Money Pits in Rocket Production

Purpose of Article I

In this article, the costs of rockets are broken down into its parts, so we can point out what specifically makes rockets more expensive. We use the Falcon-9 rocket as a proof-of-concept, although the conclusions we derive in this article are applicable to any rocket. As it turns out, approximately 13.33% of a rocket’s cost originates from the large number of parts that the Turbo-pump carries.

There are so many different models of rockets, intuitively making each one’s cost breakdown different than the others. For this reason, it is extremely important to narrow down on a specific rocket when breaking down its costs, to be as accurate as possible.

For the purposes of this article, I have decided that the specific rocket that we will be looking into is the Falcon-9. I chose the Falcon-9 since it is SpaceX’s most commonly used rocket. As the Space Industry is picking up speed, it is very likely that SpaceX will hit very steep growth curves. Selecting a rocket that is a big part of the success of one of the largest space companies in the world makes all of the following calculations more relevant for use in the modern day.

It is generally the case that money pits are not so obvious in first glance. Therefore, the approach I will be taking while hunting down a money pit will optimize for a balance of magnitude (how costly is the money pit?) but also specificity (the more specific, the more fixable — it doesn’t help to say that the most expensive part of the rocket is the rocket itself)

In order to do this, I will be breaking down costs in “layers”. Basically, I will break down the costs of a rocket based on its main parts, select the most costly of those parts, and further break down the selected part’s cost by looking at even smaller components that make it up. It’s a recursive process.

Layer I — Large Rocket Compartments

The main compartments in a rocket are considered to be the First Stage, Second Stage, Fairings, and Launch Associated Logistics. Of course, we will talk about costs in scope of the Falcon-9 to be as accurate as possible. It is estimated that the Falcon-9 costs ~$60 Million to produce and launch.

• The First Stage — 60% of the cost — $37 Million
• The Second Stage — 20% of the cost — $12 Million
• Fairings — 10% of the cost — $6 Million
• Launch Associated Logistics — 10% of the cost — $6 Million

Going by the logic that we have established earlier, we would have to look into the First Stage in a more detailed manner, since it makes up a significant percentage of the cost of a rocket.

Layer II — Main Parts of the First Stage

The first stage has 3 main components. These components are:

• Engines (x9)
• Fuel, Oxidizer, Body Tanks Material Cost
• Engineering Costs

According to the ULA (United Launch Association), each Merlin engine costs around $2.17 Million. This means that consequently, around $20 Million is spent on the engines in the first stage, making engines cover up around 54% of the rocket’s first stage’s costs.

As for the cost of the fuel tanks and the rocket body, the numbers aren’t as clear. However, it is possible to estimate the cost of the materials of the Falcon-9 first stage. Without propellants present, the Falcon-9 weights 25.6 metric tons. Subtracting the weight of the engines, the first stage should place around 21 metric tons. The aluminum-lithium alloy used in order to build the first stage costs around $2,000 per metric ton, making the cost of raw materials for the first stage is at most $50,000.

Obviously, the first stage is much more than simply the cost of its raw materials. There is not enough information out there to accurately predict how the cost breaks down for the rest of the assembly of the first stage, so it is best if we leave it there. Cutting the raw materials cost and transportation some slack, and giving them 6% share in the cost of the first stage we would be assuming that the assembly, engineering and put-together costs for the first stage cover up 40% of the first stage’s costs.

• Engines (x9) 54% of first-stage cost = $19.98 M
• Fuel, Oxidizer, Body Tanks material cost 6% of first-stage cost = $2.2 M
• Engineering Costs 40% of the first-stage cost = $14.8 M

Once again, optimizing for impact, we would reduce our scope down to reducing engine costs specifically.

Layer III (Part 1) — Engine’s Costs

Like mentioned, each engine is estimated to cost ~2.17 million dollars. A major component of the Merlin engine is the LOX/RP-1 Turbo-pump. A report released by Barber Nichols estimates the cost of each Turbo-pump used in the Merlin engines to be about 1.2 million dollars. This would mean that 55% of a rocket’s engine costs originate from a Turbo-pump. But what are the main costs when it comes to creating a Turbo-pump?

NASA has published a case study on the costs of contributing factors of engine production. They have analyzed the general costs for each specific step in engine development, and how much they contribute overall (as there are many cost drivers, most of which cover <1% of costs, I will be listing major cost drivers, >1% to keep the list to-the-point):

• Design Operations Analysis = 1.23% of engine costs
• Turbo-pump Development Tests =1.95% of engine costs
• Production Fabrication Operations = 42.40% of engine costs
• Production Test Operations = 39.50% of engine costs
• Field Maintenance & Repair Operations = 1.24% of engine costs

Just looking at Fabrication Operations (42%), it seems to be roughly evenly split in between: Procurement Processing Planning, Tooling Fabrication, Raw-stock Procurement, Casting or Forging, Machining, Welding, Subassembly & Assembly, Final Assembly (Engine), Inspection, Storage and Shipping.

As for Test Operations (39.50%), the reason they are so expensive isn’t because there are many different steps involved in the execution of this step, it is because the test + improve operations are extremely iterative. In contrast to Fabrication Operations, Test Operations aren’t split 10-way to increase costs, rather, it is constantly repeating itself. Engineers constantly test their engines, learn failures and update their design until the engine is ready to go.

It can be observed that both of these Operations are affected by the number of parts that exist within the Turbo-pump.

For fabrication costs, the amount of money spent should almost proportionally increase with the number of parts in the Turbo-pump, since each part adds another element to every one of those categories. In other words, the more parts you have, the more planning, fabricating, raw-stock procuring … welding, inspection, storage etc. you have to do. This causes fabrication costs to be almost directly dependent on the number of parts needed.

This also goes for Test Operations, the more parts, the more scientists and engineers have to test individual parts. They also are more vulnerable to inconsiderate failure. Basically, the number of parts in the Turbo-pump effects the number of iterations needed in the test operations step. The less the better.

So from this observation, we can conclude that 42% (Fabrication Costs) and (39.50% Test Operation Costs) are largely influenced by the number of parts (or the physical complexity) of the engine. Pretty much, ~80% of the cost of a Turbo-pump is heavily dependent on the existence of more parts.

Essentially, the question boils down to, how to we decrease the amount of parts (simplify) a Turbo-pump has, to drive down the overall cost of an engine, therefore a whole rocket.

Layer III (Part 2) — A Real Life Case Study

Numbers are a great way to predict the outcomes of ideas, but it is almost impossible to know how products will shape up to be in real-life implementation, due to the amount of detail in commercial industries. For this reason, it is important to support ideas via examples in real life.

How would we prove that simplicity leads to cost reduction in rocket engines?

For seeing how the reduction of complexity has lead to success in the Space Industry, Rocket Lab is a prime example. The business philosophy of Rocket Lab is to keep rocketry extremely simple by 3D-printing their rockets, and mainly using battery-powered engines to propel their relatively small rockets.

The Electron Rocket, Rocket Lab’s “main” product, is propelled by 10 Rutherford engines. These engines are battery-powered engines, something that can only be applicable in small-sat launching rockets. The fact that these engines are battery-powered gives them an absurd amount of part reduction in the Turbo-pump, gifting the company a huge dip in engine production prices.

Each Electron Rocket costs around $7.5 Million. It would be a ceiling-high estimate to place the costs of the engines at ~50% of the total cost, but so be it. This would mean that 10 Rutherford engines cost $3.75 Million, placing each Rutherford engine at $375,000 compared to SpaceX’s Merlin well above that price at $2.17 Million. Remember, this is an extremely unlikely high placed estimate. It is likely that the Rutherford engine is well below $375,000. This isn’t because the engines are much smaller. Yes, they are much less than half the size of a Merlin engine, but it is important to highlight that both engines use cheap alloys, so metal volume takes minuscule effect.

It can safely be said that this huge reduction in cost was because Rocket Lab simplified their engines, lowering the amount of parts. The reduction of the number of parts lead to the drop of testing costs and fabrication costs (as explained in Layer III Part 1) resulting in an extremely cheap engine.

It is agreed upon by both NASA’s cost breakdown, and Rocket Lab’s business model that simplification can lead to rather surprising reductions in total rocket costs.

Conclusion

A considerable amount of the price of a rocket comes from engines, specifically the Turbo-pump. By decreasing the number of parts which are within the Turbo-pump, it is possible to make it cheaper, transitively reducing the engine’s, and therefore the rocket’s price as well.

Turbo-pumps make up 16% of the overall cost of the rocket. By the conclusion we drew, we could say that 80% of this amount is dependent on the number of parts. This would mean that 13.33% of the cost of the rocket is caused by a large number of parts in the Turbo-pump. So if we reduce the amount of parts in the Turbo-pump by half, we could make rockets 6.65% cheaper. That would be $4 Million saved per Falcon-9.