## Data porosity and permeability

### Cretaceous and older

All publicly available onshore data were used for creating the porosity and permeability maps. These are data that are public under the Mining Law (age > 5 years), and data less than 5 years old that are public because they underlie RNES-Garantiefonds applications. The latter can be found on the RVO website. Generally speaking, more porosity (~1400) than permeability (~600) data points are available.

For the selection of data points, a ranking order was used that is based on the analysis type, which is linked to accuracy. In order of decreasing accuracy this is:

Porosity:

- Full petrophysical analysis
- LogQM average, calibrated using core plug data
- Petrophysical analysis of the pay zone only
- Core plug data average
- LogQM average, not calibrated using core plug data

Permeability:

- Well test analysis
- Petrophysical analysis
- LogQM average
- Core plug data average

LogQM is a tool developed by TNO that calculates aquifer porosity and permeability averages from logs and core plug data stored in the DINO database in a semi-automatic fashion. Not all data points have been used for generating the maps. The main reason for ignoring a data point is when its value is anomalous and can be considered as unrepresentative for the regional trend.

No porosity or permeability map has been made for the Dinantian reservoir. The carbonate rocks of this unit have a very low primary porosity and permeability. However, these rocks sometimes have secondary porosity and permeability due to dissolution and/or faulting. The spatial distribution of this type of permeability is very heterogeneous. The techniques that are used for calculating permeability maps of the other (clastic) aquifers is unsuitable for the Dinantian.

### Paleogene and Nederweert Sandstone Member

Little data are available for aquifer rocks of the Paleogene period. Apart from the Asten and Zevenbergen wells, these aquifers have not been targeted for geothermal exploration yet. The available data are too sparse and inconsistent to allow the generation of accurate regional porosity and permeability maps. The same holds for the rocks of the Nederweert Sandstone Member.

## Porosity maps

### Cretaceous and older

The porosity maps were generated by interpolating the average reservoir porosities using a geostatistical method: co-kriging. This interpolation method uses the assumed relation between primary (average reservoir porosity) and secondary variables (maximum burial depth) to estimate an average reservoir porosity map. This method is especially useful in areas with sparse primary data and abundant secondary data. Porosity generally decreases with increasing burial depth (especially due to compaction). This is an irreversible process; rocks that were buried deeply and have later been uplifted still retain the decreased porosity belonging to the deep burial. Therefore, maps depicting the difference between estimated maximum burial depth and current depth, named the burial anomaly, is used as secondary variable. Figure 3 shows that, for the Lower Detfurth Member, the correlation between porosity and current depth is lower than the one between porosity and maximum burial depth.

### Paleogene and Nederweert Sandstone Member

See below.

## Permeability maps

### Cretaceous and older

**Permeability trend:** a linear relation between porosity and the natural logarithm of permeability is often used to derive permeability from porosity. A study carried out by TNO shows that, especially for higher porosities, this relationship is curved rather than linear (figure 4). Assuming a linear relationship overestimates the permeability for both low and high porosities and underestimates the permeability for intermediate porosities. For ThermoGIS v2.0, a polynomial trend line was fit to the average reservoir core plug measurements. This trend line was used to convert average reservoir porosities to permeabilities (see Poster 3: "Reservoir properties revisited" on www.NLOG.nl)**.**

**Final permeability:** The permeability trend provides a first impression of the regional trend of each aquifer. On a local scale, however, the average reservoir permeability may differ from the trend. Therefore, the trend permeability map was corrected for the local average permeability derived from well data. This was achieved by interpolating the residual permeabilities (the difference between well and trend) and adding them to the trend. This yields a final permeability map.

P10 and P90 permeability maps are calculated from the geostatistically calculated uncertainty (kriging standard deviation) of the porosity maps added to an uncertainty assigned to the porosity-permeability relationship. Generally, the uncertainty increases with increasing distance from a data point.

### Paleogene and Nederweert Sandstone Member

A different method was used for calculating the porosity and permeability maps of the Paleogene and Nederweert Sandstone Member aquifers. The reason for applying a different method is lack of sufficient reservoir property data. The sediments of the Paleogene aquifers were predominantly deposited in a shallow marine environment (Wong et al., 2007). It was therefore decided to adopt a single porosity-depth and porosity-permeability relationship for the entire Paleogene aquifer stack. Various relationships found in the literature were compared to the sparse available data, and the best possible fit was used. The uncertainty of the resulting map is therefore large. Because the workflow is not data-driven, a (geo-)statistical uncertainty (standard deviation) on which the P10 and P90 maps are based, cannot be calculated. Therefore, a fixed value was adopted which corresponds to the maximum standard deviation of the deeper layers.

**Note:** the porosity maps are not available in the Map viewer because they were not used in the ThermoGIS calculations. They are nevertheless described here because the permeability maps are based on them.