The lowermost layer in the DGM-deep v4.0 model is the base of the Upper Carboniferous Limburg Group (DC), which corresponds to the top of the Lower Carboniferous. The carbonate rocks of the Lower Carboniferous are known, in the Netherlands, under various names: Dinantian, Carboniferous Limestone Group, Zeeland Formation, Kolenkalk, and sometimes "Kulm". The rocks are currently the target of geothermal exploration and production in the southern part of the Netherlands, where they are found at relatively shallow depths (~2 km). In a large part of the Netherlands, however, the rocks are buried very deep (~4-9 km), and therefore possibly interesting for ultra deep geothermal ("UDG"). The rocks have a very low primary permeability, but may exhibit karst and/or fault permeability. The Dinantian is not represented in ThermoGIS v2.0 like the other clastic aquifers. This is due to the deep burial, which hampers mapping using seismic and/or well data, and the permeability distribution, which is very different from that of the clastic reservoirs.

A first subdivision of the rocks of the Zeeland Formation into two groups can be made using the supposed water depth during its genesis. Rocks that were formed in relatively shallow water depth (as "platform" or "ramp") contain little clay and have a relatively high chance of having been karstified, either hydrothermally or by subaerial exposure, which could have enhanced the permeability. These rocks were drilled into by several wells in the Dutch subsurface. Rocks that were formed in deeper water are likely to contain more clay. Those are sometimes referred to as "Kulm". They have a smaller chance of being karstified. Therefore, the latter type is considered to be less attractive for geothermal application. The nature of these rock in the Netherlands is still very uncertain, because they have not been drilled into yet. Because of this, the geothermal potential cannot be established.

The ThermoGIS v2.0 map of the top of the Zeeland Formation is for the largest part the same as the DGM-deep v4.0 map of the base of the Limburg Group. This map was constructed using seismic interpretations. For the southern part of the province of Limburg, well data were used to update the map, and for the provinces of Zeeland and North Brabant, data came from the publication by Reymer et al. in the Netherlands Journal of Geosciences (2017). From the resulting map, the areas where the clayey facies is thought to be present were removed (e.g. around the carbonate platforms in the northern part of the country), or where the rocks were buried very deeply, to the extent that the depth could not be determined with sufficient accuracy (like in the West Netherlands Basin). The uncertainty with respect to the nature of the rocks is very large.

The thickness of the Dinantian can seldom be established using seismic data. For the carbonate platforms in the north of the Netherlands, it was decided to model the base as a roughly horizontal plane that intersects the top of the Dinantian at its outer edge. In parts of Zeeland and North Brabant, as well as the northern part of Limburg (where the California doublets are located) the base of the Dinantian is visible on seismic. The thickness in those areas is therefore based on seismic data, using a seismic velocity derived from sonic velocity logs of 6.21 km/s.

The permeability of the rocks is determined by karst and fault activity, and is distributed heterogeneously. It is therefore difficult to model this spatially using interpolation techniques. Thus, it was decided not to include a Dinantian permeability map in ThermoGIS v2.0. Consequently, it is impossible to calculate the geothermal potential.

In 2019, the SCAN project (Seismic Campaign Geothermal Netherlands, or Seismische Campagne Aardwarmte Nederland in Dutch), which was initiated by EBN and TNO, will start. This project will greatly increase the knowledge of the Dinantian reservoir in the Netherlands. Relevant updates of the Dinantian maps will be implemented in a future release of ThermoGIS.