Buoyant launch + cold-gas propulsion - built to be audited
A hydrogen balloon or airship lifts a launch platform into the stratosphere (20-40 km), where air density is 1-3% of sea level. From near-vacuum, a small rocket fires with far less atmospheric drag and a modest gravity-loss reduction - the rockoon concept (real, flown in the 1950s) and the airship-to-orbit proposal (JP Aerospace). Cold thrust - a cold-gas thruster venting pressurized H2/He - needs no combustion and is extremely simple, but its low exhaust velocity means it can reach space (the Karman line, suborbital) but not orbit. This lab computes the real physics and states the gaps openly.
Net lift at altitude
0.03 kg/m3
air density 0.036 kg/m3
Lift at sea level
1.03 kg/m3
3% of ground
Balloon volume
17k m3
to lift the mass at altitude
Equivalent sphere
32 m
diameter if spherical
At pressure equilibrium, H2 lift = air density x (1 - MW-H2/MW-air) = air density x 0.93. Buoyancy collapses with altitude because air density falls exponentially - lifting a 500 kg platform to 30 km needs a balloon roughly 32 m across. This is the honest cost of buoyant launch: enormous envelopes for modest payloads, and lift that vanishes the higher you go.
Delta-v to orbit
8.87 km/s
from 30 km
Delta-v from ground
9.80 km/s
for comparison
Savings
-0.93 km/s
-9% vs ground
Delta-v to Karman line
1.17 km/s
suborbital pop
| Propulsion | Isp (s) | Mass ratio m0/mf | Propellant % | Orbit? |
|---|---|---|---|---|
| Cold-gas H2 | 250 | 37.3 | 97% | no |
| H2O2 monopropellant | 160 | 285.5 | 100% | no |
| Chemical bipropellant | 350 | 13.3 | 92% | 2-stage yes |
The Tsiolkovsky equation is unforgiving: cold-gas H2 needs a mass ratio of 37.3(97.3% propellant) to reach orbit - physically impossible because the tanks alone weigh more than that. A chemical stage at Isp 350s needs a ratio of 13.3, achievable only as a multi-stage rocket. Air launch buys ~0.9 km/s - real but modest; it does not make cold thrust orbital.
Delta-v to Karman line
1.17 km/s
from 30 km
Cold-gas H2 mass ratio
1.61
to reach 100 km
Propellant fraction
38%
achievable for a pop
Verdict: cold-gas H2 can reach the Karman line (suborbital space) from 30 km with a mass ratio of1.61 - feasible for a small payload on a one-way pop. But reaching orbit needs8.9 km/s and a mass ratio of 37 (impossible). Cold thrust puts objects into space, not into orbit; orbital insertion still requires a chemical or electric stage.
Double envelope
H2 lift cell inside an inert helium/nitrogen jacket so any leak dilutes below the 4% LEL before reaching air.
Inter-envelope inerting
Continuously purge the space between envelopes with N2/He; alarm if O2 rises above 2%.
Leak detection
Distributed electrochemical H2 sensors alarm at 0.25% of LEL (1,000 ppm); auto-isolate and vent on trigger.
Bonding & grounding
All conductive surfaces bonded and grounded to earth during fill and launch — H2 ignites at <1 mJ of static.
Spark-free materials
No reactive skin doping (the Hindenburg's iron-oxide + aluminum skin was the real igniter); aluminized Mylar/PE only.
Overpressure relief
Burst disks and vent valves rated to 1.25× max differential; zero-pressure balloons vent automatically as they rise.
Remote / uncrewed ops
Fill and launch by remote actuator; no personnel within the blast radius during pressurization.
Over-water launch
Trajectory over ocean or an unpopulated corridor so an in-flight fire cannot reach the ground.
Real-time telemetry
Continuous H2 concentration, envelope pressure, and temperature telemetry with auto-abort-to-vent on threshold breach.
Why hydrogen demands all of this
H2 flammability range 4-75% in air (vs methane 5-15%) - the widest of any fuel. Autoignition 500C, flame 2045C, and the flame is nearly invisible in daylight. A static spark under one millijoule ignites it. H2-air flammability range is 4–75% (vs methane 5–15%) — extremely wide. The flame is nearly invisible in daylight. A static discharge of under 1 mJ can ignite it.
Cold thrust cannot reach orbit. The lab proves cold-gas H2 reaches the Karman line suborbitally (mass ratio 1.61) but needs a ratio of 37 for orbit - physically impossible. Any satellite must carry a chemical or electric orbital stage.
Path to closing: Add a small bipropellant kick stage (Isp 350s) to the payload. From 30 km it needs only a mass ratio of 13.3 - a feasible two-stage design. Alternatively, a solar-electric Hall thruster (Isp 1600s+) reaches orbit with a mass ratio of ~1.76 but requires weeks of spiral-up, limiting it to non-time-critical satellites. The architecture: H2 balloon lifts the stack, cold-gas H2 does the suborbital pop, chemical stage circularizes the orbit.
Air-launch delta-v savings are modest. Launching from 30 km saves ~0.9 km/s (9%) - real but not transformative. The main wins are lower drag, a smaller rocket, and no fixed pad, not a dramatic fuel cut.
Path to closing: Reframe the value proposition. The savings are not in delta-v but in infrastructure and logistics: no fixed launch pad (saves $100M+), mobile launch site (any azimuth, any latitude), reduced range-safety overhead, and a smaller rocket that can be recovered and reused more easily. Target markets where flexibility beats cost-per-kg: responsive launch, polar/sun-synchronous orbits from non-equatorial sites, and suborbital science payloads. The delta-v savings become decisive for a two-stage-to-orbit design where stage 1 is the balloon and stage 2 is a small chemical rocket.
Envelope scaling is unverified. A 500 kg lift at 30 km needs a ~32 m balloon - materials, wind loads, and controlled ascent of a structure this large are unsolved at scale. JP Aerospace's airship-to-orbit is a proposal, not a flown vehicle.
Path to closing: Incremental envelope scaling - start with sounding-balloon payloads (10-50 kg, proven by NASA's CSBF and zero-pressure balloon programs) and scale up in 5x steps. Use a zero-pressure balloon design (vented, not super-pressured) to avoid envelope stress at altitude. Materials: aluminized Mylar/polyethylene film (proven to 40 km by NASA). Solve wind loads by launching at dawn when stratospheric winds are calmest, and by using a guided ascent trajectory that avoids the jet stream (~10-15 km). A 32 m envelope is large but not unprecedented - NASA's ULDB program has flown 1.1 million m3 balloons.
No cryogenic hydrogen handling. This lab models gaseous H2 lift only. Liquid H2 (for a chemical stage) introduces boil-off, cryogenic embrittlement, and venting hazards not modeled here.
Path to closing: Avoid liquid H2 entirely in the buoyant stage. Use gaseous H2 for lift (no cryogenics) and switch to storable propellants (LOX/RP-1 or N2O4/MMH) for the chemical kick stage - these are room-temperature, proven, and compatible with a balloon-launched platform that may loiter for hours before launch. If LH2/LOX is required for performance, use a vacuum-jacketed tank with active boil-off recondensation (cryocooler) sized for the loiter duration, and ground-test at a facility like NASA Stennis or White Sands. Model the boil-off rate (typically 0.1-0.3% per day for good insulation) in the mass budget.
Atmosphere model is ISA exponential (valid to ~50 km); real stratospheric winds, the jet stream, and balloon dynamics require a full flight simulator, not this static estimate.
Path to closing: Integrate a real atmospheric flight simulator. Use the NRL MSISE-00 model (density, temperature, composition to 1000 km) plus HWM-14 wind model (horizontal wind profiles) for trajectory simulation. Couple these to a 6-DOF balloon dynamics solver (buoyancy, drag, wind drift, thermal expansion) validated against NASA CSBF flight data. Add a Monte Carlo dispersion analysis (100+ runs) over seasonal wind variability to define launch windows and safe impact corridors. This turns the static estimate into a launch-ready mission-planning tool - the same methodology used by NASA's Columbia Scientific Balloon Facility.
Three-stage architecture: buoyancy (free) + cold-gas (simple) + chemical (proven). Each stage is independently testable and recoverable.
Can do
Can't do (yet)
Traditional rocket company
Rocket Lab / Astra / Firefly class
Dev to first launch
$250M
5-7 yrs
Fixed launch pad
$100M
concrete + GSE
Cost per launch
$7.5M
Electron class
Cost per kg to LEO
$7,500
small payload
~600 FTE - engine test stands, fuel farms, range safety, launch complex
Buoyant launch (projected)
Early-adopter scenario - not guaranteed
Dev to first launch
$15M
2-3 yrs
Fixed launch pad
$0
mobile - no pad
Cost per launch
$0.5M
balloon + small stage
Cost per kg to LEO
$2,000
projected
~40 FTE - no engine dev, no test stands, no fuel farm. Balloon ops + small chemical stage.
17x cheaper to start - the six reasons why
No rocket engine development
Rocket Lab spent ~$250M and 5 years on the Rutherford engine. Buoyant launch uses off-the-shelf cold-gas valves and a proven chemical kick stage - zero engine R&D.
No fixed launch pad
A traditional pad costs $50-100M in concrete, GSE, and range safety. A balloon needs an open field and a hydrogen truck. Launch from any latitude.
90% lower startup cost
$250M vs $15M - a 17x reduction. Opens space launch to startups, universities, and developing nations.
Responsive launch on demand
Traditional rockets need weeks of pad prep and range coordination. A balloon can be inflated and launched in < 24 hours from any open field.
Environmentally cleaner
A rocket dumps tons of CO2 and soot into the upper atmosphere. A balloon ascent is zero-emission. The only combustion is a small chemical stage above 99% of the atmosphere.
Fully recoverable first stage
The balloon IS the first stage - and it descends gently by parachute after payload release. No ocean recovery fleet, no thermal stress from reentry. Ready for inspection and reuse.
Projected costs are early-adopter estimates based on NASA CSBF balloon hardware costs (~$200k per sounding balloon), small chemical stage costs, and workforce ratios. A chartered engineer must validate before investment.