Bavaria Geothermal Reservoir Study Maps Where Water Flows

Estimated reading time: 8 minutes · Last updated: 31 October 2026

A study published in 2026 in the journal Geothermal Energy maps how easily water moves through Bavaria's deep geothermal reservoir. It also pins that flow to particular rock layers inside individual wells. The authors are V. Ernst, F. Schölderle, A. Crinière and K. Zosseder. They paired a probabilistic, basin-wide map of transmissibility with flowmeter logs from 14 wells. They then set the active zones within the layered 'Malm'-'Purbeck' rocks of the North Alpine Foreland Basin, which span the Upper Jurassic to Lower Cretaceous. Flow concentrates in the lower 'Purbeck' and the top of a unit called LS3R. The two deepest units carry none. The results will feed the BeM-TG Bayern assessment model, which is meant to improve how wells are targeted.

Why the Authors Rebuilt Bavaria's Transmissibility Data

Transmissibility measures how readily a reservoir passes fluid. It is permeability multiplied by reservoir thickness, and it is expressed in millidarcy-metres. Earlier regional assessments blended well tests that differed in method, in length and in how much aquifer each one sampled. That made comparisons between wells unreliable.

The new dataset draws on three pools. There are 37 values from previous analyses and 11 from operators' hydraulic test reports. The authors evaluated another 13 from raw well test data with one consistent method, giving 61 values in all.

For each well, they took the longest hydraulic test run after acid stimulation and analysed its infinite-acting radial flow regime on a Horner semi-log plot. Flow rates were used as measured rather than averaged. Water viscosity was derived from measured salinity plus the reservoir's pressure and temperature, which removes one of the larger uncertainties in the calculation.

Where the Probabilistic Map Breaks From the North-South Trend

To spread point values across the basin, the authors used Empirical Bayesian Indicator Kriging. Kriging is a geostatistical technique that estimates values between measured points. This variant copes with small, uneven datasets and carries uncertainty in the spatial model through to the result. Instead of one map, it draws contours at the p10, p50 and p90 levels, marking where a transmissibility threshold is exceeded under worst-case, expected and best-case assumptions.

Around greater Munich, where wells cluster, the contours bunch together, signalling low uncertainty. In the thinly drilled west they spread far apart.

The broad pattern holds. Transmissibility falls going south, where the reservoir sits deeper and is more compacted and diagenetically altered. There are exceptions, though. Northeast of Munich, values rise significantly, backed by plentiful data. Around Landshut they dip below the trend, though the authors warn that the dataset there is very limited. South of Munich, in a small area with strong data, the p50 contour for 10,000 millidarcy-metres juts northward. That marks a pocket of lower transmissibility that earlier classification maps had also shown.

How Permeability Separates Munich's East From Its West

The vertical analysis used 14 wells that had both well tests and flowmeter spinner logs. A spinner log records how much fluid enters the borehole at each depth. Each well's total transmissibility was shared among the active zones the logs revealed, weighted by each zone's portion of the flow and by its thickness. Zones carrying five percent of total flow or less were left out. Dividing a zone's transmissibility by its thickness gave an effective permeability that lumps matrix, fracture and vuggy contributions together.

Permeability falls with depth overall, but region matters. Type A (East) wells reach the highest permeabilities, at depths of roughly 1,900 to 2,600 metres, with wide scatter and no link to depth. Type A (West) zones at similar depths are noticeably less permeable. They show a depth trend that scores 0.64 on the coefficient of determination, and they vary the most of any type.

Hydrochemistry had been the main feature telling the two Type A groups apart. Permeability now separates them as well, which the authors read as a sign of geological differences across space that nobody has yet identified.

Which Rock Layers Carry Bavaria's Geothermal Water

The authors sorted the active zones into four large sequence-stratigraphic units, each split into transgressive and regressive phases. Sequence stratigraphy divides rock into packages laid down as sea level rose and fell. Its marker horizons can be traced across the basin even where individual rock types thin out.

Average flow share shrinks step by step going down, from LS4R, the uppermost regressive phase, to LS1R. The lower 'Purbeck' and the top of LS3R, which together correspond to the lower 'Purbeck' and Bankia interval, dominate inflow. The upper quarter of LS3R averages about 25 percent. LS2T contributes less, and LS1R and LS1T hold no active zones. That fits published work treating the 'Malm' alpha to gamma sections lower down as usually impermeable.

The authors are careful about cause. In their view, the stratigraphy gives a consistent way to describe where flow zones sit, but it does not set how permeable a layer is. Because the units deepen in the same direction that flow share falls, the two effects cannot be separated. Younger units above the 'Malm'-'Purbeck' survive better in the south. On that basis, the authors suggest that permeability from dissolution and karstification may have shifted upward there, into younger rocks such as Hauterivian limestones.

What the Findings Change for Drilling in Bavaria

The results go into BeM-TG Bayern, a project building an assessment model for Bavaria's deep hydrothermal resource. They will feed static and dynamic geological models that forecast production and interference. The authors argue that knowing statistically where active zones sit, both in the rock sequence and by region, should help operators target wells and avoid drilling unproductive ones.

They also say the combined workflow could carry over to hydrothermal reservoirs in other sedimentary foreland basins, and they point to similarities with the Western Canada Sedimentary Basin. That claim has not yet been tested outside Bavaria.

The limits are real. The vertical analysis rests on 14 wells, mostly around Munich. Flowmeter logs cover only open-hole sections, which do not always reach across the whole reservoir. Data stay sparse across much of the basin.

What Supports and What Limits the Findings

The bottom line

The strongest contribution here is discipline rather than discovery. Putting 61 transmissibility values on one footing and publishing confidence contours instead of a single map gives planners an honest picture: dependable around Munich, thin in the west and near Landshut. The flowmeter work adds a vertical layer that regional maps lack, showing the lower 'Purbeck' and upper LS3R doing most of the work while LS1R and LS1T stay dry. Its weakness is sample size. Fourteen wells clustered near one city cannot speak for a whole basin, as the authors concede. The next tests are the BeM-TG Bayern models, the planned log-based porosity comparison and new drilling beyond Munich.

What to watch

• Watch for the BeM-TG Bayern project to fold these results into its static and dynamic models for production and interference forecasting; no date has been set.
• Watch for the authors' planned comparison of their hydraulic zones with porosity-permeability relationships derived from geophysical logs, which could explain why some intervals flow and others do not; no date has been set.
• Watch for new wells drilled outside greater Munich, which would test whether the Type A and Type B permeability patterns hold across the wider basin; no schedule has been published.