Deuterium Production Guide

Strange Matter,
Deuteron Fuel Rod, and the late-game lines that depend on them. The fix is not to treat Deuterium as something you can casually burn or make later. Treat it as a dedicated supply chain, decide what it must feed first, and then build a production line that keeps Hydrogen moving without clogging the rest of your factory. If you do that, Deuterium stops being a panic resource and becomes a steady industrial input.
Decide what Deuterium supply has to feed first
Start by assigning Deuterium to the jobs that actually matter. Its main uses are Deuteron Fuel Rod,
Strange Matter, and Plasma Capsule production, and that order matters in practice because Strange Matter becomes the real demand spike once you scale up. Strange Matter needs 10 
Small Carrier Rocket production also depends on Deuteron Fuel Rod, so if you are pushing toward launches, you are also pushing Deuterium demand higher.
Do not plan around burning Deuterium for power. You can use it in Thermal Power Plant and in Icarus’s fuel chamber, but it performs the same as Hydrogen there, so there is no meaningful payoff for spending a more valuable industrial material as fuel. Keep your Deuterium reserved for advanced production first, and only burn it if you are in an emergency and have no better option.
Start with the simplest source you can actually sustain
Your first real question is not “what is the best Deuterium source?” but “what can I keep running continuously?” 

Build your main plan around Hydrogen reprocessing. That gives you a source you can scale with your own factory demand instead of waiting on orbital output to catch up. If you already have gas giant infrastructure, treat it as a supplement that smooths out your supply curve, not as the backbone of the whole chain.
Here is the practical comparison you should keep in mind:
| Method | Relative energy use | Throughput | Best use |
|---|---|---|---|
Fractionator |
720 kW | 0.3 ![]() |
Efficient core production |
Particle Collider |
12000 kW | 2 ![]() |
High-throughput fallback |
Fractionator 4-stack |
3960 kW | 1.2 ![]() |
Balanced upgrade path |
Build the Fractionator loop that keeps Hydrogen flowing
If you want a dependable Deuterium line, start with Fractionators. That is the standard efficient route, and it works best when you keep things simple: make a loop, feed Hydrogen into it, and use a buffer of Liquid Storage that is slowly refilled with Hydrogen. The point is not elegance; the point is never starving the Fractionators.
Set the loop up so Hydrogen recirculates continuously. That matters more than squeezing every tile of space, because a stalled loop kills output much faster than a slightly bigger footprint ever will. If your input dries up, 
Use Mk.III Conveyor Belts if you can, because Fractionator production is roughly 2.5 times more energy efficient than Particle Colliders when using them. That is the reason to favor Fractionators as your default choice: they cost much less power, and over time that power savings is significant. The trade-off is space and throughput, so expect a larger build and do not try to compress it too aggressively.
Solve the throughput bottleneck with a 4-stack when you need more output
When a normal Fractionator line cannot keep up, do not jump straight to Particle Colliders unless you truly need the highest single-line throughput. A better middle ground is to process Hydrogen in a 4-stack using the Automatic Piler. This increases throughput while keeping the 1:1 

The 4-stack setup uses about 1.375 times as much energy per Deuterium as a normal Fractionator, but it reaches throughput much closer to a Particle Collider line. That makes it the best upgrade path when your bottleneck is output, not power. Use it when your standard loop is too slow but you still want to stay in the efficient Fractionator model.
In practice, this means you should upgrade in stages: first build a normal Fractionator loop, then add a 4-stack where you need more flow, and only consider Particle Colliders if your build truly needs their speed and you are willing to pay for it in power. Do not overbuild Particle Colliders early just because they are faster; the energy cost is steep, and the extra output is often unnecessary if you manage your loop properly.
Keep Hydrogen balanced so Deuterium does not break your other chains
The most common mistake in late-game Deuterium planning is not production at all. It is Hydrogen balance. 
That is why Hydrogen from crude-oil-based production must be recycled through the system to ensure full efficiency. Do not dump it and do not treat it as disposable byproduct. Build your routing so Hydrogen is always reused, buffered, or redirected into the next step. The final stage quantities in crude-oil-based production are only the remaining surplus available for distribution, so you should design around that surplus instead of depending on it to solve everything.
If your Deuterium line starts competing with Casimir Crystal production, stop and fix Hydrogen routing before you add more Fractionators. More Deuterium hardware will not solve a bad balance problem. What will solve it is making sure every Hydrogen stream has a destination, with the Deuterium loop getting only the share it can actually consume without starving the rest of your factory.
The safest way to think about it is this: 
Deuteron Fuel Rod, and your other late-game chains all fed at once.
Fractionator
Particle Collider