A Southern Gas Basin Prospect: An example of the Effects of Lithological Uncertainty on Fault-Seal Analysis and Hydrocarbon Flow

S.M. Clarke, M. Littler, S. Googan & D. Hughes

This example is a faulted prospect from the Southern Gas Basin, offshore United Kingdom. By contrast to the Moab Fault and Artemis Field models, this prospect provides an example where the lithological fine detail of the faulted stratigraphy is of paramount importance in controlling hydrocarbon migration and entrapment within the model. Given that this is a subsurface model based on remotely sensed data, uncertainty in the lithology is of concern.

Three-dimensional techniques of fault-seal analysis and hydrocarbon migration are applied to the model to generate deterministic predictions of hydrocarbon accumulation based on the best-case interpretations of the stratigraphy. Using the results of these models, uncertainty in the lithology is incorporated using stochastic modelling techniques to produce predictions of the most likely hydrocarbon accumulations given these uncertainties.

The three-dimensional numerical model of this Southern Gas Basin prospect is shown below. Three lithological surfaces are indicated, dividing the faulted succession into three major lithological successions. The bulk stratigraphy of the model is well known. A reservoir sandstone - the Permian Rotleigndes sand - overlies a thick, Carboniferous, coal-bearing sequence that is the gas source rock throughout the basin. The top of the Rotleigendes is represented by the yellow surface within the model. Overlying the Rotliegendes reservoir is a thick sequence of carbonates, shales and evaporates of the upper Permian Zechstein Formation. These rocks form the regional seal to the Rotleigendes reservoir and the top of this sequence is represented by the green surface. The Permian succession is overlain by a thick succession of Triassic sands, shales and evaporates (which are not recognised as part of the hydrocarbon system in this part of the basin) and a Mesozoic to Tertiary sequence up to the seabed.

Within, and towards the top of, the Zechstein sealing succession there is dolomitic carbonate, of varying thickness, that is known to be fractured in some parts of the basin. The top of this carbonate is represented by the blue surface in the numerical model.

A number of major faults trending Northwest cut and displace the full stratigraphy of the model and several additional faults cut the Rotleigendes reservoir with varying displacements.


A structural analysis of the Rotleigendes reservoir (below) indicates a major structural high and potential prospect in the footwall of Fault 1, with fault closure to the southwest (Fault 1), south (Fault 2), and northeast (faults 3, 4, & 5). Additionally, within the hanging-wall block of Fault 1 there is a small antithetic fault, with limited displacement, that cuts the Rotleigendes sequence. The integrity of this prospect depends on the extent of cross fault seal at Faults 1 to 5 and the interaction of any seal with the antithetic fault block in the hanging-wall of Fault 1.


The principal risk on this prospect is the cross-fault sealing capacity of the major faults and the minor faults that comprise the antithetic fault block. In particular the uncertainty in the lithological properties of the upper Zechstein sequence (thickness and fractured state of the dolomitic carbonate) many have a major control on the size and location of hydrocarbon accumulations. Given that, in common with many untested prospects in the Southern North Sea, the potential reserves are likely to be relatively small, this risk is of major significance.

Do the major faults seal hydrocarbons in the structural high and are the uncertainties in the lithological and petrophysical properties of the dolomitic carbonate crucial to the accumulation of hydrocarbons in this prospect?

Here, we use multiple deterministic realisations to assess the sensitivity of the results to the fractured state of the dolomitic carbonate. Based on these results we incorporate lithological uncertainty into the model and generate a stochastic solution.

The Southern Gas Basin Prospect: Deterministic Analysis

Two wells with associated logs penetrate the model and a third well lies just off the northeast corner. The lithological succession of the model is derived from these well logs and interpolated to the footwall and hanging-wall of the major and minor faults controlled by the structural geometry of the modelled surfaces.

An analysis of lithological juxtaposition (below) demonstrates that the faults controlling the prospect do have the potential to seal hydrocarbons where the Rotleigendes reservoir is juxtaposed against the Zechstein Formation. Many of the smaller faults that cut the reservoir have limited juxtaposition seals but all tip out within the model and provide spill pathways via relay-ramp structures to the main prospect in the footwall of Fault 1.

juxtamodel juxtamaintrap

However, across Fault 1 the upper carbonate is juxtaposed against the reservoir and therefore has the potential to severely affect the volume of any accumulation depending on its fractured state. In the image below, the juxtaposition of the upper carbonate against the reservoir across Fault 1 is highlighted in blue.

sealmap determ

In the examples shown ABOVE, migration pathways have been predicted through Southern Gas Basin model based on permeability distributions within the lithological units interpolated from the well data and the juxtaposition of such distributions across the faults. Two realisations have been performed. In the first (Model 1), the upper carbonate is defined as completely 'tight' (HC impermeable) and assigned a permeability accordingly. In the second (Model 2), the carbonate is defined as completely fractured. Click on the thumbnail image for more information.

Model 1: A significant accumulation is observed in the footwall of Fault 1, sealed by juxtaposition across faults 1 to 5.

Model 2: A much smaller accumulation is observed in the footwall of Fault 1. Juxtaposition of the fractured carbonate provides spill points to the antithetic fault block in the hanging-wall of Fault 1. This block traps a small accumulation in the fractured carbonate sealed by juxtaposition across faults 6 & 7.

Rapid, invasion percolation-driven hydrocarbon flow pathway modelling techniques allow multiple deterministic scenarios such as these to be modelled and a risk driven approach can be adopted. It is clear from these two models that the accumulation common to both scenarios can be considered low risk and largely insensitive to the fractured state of the upper carbonate. However, the two models presented here represent end-member cases. The dolomitic carbonate is defined as completely tight (Model 1) or completely fractured (Model 2). As a consequence, the relative risk associated with the accumulations exclusive to either model cannot be assessed and clearly the volume and location of potential traps as a whole is sensitive to the fractured state of the carbonate.

The permeability distribution of the carbonate may be altered further to reflect the degree of fracturing but the distribution of fracturing within the unit and uncertainties in the overall lithological sequence cannot be easily incorporated and their effects on hydrocarbon migration can not be easily ascertained by manually remodelling in this manner.

Which of the two potential traps indicated by the modelling outlined above, is the most likely to fill given the uncertainty in the lithological sequence?

The Southern Gas Basin Prospect: Lithological Uncertainty

In the following examples, uncertainty within the lithological succession that comprises the Southern Gas Basin prospect has been captured within the model. For four points within the model, two aspects of lithological uncertainty have been introduced. Each lithological unit thickness varies randomly by as much as 10% and, additionally, the fractured state of the dolomitic carbonate near the top of the Zechstien sequence varies randomly, with the constraint that up to 50% of the unit thickness is fractured. The image below shows twenty such computer-generated variations in the lithological succession for one point, along with the original succession. Similar variations are generated for the other three points. In all, five-hundred computer generated, randomly varying lithological successions for each point are used in the modelling.


Full fault-seal analysis and hydrocarbon flow modelling is performed with each lithological variation. In the image below, the results of each lithological juxtaposition realisation are combined to indicate the likely leaky state of any point on the main fault surface. Those points that leak in 100% of the realisations are coloured red and those points that never leak are coloured blue. Clearly there are a number of points within the structural high of the fault-sealed footwall trap of the main fault (indicated by the oval) that leak for a significant proportion of these randomly generated realisations and therefore have a high likelihood of leaking in reality.


Hydrocarbon flow models can be calculated for each realisation and the results combined. In the image below, invasion percolation techniques have been used to ascertain flow pathways and accumulations for each realisation with each variation in lithological succession. The results are combined and the model nodes are shown coloured by their frequency of fill. Those volumes that fill 100% of the realisations are coloured red. These volumes are likely to always be filled of hydrocarbons irrespective of the uncertainty in the stratigraphy of the model. Points coloured blue represent those volumes that fill only 1% of the time. These volumes require a particular specific lithological succession in order that hydrocarbons can migrate into them and are therefore extremely sensitive to the lithological succession of the model. Colours between red and blue indicate varying frequencies of fill over the 500 simulations and therefore indicate varying sensitivities of those accumulations to the lithological succession of the model.


Low risk accumulations are evident in the footwall structural high of the main fault although these are limited in volume. The risk of greater volumes in this trap is significantly larger as large volumes require a non-fractured dolomitic carbonate. The antithetic hanging-wall fault block represents a trap with a medium risk as accumulation here largely depends on at least some fracturing of the dolomitic carbonate in the Zechstein sequence. Accepting its existence, the volume of this trap is largely unaffected by lithological uncertainty. The trapped volume here is smaller than the footwall accumulation but it is lower risk than a significant proportion of the footwall trap. However, none of the volume of the antithetic fault block trap is as low risk as the upper part of the main fault footwall high.

Exploring fault seal and hydrocarbon migration in three dimensions reduces many of the inherent inaccuracies introduced by one- and two-dimensional analysis. The three-dimensional environment allows the interaction of many faults within the prospect, the structural geometry of the faulted blocks and their combined effects on fault seal to be examined more completely. Combining these techniques with fast invasion percolation-driven hydrocarbon flow pathway modelling allows fault leak points and fault-controlled accumulations to be determined quickly. In this way multiple scenarios can be analysed to examine the effect of various uncertain input parameters on the outcome. However, multiple realisations with user-controlled inputs produce a limited set of results. By using these results to automatically incorporate sensitive lithological uncertainties into the modelling process it is possible to risk outcomes accordingly. The incorporation of aspects of lithological uncertainty into the fault seal modelling process for this Southern Gas Basin prospect allows risking of possible hydrocarbon accumulations based on their sensitivity to the uncertain lithological succession of the model.

Further Information

The three-dimensional modelling presented here was performed using in-house modelling software developed by the Basin Dynamics Research Group at Keele. This software has now been developed into a commercial three-dimensional fault seal analysis and flow pathway modelling software package called Qfault. The software combines all of the three-dimensional modelling techniques presented here with a powerful advanced visualisation and model-building tool. Images produced specifically using Qfault have the logo on them.

For further information on Qfault or any of the techniques demonstrated here please contact the corresponding author:

For further information on any of the techniques demonstrated here please contact the corresponding author:

Stu Clarke