Hydrology and hydrogeology

The conceptualisation of the hydrological-hydrogeological system at Forsmark is summarised in chapter 8; more detailed descriptions are given in the background reports for surface hydrology and near-surface hydrogeology /Johansson 2008/ and bedrock hydrogeology /Follin 2008/. The present section summarises results from surface system modelling activities and joint interpretations considered important for the bedrock modelling. In particular, hydrogeological properties and time-series data on groundwater levels are discussed.

As described in some detail in e.g. /Follin et al. 2007c/, the uppermost rock is a key component in the overall groundwater flow system at Forsmark. The transmissivities measured in this part of the rock are, in many cases, exceptionally high, in particular in the north-western part of the candidate area (the target area). Therefore, it was suggested that there can be a “hydraulic cage phenomenon”

caused by a lattice of high-transmissive structures consisting of extensive horizontal fractures/sheet joints. The shallow and anisotropic bedrock aquifer (see section 8.2), intercepts the recharge from above as well as the discharge from below. This implies that it has important implications for the groundwater exchange between the Quaternary deposits and the rock within the target area.

The geology of the regolith in the Forsmark area is described in /Hedenström and Sohlenius 2008/

and in section 4.2.1. Till is the dominant Quaternary deposit in the area, whereas fine-grained materials such as gyttja and glacial and post-glacial clay are found below lakes and wetlands. Based on a simplified conceptual model in which the regolith was subdivided into a limited number of

Regarding hydrogeological properties, the following should also be noted /Johansson 2008/.

• The hydraulic conductivity (K) values for till in the Quaternary deposits model /Johansson 2008/

are horizontal conductivities obtained from slug tests. The only vertical K-values from till are from the laboratory permeameter tests, which indicate a Kh/Kv ratio of about 30. Note that this result should be used with caution since the scales of the tests are not the same.

• The hydraulic conductivities are considerably higher in the uppermost c. one metre of the soil profile. The difference, due to impact of soil forming processes, is 2 to 3 orders of magnitude. In the model, this was considered by including a differentiation of the hydraulic properties between the upper 0.6 metre of the profile and the deeper part of the profile for some of the Quaternary deposits.

• A pumping test in till with observation wells near and below Lake Bolundsfjärden showed that there is a limited hydraulic contact, potentially determined by low-permeability lake sediments between Lake Bolundsfjärden and the pumped aquifer. The evaluation of the vertical leakage through the gyttja sediments indicates a vertical K of the gyttja of 10^–8 to 10^–9 m/s.

Groundwater levels in Quaternary deposits are of potential interest as a basis for setting the upper boundary conditions in hydrogeological models. Measured groundwater elevations in Quaternary deposits at Forsmark range from about –1 m to +13 m. However, the range in groundwater levels is only about 5.5 m when represented as depths below surface. The majority of wells form a tight cluster with reported groundwater levels in the range of approximately +0.25 to –1.5 m relative to the surface /Johansson 2008/. These wells typically show a strong uniformity in response to drier summer conditions in July and August. Similarly, these wells also display uniformity in response to recharge events following major precipitation and snow melting.

Figure 4-11 summarises the strong correlation that was observed between mean observed groundwa-ter and ground surface elevations in the Quagroundwa-ternary deposits. With a few exceptions, it can be stated that the average position of groundwater in the Quaternary deposits appears to be largely determined by the local ground surface elevation. In other words, the shape of the groundwater surface in the Quaternary deposits generally follows that of the ground surface. The most pronounced outliers are located below the ridge of the glaciofluvial deposit Börstilåsen, or in typical recharge areas (see /Johansson 2008/ for a detailed discussion on the outliers).

Modelling of groundwater flow in the bedrock at Forsmark has also been performed using flux boundary conditions at the top boundary; in fact, this has been the most common option in the later stages of the site descriptive modelling. Water balances provide a basic input when assigning flux

Figure 4‑11. Cross-plot of average groundwater level elevations in the Quaternary deposits versus ground elevations at the well locations. The red dots represent outliers; the ID numbers of these wells are listed in the figure.

0 2 4 6 8 10 12 14

0 2 4 6 8 10 12 14

Ground elevation (m, RHB70) Mean groundwater level in Quaternary deposits (m, RHB70).

SFM0008, SFM0058, SFM0059, SFM0061, SFM0077, SFM0104, SFM0080, SFM0107

boundary conditions, and for Forsmark both measured and calculated water balances are available.

In particular, the water balance of the saturated zone is needed, so that the groundwater recharge can be estimated. However, at Forsmark overland flow is small, implying that the groundwater recharge is approximately the same as total discharge.

According to /Johansson 2008/, the long-term average specific discharge in the Forsmark area is estimated to be 150–160 mm/y. However, it should be noted that this is a spatial average for the whole model area, including both (recharge) areas, where recharge actually takes place, and the discharge areas. In some cases, spatial distributions or values representing recharge areas only are more relevant. As indicated by Figure 4-5, this information can be obtained from the available modelling results.

/Johansson 2008/ used a systems approach to describe the hydrological and near-surface hydro-geological flow system of the central parts of the regional model area. The description is to a large extent based on a joint evaluation and interpretation of time-series data from Quaternary deposits and rock, see also /Johansson and Öhman 2008, Juston et al. 2007/. Some results of particular interest for the bedrock modelling and the interactions between Quaternary deposits and rock are summarised in the following.

• Correlations between seawater levels and groundwater levels in rock and in Quaternary deposits were investigated by several methods. The regression coefficients for the groundwater levels in Quaternary deposits with sea levels were very low with the exception of two wells located in glaciofluvial material within 100 m distance from the sea.

• The regression coefficients were also quite low for groundwater levels in the rock and sea levels, with the exception of boreholes located at the SFR-peninsula. The analysis indicated that seasonal precipitation and variations in evapotranspiration determined the main part of the variations in groundwater level.

• During events of very high seawater levels, seawater flows into several of the lakes. During these events, the sea obviously has an impact on both surface and groundwater flow systems in these lakes and their surroundings.

• Direct recharge from precipitation is the dominant source of groundwater recharge. However, the groundwater level measurements in the vicinity of Lake Bolundsfjärden and Lake Eckarfjärden show that the lakes may act as recharge sources to the till aquifers in this immediate vicinity during summer. Also, the Baltic Sea can potentially act as a source of groundwater recharge during periods of high seawater levels.

• The strong correlation between the mean groundwater elevations observed in the till and ground elevation data means that the average vertical hydraulic flux at some point below the surface is considerably less than the net infiltration into the saturated zone of the till. This could partly result from a contrast in vertical hydraulic conductivity between the till and the bedrock (with the uppermost bedrock having a lower Kv).

• The groundwater level in the till seems to be considerably higher than that in the rock within the tectonic lens. At drill site 1, for example, absolute groundwater levels in the Quaternary deposits are well above the point water heads in bedrock except during dry summer conditions.

• The groundwater levels in the bedrock boreholes are above the interface between Quaternary deposits and rock under undisturbed conditions, indicating that no unsaturated zone exists below this interface.

• In general, there is no discernable response in the groundwater levels in Quaternary deposits to disturbances in the groundwater levels in the bedrock. On the other hand, groundwater levels in both Quaternary deposits and bedrock are correlated to precipitation and evapotranspiration.

• Variations in groundwater levels in the till and in the bedrock are correlated. The natural

• Outside the tectonic lens, for example at drill site 4 and in the area around Lake Eckarfjärden, the groundwater level in the bedrock may be well above the groundwater levels in Quaternary deposits in nearby low-lying areas, implying that flow systems involving the bedrock may have local discharge areas.

• The lake water level-groundwater level relationship indicates that the lake sediments and the underlying till have low vertical hydraulic conductivities. If the hydraulic contact had been good, the situation with groundwater level drawdown from evapotranspiration extending below the lakes, and the quick and extensive drawdowns from the pumping would not have appeared.

Figure 4-12 shows the groundwater levels in wells in Quaternary deposits and bedrock at drill site 1 located within the tectonic lens. The groundwater levels in the Quaternary deposits are above those in the rock except during the summers of 2003 and 2006. Figure 4-13 shows the groundwater levels in a monitoring well in till below the middle of Lake Bolundsfjärden (SFM0023) and in a nearby percussion-drilled borehole (HFM32) located on a small island in the lake. Water levels in the sea and in the lake are also shown in the figure. The lake level and the groundwater level in till are considerably higher than the levels in the four sections of HFM32. The heads are lowest in the two deepest sections. The results indicate a downward flow gradient from the lake and the Quaternary deposits to the bedrock.