Bavaria’s Deep Geothermal Reservoir Mapped: Where Water Flows and Why It Matters

Beneath the rolling landscape of southern Bavaria lies one of Europe’s most valuable deep geothermal reservoirs, and researchers have now produced the most detailed picture yet of where water actually moves through it. A new study published in the journal Geothermal Energy has, for the first time, linked basin-scale measurements of reservoir transmissibility with the specific rock intervals that carry flow in individual wells, and placed both within a sequence-stratigraphic framework of the Upper Jurassic to Lower Cretaceous ‘Malm’-‘Purbeck’ succession in the North Alpine Foreland Basin.

Thirty-seven of these values came from previous analyses, eleven from operators’ hydraulic test reports, and thirteen were newly evaluated from raw well test data using a single uniform method. This standardisation matters, because earlier regional assessments had combined hydraulic information gathered with different test methods, durations and tested aquifer thicknesses, which made comparisons between wells unreliable.

Transmissibility, expressed in millidarcy-metres, describes the capacity of a reservoir to transmit fluid and is the product of permeability and reservoir thickness. The team calculated it using the Horner semi-log plot method, focusing on the infinite-acting radial flow regime of the longest hydraulic test following acid stimulation at each well. Flow rates were measured directly rather than averaged, and water viscosity was computed from salinity measurements together with reservoir pressure and temperature data, reducing a key source of uncertainty.

To map how transmissibility varies across the basin, the researchers applied Empirical Bayesian Indicator Kriging, a geostatistical method chosen for its ability to handle small, heterogeneous datasets and to incorporate uncertainty in the underlying spatial model. Rather than producing a single deterministic map, the method yields probability distributions: contour lines at the p10, p50 and p90 probability levels show, for each transmissibility threshold, where values exceed that threshold with worst-case, expected and best-case confidence. In areas dense with wells, such as greater Munich, the contour lines converge, signalling low uncertainty; in sparsely drilled regions to the west, they diverge markedly.

The resulting spatial distribution confirms a well-known trend: transmissibility generally decreases from north to south as the reservoir becomes progressively deeper and more compacted and diagenetically altered. But the probabilistic map also reveals regional deviations from this trend. In the northeast of the Munich area, transmissibility increases significantly, supported by abundant data. Around Landshut, values fall below the general trend, although the authors caution this observation rests on a very limited dataset. And in a small, well-constrained area south of Munich, where data quality and quantity are both high, the p50 contour for 10,000 millidarcy-metres shows a distinct northward spike of reduced transmissibility, an anomaly also visible in earlier classification maps.

The study’s second major contribution goes vertically rather than laterally. In 14 wells where both well tests and flowmeter spinner logging were available, the team distributed each well’s total transmissibility among the hydraulically active zones identified by the flow logs, weighting by each zone’s share of total flow and its thickness. Only zones contributing more than five percent of total flow were considered. Dividing each zone’s transmissibility by its thickness yielded an effective permeability, combining matrix, fracture and vuggy contributions, for every individual flow zone. The analysis confirms a systematic decrease in permeability with increasing depth, but shows that the details depend strongly on region.

Using reservoir types previously defined through multivariate statistical analysis of geothermal well data around Munich, the researchers found sharply contrasting behaviours. Type A (East) wells show the highest permeabilities between roughly 1,900 and 2,600 metres depth, with a wide range and no discernible depth relationship. Type A (West) zones, at similar depths, have noticeably lower permeabilities but a clear depth trend, with a coefficient of determination of 0.64, and the highest variability of all types, spanning roughly 1.5 to 1,000 millidarcy-metres. Type B wells, further south and deeper, show a weaker but perceptible depth trend and lower variability. Notably, hydrochemistry was previously identified as the main factor separating Type A (East) from Type A (West), yet the new permeability data distinguish them clearly as well, pointing to spatial geological variations not yet identified. The authors emphasise that these results rest on only 14 wells, mostly in the greater Munich area, so transferability to the entire basin cannot be confirmed.

The third strand of the work placed the hydraulically active zones within four large sequence-stratigraphic units, each subdivided into transgressive and regressive phases, whose marker horizons can be correlated across the basin even where individual lithofacies pinch out. The analysis shows that flow zones occur predominantly within specific intervals. Overall, the average share of flow decreases from the uppermost regressive phase LS4R down to LS1R. The lower ‘Purbeck’ and the top of LS3R, corresponding to the lower ‘Purbeck’ and Bankia interval, dominate inflow, with the upper quarter of LS3R reaching average flow shares of about 25 percent, while the deeper units LS2T, LS1R and LS1T contribute progressively less, with no active zones at all in LS1R and LS1T. This aligns with literature describing the deeper ‘Malm’ alpha to gamma sections as usually impermeable. Importantly, the authors note that sequence stratigraphy does not directly control permeability magnitude; it provides a consistent framework for describing where flow zones sit, and the observed decrease in flow share cannot be separated from the fact that the stratigraphy deepens in the same direction.

Splitting the analysis by reservoir type reveals further regional patterns. Type A (East) wells each host three to four active zones distributed with similarly high transmissibility throughout the sequence stratigraphy, with no accumulated occurrence. In Type A (West), the upper intervals LS4R, LS4T and the top of LS3R dominate flow. In Type B, the lower LS3R is dominant, with two distinct flow peaks. The southward deepening of active zones from Type A to Type B areas coincides with better preservation of younger stratigraphic units above the ‘Malm’-‘Purbeck’, leading the authors to suggest that permeability associated with dissolution and karstification may, in the south, have shifted into overlying younger units such as Hauterivian limestones.

The practical payoff comes through the BeM-TG Bayern project, an assessment model for deep hydrothermal energy in Bavaria, where these results will feed static and dynamic geological models used for production and interference forecasting. Knowing statistically where hydraulically active zones occur, both in sequence-stratigraphic position and in regional context, should improve well targeting and reduce the risk of drilling unproductive wells. The authors also argue the workflow, combining standardised well test analysis, probabilistic spatial mapping and flowmeter-resolved permeability tied to sequence stratigraphy, is transferable to other hydrothermal reservoirs in sedimentary foreland basins, citing similarities with the Western Canada Sedimentary Basin. For now, the caveats stand: the vertical analysis draws on a small well count concentrated around Munich, flowmeter data cover only open-hole sections that do not always span the full reservoir, and data remain sparse across much of the basin. Future drillings will expand the database, and the researchers plan to compare their hydraulic zones with geophysical log-derived porosity-permeability relationships to explain why some intervals, and not others, carry Bavaria’s geothermal water.

Subject of Research: Integrated characterisation of transmissibility, hydraulically active zones and sequence stratigraphy in the deep geothermal 'Malm'-'Purbeck' reservoir of the North Alpine Foreland Basin in Bavaria.

Article Title: Linking transmissibility and hydraulically active zones in the deep geothermal reservoir in Bavaria

Article References: Ernst, V., Schölderle, F., Crinière, A., & Zosseder, K. (2026). Linking transmissibility and hydraulically active zones in the deep geothermal reservoir in Bavaria. Geothermal Energy, 14(1), Article 23. https://doi.org/10.1186/s40517-026-00398-1

Image Credits: AI Generated

DOI: 10.1186/s40517-026-00398-1

Keywords: deep geothermal energy, North Alpine Foreland Basin, Malm reservoir, transmissibility, hydraulically active zones, sequence stratigraphy, Empirical Bayesian Kriging, permeability, flowmeter logging, pressure transient analysis, Bavaria, reservoir characterisation