Physics & Systems

Vol. I · 01 — Design Basis

The reference frame everything is sized against.

A whitepaper proves a concept closes. An engineering volume proves it survives contact with physics. This document fixes the design basis — the assumptions, loads, and budgets — that every later spreadsheet and set design inherits.

The reference city is a single AETHER node sized for a resident-plus-worker population on the order of P ≈ 250,000, descending to a representative mid-strata depth of ≈ 300 m across nine load-bearing rings. All numbers below are order-of-magnitude design targets, not as-built guarantees.

Design basis quantities

Vol. I · 02 — Reactor & Thermodynamics

An exergy budget, not just an energy budget.

The reactor's value to AETHER is not only its electricity but the quality (exergy) of its heat. We track energy and exergy separately because the reuse cascade depends on temperature grade.

Gross balance

The reuse cascade

Waste heat is drawn down by temperature grade so each user receives the lowest grade it can use, maximising total reuse fraction φ:

1. High grade (120–180 °C) → multi-effect desalination & process heat. Allocation ≈ 30% of Q_waste.
2. Mid grade (60–90 °C) → agritech vapour-pressure-deficit control & absorption chilling. ≈ 20%.
3. Low grade (30–45 °C) → district climate & domestic hot water. ≈ 10%.
4. Residual → rejected to the rock shell (§3). With the above, φ ≈ 0.60 and Q_reject ≈ 244 MW.

Exergy check

Exergy of a heat stream at temperature T relative to ambient T₀ is B = Q·(1 − T₀/T). Cascading by grade preserves exergy that a single low-grade dump would destroy — the reason the cascade beats a cooling tower by a wide margin.

Vol. I · 03 — Transient Heat Dissipation

The rock is a slow, enormous heat sink — but it does warm.

Rejecting ~244 MW into the surrounding rock indefinitely is only safe if the rock's temperature rise reaches a tolerable steady state. We treat the shell as radial conduction into a semi-infinite medium.

Steady conduction

Transient time constant

The rock's thermal diffusivity α = k/(ρ·c_p) governs how fast the heated zone spreads. The characteristic penetration depth after time t is:

Because the engaged rock volume grows with √t, the effective thermal resistance rises slowly and the interface temperature trends toward a bounded asymptote rather than running away. Design rule: the rejection annulus must be sized so the worst-case interface stays below the structural temperature limit of the lining over the 100-year refit window; active borehole heat exchangers provide margin.

Vol. I · 04 — Atmosphere & Life Support

An occupied volume that breathes.

Unlike a surface city, AETHER must actively manage its atmosphere. The governing constraint is CO₂ removal at the occupied set-point.

Ventilation is hybrid: fresh-air shafts to the surface cap for baseline, plus closed-loop scrubbing (molecular sieve / amine, reusing low-grade reactor heat for regeneration) for resilience and surge. Air changes target ≥ 6 ACH in occupied strata; agritech vaults double as biological CO₂ sinks and O₂ sources, partially closing the loop.

Vol. I · 05 — Water & Desalination

High-grade heat closes the water loop.

The 30% of waste heat allocated to high-grade use drives multi-effect distillation (MED). Thermal desalination yield scales with the gain-output ratio (GOR) of the plant:

That covers municipal demand at ~220 L/person/day for P=250k (≈ 55,000 m³/day) with the agritech recycling greywater. Brine is managed via deep reinjection or mineral recovery (a secondary export). Potable polishing is by remineralisation + UV.

Vol. I · 06 — Maglev Dynamics

Levitation, propulsion, and the power it costs.

Pods use electrodynamic suspension (EDS) along a passive guideway with a linear synchronous motor (LSM) for propulsion. Levitation force per unit area scales with the square of field strength:

Propulsion & energy per trip

At cruise the pod overcomes mainly aerodynamic and magnetic drag (rolling resistance ≈ 0 when levitated). Tractive power and per-trip energy:

Running the main spine inside a partially evacuated tube (a metro-scale hyperloop) is optional but collapses aero drag, raising practical cruise speed and slashing E_trip — a lever the energy model in Vol. II can toggle.

Vol. I · 07 — Network & Routing

Autonomous flow as a control problem.

The guideway is a directed graph G=(V,E). Demand is a time-varying origin–destination matrix. The controller solves a receding-horizon optimisation each tick, assigning pods to paths to minimise total weighted cost while honouring headway and capacity (Eq. 5.1 in the whitepaper).

Throughput & stability

Vector-field guidance (Eq. 5.2) gives decentralised collision avoidance as a fallback if the central optimiser degrades — a graceful-degradation property required for a life-critical transit spine. Network capacity = Σ lanes × occupancy; freight rides the same graph at off-peak cost.

Vol. I · 08 — Geotechnical & Structural

Thick-walled cylinders, convergence, and seismic comfort.

Overburden at depth h sets the primary load (whitepaper Eq. 6.1): σ_v = ρ_rock·g·h ≈ 7.4 MPa at 300 m. The lining is analysed as a thick-walled cylinder (Lamé) rather than a thin ring for the larger caverns.

Ground–support interaction

Competent rock carries much of the load via arching; the convergence–confinement method sizes support against the ground reaction curve. Where rock quality (RMR/Q) is high, lining demand drops sharply — siting on competent host rock is a first-order capex lever (Vol. II §siting).

Seismic

Deep, embedded structures experience attenuated ground motion versus the free surface (no resonance amplification, motion moves with the rock). The design controls for fault crossing and shaft connections, not for surface-style sway. Per-ring isolation joints localise damage.

Vol. I · 09 — Daylight & Circadian

Real sunlight, piped 300 m down.

Surface heliostats/concentrators couple sunlight into large-core fibre or mirrored light-guides. Delivered illuminance at a terrace falls with collector area, optical efficiency, and guide losses:

Where piping daylight is impractical, circadian LED lighting reproduces the daily colour-temperature curve. The combination preserves the diurnal cues that underground habitation otherwise destroys — a habitability requirement, not a luxury.

Vol. I · 10 — Power & Redundancy

One reactor is a single point of failure — so it isn't the only thing.

Baseload is the SMR, but the grid is engineered N+1: battery/flywheel storage for ride-through, an islanding controller, and a surface grid-tie that runs both ways (import on outage, export surplus for revenue — Vol. II §revenue). Critical life-support and transit buses have independent backup.

• Ride-through: storage covers the reactor trip-to-backup gap; life support never loses power.
• Islanding: the city can disconnect from the surface grid and run autonomously.
• Bidirectional tie: surplus baseload is sold; deficits are covered — the tie is a financial hedge, not just a safety net.
• Refuel/refit: SMR modular design enables maintenance without full city shutdown.

Vol. I · 11 — Construction Sequence

Bored, not built — and timed by TBM throughput.

Schedule is gated by tunnel-boring-machine advance rate and shaft logistics. Rough excavation time for a ring scales with bored volume over effective advance:

Sequence: (1) surface cap + main shaft, (2) reactor cavern + first strata in parallel, (3) downward strata expansion funded by token pre-sales as completed rings come online, (4) maglev spine commissioning. The phased fill is what lets revenue (Vol. II) begin before the city is complete.

Vol. I · 12 — Engineering Risk Register

The honest list of what could break.

Each row above is a candidate for its own production sub-volume. The point of this register is intellectual honesty: AETHER's hardest problems are thermal saturation, nuclear licensing, and squeezing ground — and each has a credible, physics-grounded mitigation.