Case Study: Why Ariane Launches from the Jungle — Auditing the GEO Plane Change

"A degree of inclination is not free. Out at the geostationary belt it is measured in kilometers per second, and kilometers per second are measured in years of a satellite's life."

Executive Summary

Every geostationary communications satellite has to end up in the equatorial plane — inclination $0^\circ$ — because that is the only plane in which a satellite can hover motionless over a fixed point on Earth. But a rocket cannot launch into a plane more equatorial than the latitude of its launch site. Launch from Cape Canaveral at $28.5^\circ$ north and the cheapest orbit you can reach is inclined $28.5^\circ$; launch from Europe's Guiana Space Centre at Kourou, only $5.2^\circ$ from the equator, and you start almost in the right plane already. That single geographic fact, filtered through the plane-change relation $\Delta v = 2v\sin(\Delta i/2)$ of Section 10.4, is worth hundreds of meters per second — and, downstream, years of revenue.

In this case study we audit an existing decision: the choice of launch site for a GEO mission. We reconstruct the plane change each site imposes, compute the combined apogee burn (Section 10.6) at each, and translate the delta-v difference into what operators actually care about — propellant, and therefore operational lifetime. The numbers are representative of real GTO deliveries by Falcon-9-class vehicles from the Cape and Ariane-5-class vehicles from Kourou (Tier 2 — inclinations and station-keeping rates are rounded and vary by mission).

Skills applied

• The plane-change cost relation $\Delta v = 2v\sin(\Delta i/2)$ (§10.4).
• The combined circularize-plus-turn burn via the law of cosines (§10.6).
• Reading a maneuver delta-v as a difference of vis-viva speeds (§10.2, from Ch. 6).
• Converting a delta-v difference into propellant and mission lifetime through the rocket equation (Ch. 3) and station-keeping (a preview of Ch. 12).

Background

A geostationary transfer orbit (GTO) is the same LEO-perigee-to-GEO-apogee ellipse we have used since Chapter 6: $r_p \approx 6{,}771\ \text{km}$, $r_a = r_{\text{GEO}} = 42{,}164\ \text{km}$, semi-major axis $a_t = 24{,}467.5\ \text{km}$. Its apogee speed (from vis-viva) is $v_a = 1.617\ \text{km/s}$, and a circular GEO needs $v_{c2} = 3.075\ \text{km/s}$ — both independent of the launch site, because the ellipse is the same. What the launch site changes is the inclination of that ellipse, and hence the plane change the satellite must buy at apogee:

• Cape Canaveral: minimum inclination $\Delta i \approx 28.5^\circ$.
• Kourou: minimum inclination $\Delta i \approx 6^\circ$ (Ariane's standard GTO; the site's $5.2^\circ$ latitude plus a small dogleg).

The perigee injection burn (LEO onto GTO) is essentially the same $\approx 2.40\ \text{km/s}$ from either site, so we hold it fixed and audit only the apogee burn, where the plane change lives.

Phase 1: The plane change as a standalone cost

First, price the plane change by itself at GEO speed ($v = 3.075\ \text{km/s}$), using $\Delta v = 2v\sin(\Delta i/2)$:

$$ \text{Cape } (28.5^\circ):\quad \Delta v = 2(3.075)\sin(14.25^\circ) = 6.150 \times 0.2462 = 1.514\ \text{km/s}, $$ $$ \text{Kourou } (6^\circ):\quad \Delta v = 2(3.075)\sin(3^\circ) = 6.150 \times 0.05234 = 0.322\ \text{km/s}. $$

The Cape's plane change alone ($1.514\ \text{km/s}$) is more than the entire circularization burn ($1.458\ \text{km/s}$); Kourou's ($0.322\ \text{km/s}$) is a fifth of that. A difference of nearly $1.2\ \text{km/s}$ hangs on where the rocket left the ground.

Phase 2: The combined apogee burn (the smart way)

No competent operator does the plane change separately — they fold it into the circularization burn at apogee, as Section 10.6 prescribes, using the law of cosines on the arrival velocity $v_a$ (in the inclined plane) and the target velocity $v_{c2}$ (equatorial), separated by $\Delta i$:

$$ \Delta v_{\text{combined}} = \sqrt{v_a^2 + v_{c2}^2 - 2\,v_a\,v_{c2}\cos\Delta i}. $$

We can compute both sites with one small function:

import math

def combined_apogee(v_a, v_geo, di_deg):
    """Single burn that circularizes AND changes plane at GTO apogee (km/s)."""
    di = math.radians(di_deg)
    return math.sqrt(v_a**2 + v_geo**2 - 2 * v_a * v_geo * math.cos(di))

V_A, V_GEO = 1.617, 3.075
cape   = combined_apogee(V_A, V_GEO, 28.5)
kourou = combined_apogee(V_A, V_GEO, 6.0)
print(round(cape, 3))
print(round(kourou, 3))
print(round(cape - kourou, 3))
# Expected output:
# 1.825
# 1.477
# 0.348

So the combined apogee burn is $1.825\ \text{km/s}$ from the Cape and $1.477\ \text{km/s}$ from Kourou. The launch-site penalty, even after the combined-maneuver trick has done its best, is

$$ \Delta v_{\text{site}} = 1.825 - 1.477 = 0.348\ \text{km/s} = 348\ \text{m/s}. $$

Sanity check. Both combined burns must exceed the pure circularization cost of $1.458\ \text{km/s}$ (you still have to speed up from $1.617$ to $3.075\ \text{km/s}$ and turn), and they do: $1.825$ and $1.477\ \text{km/s}$. Kourou's is only $0.019\ \text{km/s}$ above the no-turn floor, because $6^\circ$ is a nearly negligible turn; the Cape's sits well above it. The ordering and magnitudes are exactly what the geometry demands.

Phase 3: Turning delta-v into years of life

A GEO satellite carries a fixed tank of propellant. Whatever it does not spend reaching its slot, it keeps for station-keeping — the small yearly burns that fight the perturbations (Chapter 12) trying to drag it out of position. The dominant cost, north-south station-keeping against the Sun and Moon's pull on the orbital plane, runs about $45$–$50\ \text{m/s}$ per year.

So the $348\ \text{m/s}$ that a Kourou launch saves on the apogee burn can instead be banked for station-keeping:

$$ \frac{348\ \text{m/s}}{\sim 48\ \text{m/s/yr}} \approx 7\ \text{years of extra life.} $$

For a satellite whose business case is measured in revenue-years, seven extra years is enormous — often the difference between a marginal mission and a profitable one. (FLAG: this conversion is illustrative — real satellites size propellant to a fixed design life, so the saving more often shows up as a lighter, cheaper satellite or a larger payload than as literal extra years. Either way, the $348\ \text{m/s}$ is real money.)

🔧 Engineering Reality: electric propulsion is rewriting this audit. Modern all-electric comsats circularize with ion or Hall thrusters (Chapter 20) at $I_{sp}$ ten times a chemical engine's, so the plane-change delta-v — still governed by the same $2v\sin(\Delta i/2)$ — costs far less propellant, even though the delta-v is identical. The trade is time: electric raising to GEO takes months of slow spiral, not five hours. The launch-site advantage shrinks in propellant terms but the delta-v arithmetic of this audit is unchanged; physics sets the delta-v, the engine sets what that delta-v costs in mass.

Phase 4: The energy view, as a cross-check

We can confirm the story in the language of Chapter 6. The apogee burn does two things: it adds the energy to climb from the transfer ellipse ($\varepsilon_t = -\mu/2a_t = -8.15\ \text{MJ/kg}$) up to the GEO circle ($\varepsilon_{\text{GEO}} = -4.73\ \text{MJ/kg}$), a gain of $3.42\ \text{MJ/kg}$ that is identical for both sites — and it redirects the velocity into the equatorial plane, which adds no energy at all (a plane change keeps the speed the same). So the entire launch-site penalty is spent on turning, not on climbing: it buys zero extra orbital energy. That is precisely why plane changes feel so wasteful — every one of those $348\ \text{m/s}$ goes into rotating a velocity vector the satellite already had, and none of it into raising the orbit.

Discussion Questions

1. The apogee plane change adds no orbital energy, yet costs hundreds of m/s. Reconcile these two facts using the vector picture of a plane change from Section 10.4.
2. Why is the perigee injection burn ($\approx 2.40\ \text{km/s}$) roughly the same from both launch sites, while the apogee burn differs by $0.348\ \text{km/s}$?
3. Kourou's combined burn ($1.477\ \text{km/s}$) is barely above the no-plane-change circularization cost ($1.458\ \text{km/s}$). Why does a $6^\circ$ turn cost so little when combined?
4. If a Cape launch could somehow do its $28.5^\circ$ plane change down in LEO instead of at GTO apogee, how much would that cost (use $v = 7.673\ \text{km/s}$), and why would that be a terrible idea?

Your Turn: Extensions

• Option A (analysis). Baikonur Cosmodrome sits at $45.6^\circ$ north — the highest-latitude major spaceport. Compute the combined GTO-apogee burn for a $46^\circ$ plane change and compare with the Cape and Kourou. How much of a lifetime penalty does Baikonur's latitude impose on a GEO mission?
• Option B (computation). Extend the combined_apogee function to also return the separate cost (circularize then turn), and print the saving from combining for each of the three sites (Kourou $6^\circ$, Cape $28.5^\circ$, Baikonur $46^\circ$). Do not run it — hand-trace and add # Expected output:.
• Option C (your mission). If your mission (Track A/B/C/D) ends in an inclined-to-equatorial transition or any plane change, audit it: compute the plane change at both the fast (low) and slow (high) points of your trajectory, and record in your MDR which one you will use and why.

Key Takeaways

1. Launch latitude sets the minimum inclination, and inclination you cannot avoid becomes a plane change you must pay for: $\Delta v = 2v\sin(\Delta i/2)$.
2. Even after the combined-maneuver trick, a Cape ($28.5^\circ$) GEO delivery costs about $0.348\ \text{km/s}$ more than a Kourou ($6^\circ$) one — the launch site is a first-order design choice, not a detail.
3. That delta-v is real money: banked as station-keeping it is worth years of a satellite's operational life (or, equivalently, a lighter and cheaper spacecraft).
4. The plane-change delta-v buys no orbital energy — it is spent entirely on turning the velocity vector — which is exactly why the discipline works so hard to avoid it, and why "combine and defer to apogee" is the universal rule.