NASA unveiled the contracts at Kennedy Space Center on Nov. 7 in the flat language procurement lawyers prefer, with the number that matters buried in the ceiling. The program, CLPS 2, runs for ten years, carries a maximum value of $10 billion, and places eight companies in a pool that will compete for individual delivery orders rather than for a single award. The eight are SpaceX, Blue Origin, Firefly Aerospace, Intuitive Machines, Lockheed Martin, Northrop Grumman and two others the agency did not single out. Each will bid on task orders, and NASA will choose per mission rather than per company. The existing program, CLPS 1.0, had its own ceiling raised from $2.6 billion to $4.2 billion, which says something about how much demand accumulated under the original figure.
The architecture is the part worth understanding. Under the first CLPS, NASA bought lunar delivery as a service: a company builds a lander with its own money, sells the agency a ride, and keeps the design. NASA pays a fixed price and accepts that some missions fail, because a fixed-price contract is cheaper than a cost-plus one and the failure rate is the price of learning. That model produced the first American soft landings in half a century and also produced a run of crashes, a lander that tipped over and a mission that arrived on its side. CLPS 2 keeps the fixed-price structure and changes the pool. Eight companies competing for orders means NASA can steer work toward whichever supplier is performing, and it means a company that misses a schedule does not get the next mission by default. It also means the companies carry the cost of bidding, and the smaller ones will feel that cost more than Lockheed Martin will. A pool is a way of buying optionality, and optionality is paid for by the bidders.
The lunar south pole is where the money is going. Blue Origin received a separate award worth $94 million as a base contract to deliver a rover called PROMISE to the polar region on its Mark 1 lander, with a launch targeted for the end of 2028 and a performance option worth another $140 million. PROMISE carries a nuclear power source, which is the detail that matters. A solar-powered rover at the south pole spends half of every lunar day in darkness and the other half with the panels near the horizon. A radioisotope power system does not care.
The scientific case for the pole is water. Permanently shadowed craters there hold ice that has been in place for billions of years, and ice can be split into hydrogen and oxygen, which is propellant. Every architecture for a sustained lunar presence assumes that propellant will eventually be made on the surface rather than shipped from Earth at a cost measured in tens of thousands of dollars per kilogram, and every version of that plan begins with finding out how much ice exists, where it sits and what form it takes.
The $10 billion ceiling is not $10 billion of funding. It is the maximum the program can spend if every option is exercised and Congress appropriates the money, and NASA’s science budget is under pressure from cost overruns elsewhere. The number that matters in any given year is the sum of task orders actually placed, and those will be small at first. A program with a ten-year ceiling and a flat budget spends its money slowly, and the gap between the ceiling and the spending is where programs usually go wrong.
The industrial consequence is consolidation pressure. Fixed-price delivery rewards companies with a balance sheet that can absorb a failure and a factory that can build a second lander after the first one tips over. Firefly and Intuitive Machines have both flown and both have had bad outcomes. Lockheed and Northrop have the balance sheets and less appetite for the risk. The pool format lets NASA keep all of them engaged while it learns which ones can hit a date, which is a more efficient way to run a competition than picking a winner before anyone has flown.
There is a budget politics layer that will decide more than the engineering does. A lunar delivery program is cheap relative to a rocket or a space station, and it is visible in a way that a research grant is not. That makes it attractive to lawmakers who want a concrete achievement and vulnerable to lawmakers who want the money spent in their districts instead. The ten-year ceiling is a promise that has to be renewed in five appropriations cycles, and the promise is only as good as the last one. What to watch is the first round of task orders. A pool of eight is a statement of intent, and the list will narrow in practice as NASA places missions with the suppliers that hit their dates. The second thing to watch is PROMISE. A nuclear-powered rover on the south pole by the end of 2028 would be the first surface power system of its kind since Apollo, and it would settle a question that has been open since the ice was confirmed: whether the polar deposits are accessible or buried under a layer of regolith that no wheeled vehicle can dig through.
NASA has spent twenty years explaining that it cannot afford to return to the moon the way it went the first time. The commercial delivery model is the answer it settled on: buy the ride, insure against failure by buying several, and let the market work out the lander. Friday’s announcement extends that bet for another decade and raises the amount at risk, and the agency has now committed to a procurement strategy that assumes its suppliers will get better at landing because the alternative is too expensive to contemplate.
