Haute-Sorne's Quiet Number: M0.3, and Why a Swiss EGS Well Didn't Shake Anyone
Switzerland's Haute-Sorne enhanced-geothermal well cleared stimulation at a felt-maximum of M0.3 — 60x under Basel 2006. What the numbers mean for your desk.
Geo-Energie Jura AG has closed the stimulation phase of the Haute-Sorne enhanced geothermal (EGS) project in the Jura, and the headline number is the one that didn’t happen: no tremor felt at the surface. Between 24 June and 31 July 2025, downhole seismometers logged 3,387 seismic events. Only four reached the Swiss Seismological Service (SED) at ETH Zurich’s independent surface network. The strongest was magnitude 0.3.
Hold that against the reference case every geothermal engineer in the DACH region carries: Basel, December 2006, a Deep Heat Mining well that induced an M3.4, rattled the city, and killed the project. Haute-Sorne came in more than three magnitude units quieter.
What the stimulation did
EGS is a permeability problem, not a heat problem. The heat is simply there — the crew measured 135 °C at 3,986 m, on trend for the 150 °C-plus expected at 4,500–5,000 m. The hard part is coaxing water through hot but tight crystalline rock. So you inject: the operators ramped to 80 l/min at 270 bar, pushing 430 m³ before opening the well to relieve pressure. Fluid pressure lowers the effective stress on existing faults; they slip; that slip is your seismicity and, if you’re lucky, your new fracture permeability.
The control loop is the engineering. A traffic-light protocol paused injection at M0.3, would have stopped it permanently at M0.9, and vented pressure at M1.5. Eight seismometers sat at the bottom of the borehole — calibrated, per thinkgeoenergy’s account, with a vibrating truck — and the injection pumps were the ones ETH Zurich runs at its Val Bedretto underground lab. Injection stopped on 14 July when the threshold was reached twice. That is the system working, not failing.
←TODAY: A Jura EGS well cleared stimulation at a felt-maximum of M0.3, sixty-odd times weaker than Basel 2006. →3012: The Zurich-3012 grid runs on heat mined under inhabited ground, governed by seismic control loops nobody notices. Fulcrum: Deep geothermal under towns only scales because the monitoring got dense enough to trade megawatts against millimetres of fault slip in real time.
Why this reaches your desk
Because the reservoir is under a village, not a desert. Seismicity decayed to two or three microseisms per day once injection stopped — textbook post-shut-in relaxation. For anyone assessing structures above a future EGS field, the live question is no longer “will it shake?” but “what magnitude, how often, against which fragility curve?” That is a digital-twin conversation: a monitored asset coupled to a hazard model that recommends an action under uncertainty, not a dashboard that merely shows telemetry.
Atelier: Offices that have sat near induced-seismicity projects report the same gap — the structural team learns the seismic protocol only after the drilling permit is public. This Monday, if any deep-geothermal or wastewater-injection project sits within a few kilometres of a site you’re carrying, pull its published traffic-light thresholds and set the design-basis induced magnitude next to your building’s code seismic demand. Usually the induced number is trivially small — but you want that on paper before a client asks.
Hack: Convert a magnitude into the energy that reaches your foundations, so “M0.3 versus M3.4” stops being an abstraction. The Gutenberg-Richter radiated-energy relation puts each magnitude unit at about 31.6× the energy. Run it:
import math
def energy_j(M): # radiated seismic energy, joules
return 10 ** (1.5 * M + 4.8)
print(energy_j(3.4) / energy_j(0.3)) # Basel 2006 vs Haute-Sorne
That ratio comes back near 90,000 — Basel released roughly five orders of magnitude more energy than the strongest Haute-Sorne tremor, which is why one was a civic incident and the other a line in a log.
The move
Keep one number for every wonder-claim. The frontier results my generation scaled wrongly were never the ones that failed in the lab; they were the ones deployed at city scale before anyone read the error bar. Haute-Sorne got the arithmetic right — dense array, honest thresholds, an independent monitor at ETH. Before you specify anything downstream of deep geothermal, ask for that error bar and ask who watched the seismometers.
Source: thinkgeoenergy.com
SOURCE · ↗
PAZ Kaffi · multidisciplinary editorial, led by PAZ Academy