• No results found

23 History of pore pressures and groundwater flow

23.2 History of groundwater flow

The model predictions of the history of groundwater flow patterns and flow rates will be discussed below in relation to the previously identified periods of active paleo groundwater flow (Figure 83 and Table 16).

Pre- and Early-rift phase

Continuous regional sedimentation during the pre- and early rift phase of basin evolution was the primary mechanism influencing the rates and patterns of groundwater flow accounted for in the modelling. In general, the predicted groundwater flow induced by the continuous regional sedimentary loading during the pre- and early rift phase, was directed vertically upwards in the poorly permeable units and was bedding-parallel in the relatively permeable units. The calculated mean vertical flow rates at shallow depths in newly deposited sediments were in the order of 10 - 50 m/My (mean Darcy flow rates). The bedding-parallel flow rates were higher than the vertical rates: >100 m/My. The basin was subaerial during deposition of the terrestrial Slochteren Formation of the Upper Rotliegend Group, and rates – and as a consequence directions of groundwater flow – were also influenced by topography-driven flow: in the newly deposited sediments of the Slochteren Formation the flow directions varied and mean flow rates were 100 - 500 m/My. Figure 84 shows the

model-predicted directions of groundwater flow for parts of the flow system with the highest groundwater flow rates at the end of the early-rift subsidence phase.

At a depth of approximately 1000 m, the mean rates of vertically upward groundwater flow in the poorly permeable Aalburg Formation were less than 5 m/My. At depths of more than 3000 m the vertical flow rates reduced to <1 m/My (in Limburg Group).

Predicted-bedding parallel flow occurred in the following relatively permeable units:

Middle Werkendam Member, sandstone units of the Lower Germanic Trias Group, Solling Formation and the Slochteren Formation.

SW

1

Overpressure in MPa

2 4 6 8 10

Depth (m)

P9 P6 Q1

Broad Fourteens Basin

4000 1000

2000

3000

6000 5000 0

7000

70 Distance (km)

0 10 20 30 40 50 60

NE

Overpressure in MPa

Depth (m)

P9 P6 Q1

Broad Fourteens Basin

4000 1000

2000

3000

6000 5000 0

7000

70 Distance (km)

0 10 20 30 40 50 60

NE SW

<1 2 4 6 8 10

Figure 82 Predicted distribution of present-day overpressures at present-day (modelling assumption: open hydraulic boundary at northeastern side of cross-section)

Figure 81 Predicted distribution of overpressures at the end of the post-rift phase of basin evolution (at 74 Ma; modelling assumption: open hydraulic boundary at northeastern side of cross-section)

Before the start of major gas generation in the syn-rift phase, the Carboniferous source rocks in the basin generated other volatile non-hydrocarbon compounds, such as CO2(Section 21.1). The continuous expulsion of CO2-rich groundwater from the Limburg Group into the Upper Rotliegend Group and overlying Zechstein Group is in accordance with the flow conditions required to explain kaolin cement and leached K-feldspar in the Slochteren Formation and calcite cements in Z3 Carbonate Member (Figure 83 and Table 16). This is also illustrated by the modelled history of overpressures in the Limburg Group, Slochteren Formation and Zechstein Group in the central part of the basin (Figure 80). During the Early-rift phase, overpressure gradients are consistent with vertical upward expulsion of groundwater from Limburg Group into Slochteren Formation and Zechstein Group.

Topography-driven flow developed again at shallow depths in the basin and flanking regions during the Mid Kimmerian phase of uplift at the end of the early-rift phase.

Mean lateral groundwater flow in the Werkendam Formation attained mean Darcy flow rates of approximately 100 m/My at shallow depths of 500 m.

Main syn-rift phase

Groundwater flow in the early part of the main syn-rift phase was modelled to be affected by the presence of fault zones with increased permeabilities, increased rates of sedimentation during deposition of the Delfland Subgroup, gas generation and the elevation of the basin above sea level. The resulting groundwater flow was dominated by bedding-parallel flow in the relatively permeable units, including the recently deposited Delfland Subgroup. The permeable fault zones dewatered the relatively

Depth (m)

Variscan Pre- and early rift Main syn-riftPost-riftSyn-inversionPost-inversion

II I

III IV

300 250 200 150 100 50 0

Age (Ma) 0

500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000

Limburg Group Lower and Upper Germanic Trias Groups Zechstein Group Upper Rotliegend Group Altena Group North Sea Groups Chalk Group Rijnland Group and Schieland Group

fluid flow events I

Figure 83 Timing of paleo-groundwater flow events. Fluid flow events I, II, III and IV according to Table 16 in Part 2

permeable units (Slochteren Formation, Triassic sandstones and Middle Werkendam Member) and bedding-parallel flow was towards the P6, P9 and Q1 fault zones.

In comparison with the foregoing period, the predicted groundwater flow rates were higher both in the upper part of the basin (vertical flow regime) and in the permeable units in the deeper parts of the basin (e.g. a Darcy flow rate of 900 m/My in the Solling Formation at 26 km along the cross-section, and 400 m/My in the Slochteren Formation at 13 km). Average Darcy flow rates for the entire basin were approximately 500 m/My in Triassic sandstones and 100 m/My in Slochteren Formation.

During the main syn-rift phase the overpressure gradients of the groundwater were directed from the Limburg Group and the Zechstein Group towards the dewatering and lower overpressured Slochteren Formation (Figure 80). These modelled expulsion directions of water from both the Limburg and the Zechstein Group are consistent with one of the flow conditions required to explain illite cements in the Slochteren Formation (Table 16).

The fault zones were inactive and poorly permeable during the subsequent deposition of Rijnland Group sediments. The sedimentary loading and gas generation affected the groundwater system. Sedimentary loading induced the following vertical Darcy flow rates in the syn-rift deposits: <10 m/My in the Holland Formation and <5 m/My in the Vlieland Claystone Formation. Differential sedimentary loading induced lateral flow in the Vlieland Sandstone Formation (<– 2000 m/My). Deeper in the basin the calculated vertical flow rates fell to <– 1 m/My in the Aalburg Formation and

<– 0.1 m/My in the Limburg Group.

Overpressure in MPa

<1 2 4 6 8 10

direction of groundwater flow mean groundwater flow rates > 10m/My

Modelling scenario P3 and closed hydraulic bottom and lateral boundaries

Depth (m) 4000

1000

2000

3000

6000 5000 0

7000

P9 P6 Q1

Broad Fourteens Basin

SW NE

70 Distance (km)

0 10 20 30 40 50 60

Figure 84 Predicted distribution of overpressures, pattern of groundwater flow and mean rates of groundwater flow at the end of the early-rift phase of basin evolution (modelling scenario P3)

Post-rift phase

Post-rift regional subsidence and sedimentation and gas generation were the principal mechanisms influencing the groundwater and were calculated to result e.g. in continuation of the vertical upward flow of groundwater in the shallow part of the basin (25 m/My in Chalk; <10 m/My in Holland Formation; <5 m/My in Vlieland Claystone Formation). Figure 85 shows the modelling results at the end of the post-rift phase. Main flow was concentrated in the upper part of the basin (vertical flow) and in the relatively permeable units (bedding-parallel flow). The highest flow rates of bedding-parallel flow occurred in the sandstones of the Delfland Subgroup and in the Vlieland Sandstone Formation. Flow was restricted in the deeper poorly permeable parts of the basin as illustrated by the previously described extended zone of overpressuring (Figure 77).

Syn-inversion phase

Figure 86 illustrates the syn-inversion groundwater flow system at the end of the Cretaceous. Continued sedimentary loading in the northern platform area maintained the vertically upward flow system until depths of approximately 2000 m. The mean vertical rate of groundwater flow in the Chalk Group and Holland Formation was 50 m/My.

The pre-inversion overpressured conditions did dissipate in the inverted part of the basin. The topographic relief of the water table (in the modelling represented by the topographic relief of the ground surface) of the inverted basin was the main mechanism inducing groundwater flow. The inversion-related tilting of the geological units exposed different permeable units (e.g. Vlieland Sandstone Formation) to inflow of surface-derived water. In addition, fault zones have assigned increased permeabilities during the Late Cretaceous part of the syn-inversion period. The predicted topography-induced flow was from the central inverted part of the basin towards its edges. In the southwestern part of the cross-section the groundwater discharged by vertical upward flow in the area between 0 and 12 km. In the northeast two more local discharge areas occurred above structural highs at 53 and 62 km.

Shallow topography-induced flow systems developed in the Chalk Group deposits in the southwestern part of the basin. Regional topography-induced groundwater flow from the central part of the basin towards its edges was focussed through the outcropping Vlieland Sandstone Formation. Southward bedding-parallel flow in the P9 area had predicted mean Darcy flow rates of 700 m/My at a depth of 1300 m (corresponding to true flow rates of approximately 4400 m/My). The regional topography-induced flow towards the north through the Vlieland Sandstone Formation had predicted Darcy flow rates of 8000 m/My at a depth of 455 m in the Q1 area (corresponding to a true flow rate of approximately 46 000 m/My).

In modelling scenario P3, the permeable fault zone in the P6 area was located in the basin-wide recharge area of the topography-induced groundwater flow system and allowed deep infiltration of water feeding the permeable units in the deeper part of the basin (such as the Triassic sandstones of the Lower Germanic Trias Group, the Solling Formation, and even the Slochteren Formation). The present-day NW-SE trending fault-related fractures in the Upper Rotliegend Group and inversion-related fractures in Z3 Carbonates indicate that vertical permeability in the basin increased

Overpressure in MPa

<1 2 4 6 8 10

direction of groundwater flow mean groundwater flow rates > 10m/My

Modelling scenario P3 and closed hydraulic bottom and lateral boundaries Broad Fourteens Basin

Depth (m)4000

1000

2000

3000

6000 5000 0

7000

P9 P6 Q1

SW NE

70 Distance (km)

0 10 20 30 40 50 60

Figure 85 Predicted distribution of overpressures, pattern of groundwater flow and mean rates of groundwater flow at the end of the post-rift phase of basin evolution (at 74 Ma, modelling scenario P3)

Depth (m)4000

1000

2000

3000

6000 5000 0

7000

P9 P6 Q1

Broad Fourteens Basin

SW NE

70 Distance (km)

0 10 20 30 40 50 60

Overpressure in MPa

<1 2 4 6 8 10

direction of groundwater flow mean groundwater flow rates > 10m/My

Modelling scenario P3 closed hydraulic bottom and lateral boundaries

Figure 86 Predicted distribution of overpressures, pattern of groundwater flow and mean rates of groundwater flow during the syn-inversion phase of basin evolution (at 65 Ma, modelling scenario P3)

Modelling scenario P2

Closed hydraulic bottom and lateral boundaries Mean rate of groundwater flow in m/My

10 1000

Horizontal permeability in mD

>500

10–7 10–6 10–4 0.01 1 100

Oil accumulations

Depth (m) 4000

1000

2000

3000

6000 5000 0

7000

P9 P6 Q1

Broad Fourteens Basin

SW NE

70 Distance (km)

0 10 20 30 40 50 60

Figure 87 Predicted distribution of horizontal permeabilities, pattern of groundwater flow and mean rates of groundwater flow during the syn-inversion phase of basin evolution (at 65 Ma, modelling scenario P2)

Overpressure in MPa

<1 2 4 6 8 10

direction of groundwater flow mean groundwater flow rates > 10m/My

Modelling scenario P3 and closed hydraulic bottom and lateral boundaries

Depth (m) 4000

1000

2000

3000

6000 5000 0

7000

P9 P6 Q1

Broad Fourteens Basin

SW NE

70 Distance (km)

0 10 20 30 40 50 60

Figure 88 Predicted distribution of overpressures, pattern of groundwater flow and mean rates of groundwater flow at present-day (modelling scenario P3)

as a result of inversion (Gauthier et al. 2000, and Van der Poel 1989, respectively;

Part 2). However, there are no published data known to the author, that support the actual syn-inversion flushing of these deep Triassic and Permian sandstone units in the southern part of the basin.

Model scenarios P2 and P3 of the present modelling study (Figure 87 and 86 respectively) both indicate the existence of active topography-induced groundwater flow through Chalk and Rijnland Group. The focussed topography-induced flow through the Vlieland Sandstone Formation occurred independently of the P6 fault system (Figure 87). Such focussed topography-driven flow is supported by flow conditions required to explain e.g. K-feldspar leaching in the Vlieland Sandstone Formation (Fluid flow event IV, Figure 83, Table 16).

The modelling package applied did not allow the introduction of variable densities of the groundwater during the modelling. A constant density of the groundwater of 1030 kg/m3was used (Appendix 4). Within the scope of this Broad Fourteens study Bouw (1999) studied the syn-inversion development of a freshwater lens in the more salt-dominated part of the basin north of the cross-section, by applying a density-dependent groundwater flow model. The results of her modelling scenarios (which do not include permeable faults, but do take into account different permeability assumptions for the hydrostratigraphic units) show that permeability distribution was the main factor of influence on the development of a freshwater lens. Predicted near steady-state conditions were established within 4 My. The maximum depth of the fresh-saltwater interface was 1200 m for a high permeability scenario. For a low permeability scenario, active topography-induced flow was mainly restricted to Rijnland and Chalk Group deposits.

Post-inversion phase

The post-inversion deposition of the Lower North Sea Group deposits, the uplift and erosion of the basin during Eocene–Oligocene and the subsequent Oligocene-Miocene period of non-deposition or only minor deposition of Middle and Upper North Sea Group deposits resulted in prolonged periods of near-hydrostatic conditions in the basin (Section 23.1). The sedimentation rates increased during the Pliocene and, especially, Quaternary. Table 18 summarises the predicted evolution of overpressure build-up and changes in groundwater flow rates during the last 5 million years. The present-day result of the Pliocene-Quaternary sedimentary loading on the groundwater flow characteristics is illustrated in Figure 88, and includes the following characteristics:

— the main area of groundwater flow is located above the overpressured zone;

— in the upper part of the basin groundwater flow is cross-formational and mainly vertically upwards, flowing through Quaternary, Tertiary and Cretaceous units;

flow is concentrated above structural highs at 31, 53 - 54 and 61 - 62 km along the cross-section;

— in the Quaternary deposits (with an assigned permeability anisotropy of 5) lateral flow is also observed; the flow is from the Quaternary depocentre (at 54 km) towards the edges of the Quaternary basin;

— bedding-parallel flow of groundwater through permeable sandstone units occurs at all depths.

SW NE

Modelling scenario P3 and closed hydraulic bottom boundary

closed hydraulic boundary at southwestern side of cross-section open hydraulic boundary at northeastern side of cross-section Mean rate of groundwater flow in m/My

10 1000

Horizontal permeability in mD

>500

10–7 10–6 10–4 0.01 1 100

P9 P6 Q1

Broad Fourteens Basin

70 Distance (km)

0 10 20 30 40 50 60

Depth (m) 4000

1000

2000

3000

6000 5000 0

7000

Modelling scenario P3 and closed hydraulic bottom boundary

closed hydraulic boundary at southwestern side of cross-section open hydraulic boundary at northeastern side of cross-section Mean rate of groundwater flow in m/My

10 1000

Horizontal permeability in mD

>500

10–7 10–6 10–4 0.01 1 100

Depth (m) 4000

1000

2000

3000

6000 5000 0

7000

P9 P6 Q1

Broad Fourteens Basin

SW NE

70 Distance (km)

0 10 20 30 40 50 60

Figure 90 Predicted distribution of horizontal permeabilities, pattern of groundwater flow and mean rates of groundwater flow during the syn-inversion phase of basin evolution (modelling assumption: open hydraulic boundary at northeastern side of cross-section)

Figure 89 Predicted distribution of horizontal permeabilities, pattern of groundwater flow and mean rates of groundwater flow at the end of the post-rift phase of basin evolution (at 74 Ma;

modelling assumption: open hydraulic boundary at northeastern side of cross-section)

The increasing sedimentation rates induce increases in vertical rates of groundwater flow in the Lower Tertiary and the Cretaceous units, as well as the build-up of overpressures in relatively poorly permeable units and the extension of the overpressured zone towards younger units at shallower depths (Table 18).

The predicted history of groundwater flow assuming an open northern boundary since Early Cretaceous times, clearly shows an increase in northward bedding-parallel flow along the cross-section in the area north of 50 km. All flow through the permeable Triassic sandstone units and the Slochteren Formation in this area was northward at the end of the post-rift phase, at syn-inversion and at present-day (Figures 89, 90 and 91).