Every answer and its explanation appears here once you have finished the path. Each one then links to the matching glossary entry, where the concept is set out in full with its worked example.
1. The space sector moves from a few costly satellites to constellations of thousands of cheap craft. What shift does that produce in the insured risk?
From the major unit loss towards fleet and accumulation logic
NewSpace names the sector's transformation driven by private operators, drastically lower launch costs, reusable launchers, satellite miniaturisation and the deployment of massive constellations. It breaks with the historic model of large state programmes and a few very costly craft, and what it changes for an insurer is not the level of risk but its shape. The proliferation of cheap small satellites alters the ratio between unit value and number of objects: the loss to watch stops being the destruction of one major asset and becomes accumulation across a fleet. It would be wrong to conclude that exposure falls because each object is worth less, and that is the error the question seeks: one launch campaign can carry dozens of craft, hence concentrate considerable value in a single event. Two further consequences accompany the move and are worth knowing: growing orbital congestion sharpens collision and debris risk, degrading the environment for every operator at the same altitude; and new entrants arrive, sometimes thinly capitalised and with a short reliability record, making selection harder than when every manufacturer was known.
Glossary entry · newspace-assurance2. One first stage exceeded several dozen successful flights during the 2020s. Why did reusability first attract a caution loading, even on launchers already proven in expendable form?
Two opposite hypotheses on reliability, wear or maturity, that no record could settle
Reuse changes the nature of the statistical record on which pricing rests, and that is where the whole difficulty lies. An expendable stage is new on every flight, so each launch samples the same population; a reused stage accumulates cycles, and its reliability can move in two contrary directions. It can degrade with wear on structures and engines, which is the first intuition. It can instead improve, because a flown article has demonstrated that it works and because refurbishment procedures mature, which is the opposite intuition and no less defensible. Only a long enough flight record settles it, and at a programme's start that record does not exist. Insurers therefore apply a caution loading, including for a launcher whose expendable version was proven, because the missing data concerns precisely the new practice. What matters here reaches beyond space: it is the position of any risk where a technique must be priced before its loss experience exists, and the way out is not better reasoning but accumulated flights, which eventually made this launcher one of the cheapest to insure in the commercial market.
Glossary entry · assurance-lanceur-reutilisable3. A single launch can carry more than fifty small satellites from one programme. How does an insurer then structure the cover?
As a fleet portfolio, with an expected per-batch failure rate as an actuarial input
Insuring a constellation satellite by satellite would mean writing hundreds or thousands of policies over a homogeneous population, costing much in administration to learn nothing more. The model that prevailed structures cover as a fleet portfolio, and that is not an administrative simplification but a change of reasoning. The insurer no longer seeks the failure probability of a single high-value object, it works on a statistical distribution of losses: a failure rate of a few per cent per batch becomes an expected actuarial input rather than an exceptional loss, and it enters the price as a normal charge. What remains hard, and what the question highlights, is launch concentration: one shot carrying dozens of craft concentrates in a single event a large fraction of the programme's insured value, so diversification by number does not help at the very moment it would matter most. Reasoning by batch is therefore both the right method and the acknowledgement of an accumulation that cannot be spread, since it is the customer who decides how to load a launcher.
Glossary entry · assurance-constellation4. In February 2022, a geomagnetic storm described as moderate caused the loss of nearly forty of the forty-nine satellites just launched, at an estimated cost of about fifty million dollars. By what mechanism?
The upper atmosphere expanded, raising drag before orbit raising
This peril acts with no direct physical impact, which is what makes it counter-intuitive. A solar flare or coronal mass ejection produces a geomagnetic storm that heats and expands the upper atmosphere; the air is then denser where satellites have just been released, the drag they experience rises, and low-orbit craft lose altitude instead of gaining it. The February 2022 episode became the teaching case because it combines three striking elements: the storm was described as moderate, the satellites were at the most vulnerable moment of their lives, just after release and before reaching operational orbit, and the loss was nearly total across the batch. One also reads in it the NewSpace trait from the previous question, since a whole batch disappears rather than one satellite. It should be added that solar activity can harm by other routes, damaging onboard electronics through induced surges or disrupting communications and navigation, but that is not what happened here: the cause was mechanical, and it lay in the atmosphere rather than in the satellite.
Glossary entry · risque-meteorologique-spatial5. Between 2017 and 2019, several satellites of one all-electric propulsion family suffered leaks in their propulsion systems. Settlements were for partial loss rather than destruction. Why does this peril worry insurers particularly?
It is gradual, shortening useful life without destroying, and the technique is thinly proven over time
Electric propulsion holds a satellite on station and manages its orbit changes, and its adoption follows from two real advantages: it is lighter and more fuel-efficient than chemical propulsion, freeing mass for payload. Its drawback lies elsewhere, and it is twofold. Failure there is often gradual, a valve leak, cathode degradation or loss of redundancy reducing manoeuvring capacity without causing immediate loss, but sharply shortening the expected operational life. That is exactly the loss profile the partial loss clause exists to handle, and the 2017 to 2019 settlements were made on that basis rather than as total losses. The second drawback is one of knowledge: the technique is less proven over long periods than chemical propulsion, so a failure mode can emerge after years of service, when the exposed fleet is already large. Here again is the difficulty from the reusable launcher question: it is not the risk that is poorly understood, it is the record that is missing, and it is made only by waiting.
Glossary entry · defaillance-propulsion-electrique6. In 2020, a specialised vehicle docked with a geostationary satellite out of fuel and took over its propulsion, extending its life by about five years. What does that operation add to the risk?
A docking between two objects not designed for each other, whose failure can destroy both
In-orbit servicing has a specialised vehicle dock with a satellite at end of life, generally out of fuel, to take over its station-keeping and attitude control, extending its operation without replacement. The operation changes the risk at two levels that must be told apart. The first is technical and immediate: two objects not designed for one another are docked in orbit, an unprecedented and high-risk manoeuvre whose failure can damage or destroy both craft, doubling the stake instead of moving it. The second is economic and pulls the other way: rather than bearing the cost and risk of a new launch to replace a working but dry satellite, the operator extends an already amortised asset, removing a launch and the launch peril with it. The net outcome depends entirely on docking reliability, and that is what makes this activity worth following: every successful mission moves the needle, like the accumulated flights of a reusable launcher. The 2020 case is worth remembering because it demonstrated feasibility on a commercial satellite in service, not on a technology demonstrator.
Glossary entry · service-en-orbite7. A construction contract withholds fifteen per cent of the price, paid to the manufacturer over fifteen years of service subject to performance. What does that mechanism align, which the sale alone did not?
The manufacturer's interest with actual reliability, long after title transfers
In an ordinary sale the manufacturer's interest stops at delivery: it is paid, title passes, and what happens next concerns it only through a limited contractual warranty. Space has a peculiarity that makes that arrangement ill-fitting: most design or manufacturing failures show themselves not at launch but during the service years, when the satellite is already far away and already paid for. In-orbit incentives correct this by withholding at delivery a share of the price, of the order of ten to twenty per cent, released progressively over the operational life subject to performance. If the craft degrades early, losing transponders, shedding power, or seeing its life cut short by a propulsion failure, the remaining payments are reduced pro rata. The point worth seeing, because it often surprises, is that the mechanism is independent of insurance: it operates between operator and manufacturer, whatever settlement is received elsewhere. So two arrangements look at the same event without substituting for each other, one indemnifying a loss, the other adjusting a price, and a professional must read both in the same file.
Glossary entry · incitation-en-orbite8. Since 2020, several uncontrolled re-entries of a roughly twenty-tonne first stage have been tracked worldwide without the impact point being predictable more than a few hours ahead. What does that uncertainty impose on the analysis?
Reasoning in statistical probability of damage, no targeted evacuation being possible
The great majority of small objects burn up entirely on re-entry, which is why this peril stays marginal in numbers. It stops being marginal for high-mass stages or refractory components, which can reach the ground undestroyed, and a stage of some twenty tonnes clearly belongs to that category. What distinguishes this risk from other space perils is not its severity but its geographic uncertainty: an uncontrolled re-entry can be predicted only a few hours ahead and across a wide track, so no targeted evacuation is possible and the question shifts to a statistical probability of damage rather than a location. This is a rare situation in insurance, where one usually knows where the risk sits even without knowing when it will happen; here one roughly knows the when and does not know the where. The legal regime is finally the launching state's absolute liability for damage on the ground, laid down by the 1972 convention, meaning a victim need not prove fault and the claim is settled between states.
Glossary entry · risque-rentree-atmospherique9. On 24 February 2022, a software attack paralysed the modems of a satellite network in Europe, cutting off tens of thousands of users and disrupting the remote control of thousands of wind turbines along the way. No satellite was physically touched. What question does that raise?
Which cover applies, the service being cut with no physical space loss
A space system is not merely the satellite: it includes the ground control segment and user terminals, and those two are connected to ordinary terrestrial networks, often less protected than the craft in orbit. That is where the preferred attack surface lies, and the February 2022 episode showed it most sharply: compromising customer modems sufficed to cut the service, with no physical space loss recorded. The question of which cover applies then arises directly, since space policies are built around damage to the craft and the business interruption flowing from it, whereas here there is no damage to the craft. This is exactly the silent cyber difficulty, transposed to another line. Two elements are worth adding. The propagation to wind turbines shows that dependence on a satellite service hides inside activities that are in no way space-related, which is an accumulation problem for an insurer. And attacks on command links, with a theoretical risk of taking control of a satellite, remain rare but are taken seriously, which would move the subject from the service to the integrity of the asset itself.
Glossary entry · risque-cyber-spatial