50 m level. The simulations indicated that the open boreholes have a considerable hydraulic influence, especially on hydraulic heads at large depths. There was a difference in the hydraulic function of the open boreholes when comparing the two sites studied for the present-day hydraulic boundary conditions. In Forsmark, as a discharge area for deeper groundwater, open boreholes acted as easy path ways for groundwater from repository depth to surface. In Laxemar, on the other hand, being in part a recharge area for deeper groundwater, open boreholes acted as paths from surface to depth. The open boreholes increased the groundwater turnover in the borehole site rock volume with about 17% in the Forsmark model, but the effect on the groundwater tur nover in the Laxemar model was small. This difference is due to the much lower permeability at repository depth in Forsmark compared to Laxemar, which makes the effect of the open boreholes more apparent in Forsmark. The different influence on groundwater flow is also reflected in the influence on solute transport in the surrounding bedrock, where the open boreholes had a strong influence in the Forsmark model but not in the Laxemar model. This is shown by the influence on the distributions of advective travel times and path lengths for particles