Case Study: Why the Space Station Keeps Falling (and How It's Pushed Back)

"The Station is always falling. Our job is to keep it falling around the Earth, not into it." — a flight controller's way of putting it

Executive Summary

The International Space Station is the largest object humans have ever assembled in orbit, and it is a perfect specimen for this chapter because it disobeys, very slightly, the tidy promise of Section 2.1 that "orbits don't decay." The ISS orbits low enough — around $420\ \text{km}$ — that it plows through a whisper of residual atmosphere, and Newton's first law bites: that faint drag is an external force, and it steadily removes the station's momentum, dropping it toward the Earth. Left alone, the ISS would re-enter within a year or two. It survives because, every few weeks, a docked spacecraft fires its engines and gives the station a small forward shove — a reboost — restoring the momentum that the atmosphere stole. In this case study we use nothing but Newton's laws to reconstruct the whole cycle: the station's orbital momentum (from $\mu$), the force of drag that erodes it (first and second laws), and the thrust and propellant of the reboost that repays it (second and third laws, and conservation of momentum).

Skills applied

• Computing orbital speed and momentum from the gravitational parameter (§2.5).
• Recognizing residual drag as the external force that makes a real orbit decay (§2.1).
• Relating force, momentum, and impulse for a variable maneuver ($T = \dot m_p v_e$, impulse $= m\,\Delta v$) (§2.2, §2.4).
• Sanity-checking every figure against reality — the habit the whole book insists on.

Background

The vehicle and its orbit

We will use round, widely reported figures (Tier 2, and Tier 3 for the atmospheric inputs, which vary enormously with solar activity and the station's orientation):

The physical question is simple to state: how hard does the thin air push back on the station, how fast does that push bleed away its motion, and what does it take to put the motion back?

Phase 1: The station's speed and momentum

First, how fast is the ISS moving? Section 2.5 gives the circular orbital speed as $v = \sqrt{\mu/r}$ (we derive this properly in Chapter 6; for now take it on trust as the speed that balances gravity against a circular path). With $\mu_\oplus = 3.986\times10^{14}\ \text{m}^3/\text{s}^2$ and $r = 6.791\times10^{6}\ \text{m}$:

$$ v = \sqrt{\frac{3.986\times10^{14}}{6.791\times10^{6}}} = \sqrt{5.869\times10^{7}} = 7{,}661\ \text{m/s} \approx 7.66\ \text{km/s}. $$

That is about $27{,}600\ \text{km/h}$ — the ISS crosses a continent in minutes. Its momentum, the quantity Section 2.4 taught us to track, is

$$ p = m v = 420{,}000\ \text{kg} \times 7{,}661\ \text{m/s} = 3.22\times10^{9}\ \text{kg}\cdot\text{m/s}. $$

Three billion kilogram-metres per second is an enormous store of motion. It is exactly this reservoir that drag slowly drains, and that the reboost tops back up.

Phase 2: The force that never lets up

At $420\ \text{km}$ the "vacuum" is not empty; it holds a few million molecules per cubic centimetre — unimaginably thin, but the station is moving through it at $7.66\ \text{km/s}$, and drag grows with the square of speed. The drag force is

$$ F_{\text{drag}} = \tfrac{1}{2}\,\rho\, v^2\, C_d\, A. $$

Plugging in our numbers, and keeping the units in view:

$$ F_{\text{drag}} = \tfrac{1}{2}\,(4\times10^{-12}\ \tfrac{\text{kg}}{\text{m}^3})\,(7{,}661\ \tfrac{\text{m}}{\text{s}})^2\,(2.2)\,(2{,}000\ \text{m}^2) \approx 0.5\ \text{N}. $$

Half a newton. That is roughly the force you feel holding a small apple against Earth's gravity — and it is the entire force decelerating a 420-tonne space station. It sounds absurdly small, and it is; but Newton's first law does not care how small a force is, only that it is there. A force with no opposition acts forever, and forever adds up.

🔧 Engineering Reality: The single most uncertain number in this study is the air density $\rho$. When the Sun is active, it heats and puffs up the upper atmosphere, and the density at $420\ \text{km}$ can rise by a factor of ten — so the drag, and the reboost propellant, rise with it. This is why station planners watch space weather as closely as a sailor watches the sky: the "unforgiving environment" of this book's Theme 2 includes an atmosphere that reaches higher, and grabs harder, exactly when the Sun flares.

Phase 3: How fast the momentum bleeds away

A force of $0.5\ \text{N}$ on a $420{,}000\ \text{kg}$ body produces a deceleration of

$$ a = \frac{F_{\text{drag}}}{m} = \frac{0.5\ \text{N}}{420{,}000\ \text{kg}} = 1.2\times10^{-6}\ \text{m/s}^2. $$

Tiny — a millionth of a $g$. But run it for a full day ($86{,}400\ \text{s}$) and the lost speed is

$$ \Delta v_{\text{day}} = a\,t = (1.2\times10^{-6}\ \text{m/s}^2)(86{,}400\ \text{s}) \approx 0.10\ \text{m/s per day}. $$

Over a month that is roughly $3\ \text{m/s}$ of speed quietly removed, and — through the counterintuitive bookkeeping of orbits — the station sinks a few kilometres closer to Earth each month. (The full accounting of why a decaying orbit actually speeds the satellite up, even as drag removes energy, is a subtlety we save for Chapter 12; here we simply track the momentum the reboost must replace.) The lesson of Section 2.1 stands, sharpened: in a true vacuum the orbit is eternal, but in the real, slightly-gassy vacuum of LEO, "eternal" comes with a monthly bill.

Phase 4: Paying the bill — the reboost

To stop the station falling, you must give back the momentum drag took. This is Newton's third law put to work: a docked cargo ship (a Progress, a Cygnus) fires its engine, throwing exhaust backward, and the reaction pushes the whole station forward — exactly the "throw mass to move" principle of Section 2.3.

Suppose we let a month's loss accumulate and restore it in one burn: $\Delta v = 3\ \text{m/s}$. The impulse required is the change in momentum,

$$ J = m\,\Delta v = 420{,}000\ \text{kg} \times 3\ \text{m/s} = 1.26\times10^{6}\ \text{N}\cdot\text{s}. $$

A Progress main engine produces a thrust of about $T = 3\ \text{kN}$. Since impulse is thrust times time, the burn must last

$$ t = \frac{J}{T} = \frac{1.26\times10^{6}\ \text{N}\cdot\text{s}}{3{,}000\ \text{N}} = 420\ \text{s} \approx 7\ \text{minutes}, $$

which matches the real thing: ISS reboosts are burns of several minutes. And how much propellant does that cost? From Section 2.2, thrust is $T = \dot m_p v_e$, so with a storable propellant at $v_e \approx 2{,}900\ \text{m/s}$ (specific impulse near $300\ \text{s}$),

$$ \dot m_p = \frac{T}{v_e} = \frac{3{,}000\ \text{N}}{2{,}900\ \text{m/s}} \approx 1.03\ \text{kg/s}, \qquad m_p = \dot m_p\, t = 1.03 \times 420 \approx 430\ \text{kg}. $$

About $430\ \text{kg}$ of propellant buys back one month of drag.

Phase 5: The yearly bill and a sanity check

Scaling to a year: roughly $12$ such reboosts, restoring about $37\ \text{m/s}$ of speed and burning on the order of $5{,}000\ \text{kg}$ of propellant. Widely reported figures put the ISS's annual reboost propellant in the several-thousand-to-$7{,}000\ \text{kg}$ range, rising in years of high solar activity — so our from-scratch estimate, built on nothing but Newton's laws and one drag equation, lands in the right ballpark. That agreement is the point of the exercise: the same three laws that describe a thrown wrench describe, quantitatively, the maintenance schedule of a $420$-tonne spacecraft.

# Reboost bookkeeping (illustrative; hand-traced, not run).
m_iss   = 420_000     # kg
dv      = 3.0         # m/s to restore one month of drag loss
thrust  = 3_000       # N  (Progress-class main engine)
ve      = 2_900       # m/s (storable propellant, Isp ~ 300 s)

impulse   = m_iss * dv          # N*s  = kg*m/s
burn_time = impulse / thrust    # s
mdot      = thrust / ve         # kg/s
prop_used = mdot * burn_time    # kg
print(round(impulse), round(burn_time), round(prop_used))
# Expected output:
# 1260000 420 434

🔗 Connection: Notice that we never once used the rocket equation — we treated the $3\ \text{m/s}$ reboost as a small, near-constant-mass nudge, so impulse $= m\,\Delta v$ was accurate enough. That shortcut works because the propellant burned ($430\ \text{kg}$) is a whisker of the station's mass. When the velocity change is not small compared with the exhaust velocity — a launch, a Mars injection — the shrinking mass matters, and you must use $\Delta v = v_e \ln(m_0/m_f)$ from Chapter 3. Knowing when the simple momentum bookkeeping suffices, and when it doesn't, is a real engineering judgment.

Discussion Questions

1. Drag on the ISS is only about half a newton, yet it forces multi-tonne propellant deliveries every year. Explain, using the first law, why a tiny force has such a large cumulative cost.
2. During a solar maximum the upper-atmosphere density can rise tenfold. What happens to the drag force, and roughly what happens to the annual reboost propellant?
3. The reboost pushes the station forward by throwing exhaust backward. Identify the action-reaction pair explicitly, and state what would happen to the station's momentum if the exhaust somehow carried none.
4. Why is a reboost delivered as a burn of several minutes at a few kilonewtons, rather than an instant kick? Relate your answer to impulse, $J = T\,t$.

Your Turn: Extensions

• Option A (analysis). Recompute the drag force and the monthly $\Delta v$ loss for an orbit at $340\ \text{km}$ instead of $420\ \text{km}$, taking the density there to be $3\times$ higher. How much more often would this lower station need reboosting? Comment on the altitude-versus-drag trade.
• Option B (computation). Extend the Python snippet to a function reboost(m, dv, thrust, ve) that returns impulse, burn time, and propellant mass, then tabulate the propellant for $\Delta v = 1, 3,$ and $6\ \text{m/s}$. (Do not run it; hand-trace and add # Expected output:.)
• Option C (design). A future station masses $1{,}000\ \text{t}$ and orbits at $500\ \text{km}$ (roughly a quarter the drag per unit area). Estimate its annual reboost propellant and argue whether a higher, heavier station is easier or harder to keep aloft per tonne.

Key Takeaways

1. A real low orbit decays because the vacuum isn't perfect. Residual drag is a genuine external force, and by the first law even $0.5\ \text{N}$, unopposed, removes momentum without end.
2. Momentum bookkeeping runs the maintenance. Orbital momentum comes from $\mu$; the reboost impulse is $m\,\Delta v$; the propellant follows from $T = \dot m_p v_e$ — all pure Chapter 2.
3. The third law pays the bill. The station stays up only by continually throwing mass backward; there is no other way, in vacuum, to replace the momentum drag steals.
4. Newton's laws are quantitative, not just conceptual. Fed real numbers, they reproduce the ISS's real, multi-tonne annual propellant appetite to the right order of magnitude.