The size of ISRU rigs

[Wjolcz]

Don't know about you but I always go for the smaller one simply because it's, well, smaller.

Can we at least have some sort of power generation thing (either a fuel cell or an RTG) included in the bigger one? Or a tank to store the fuels/ore in? The efficiency is also something I don't get. Would an IRL ISRU unit be more efficient if it was upscaled?

And please don't argue that 'it's a gaem!!!!'. I'm already kind of sick of all these comments using that as a valid point. Let's have a serious discussion instead.

Edited December 15, 2016 by Veeltch

[Darnok]

Larger batteries (4k is too little for large ISRU), larger generators and larger solar panels (gigantor like, but in different shape).

As for ISRU sizes I agree they should have almost same efficiency, but smaller one should have lower "transfer rate" what would add time needed to convert ore into fuel.

 

[Jarin]

There could easily be some kind of "fixed costs" involved in the process, making a larger unit more efficient. That's far from unheard of.

[Wjolcz]

43 minutes ago, Jarin said:

There could easily be some kind of "fixed costs" involved in the process, making a larger unit more efficient. That's far from unheard of.

What do you mean by that?

[Jarin]

11 minutes ago, Veeltch said:

What do you mean by that?

It's generally called the economies of scale. But put simply, the larger ISRU unit isn't just the small one tweakscaled up, it's going to have different internal mechanisms, and work on different ratios; doing things that wouldn't work - or would work differently - if miniaturized. Miniaturization is another concern in itself. We don't have shrink-rays, so sacrifices may well have to be made in creating a smaller device for the same task. Like the smaller unit doesn't have space for equipment to pull useful material out of the last dregs of ore residue, so it's throwing away potentially useful material that the larger one could process.

Edited December 15, 2016 by Jarin

[Wjolcz]

8 minutes ago, Jarin said:

It's generally called the economies of scale. But put simply, the larger ISRU unit isn't just the small one tweakscaled up, it's going to have different internal mechanisms, and work on different ratios; doing things that wouldn't work - or would work differently - if miniaturized. Miniaturization is another concern in itself. We don't have shrink-rays, so sacrifices may well have to be made in creating a smaller device for the same task. Like the smaller unit doesn't have space for equipment to pull useful material out of the last dregs of ore residue, so it's throwing away potentially useful material that the larger one could process.

Fair enough. I can get behind that. But the power demands are still pretty crazy especially around/on distant worlds like Eeloo and Jool.

[Jarin]

Just now, Veeltch said:

Fair enough. I can get behind that. But the power demands are still pretty crazy especially around/on distant worlds like Eeloo and Jool.

Yeah, I tend to make "Stock nuclear reactors" by taking a Mk2 Service Bay and cramming like 50 RTGs in it. 

[MaxL_1023]

The power demands could be fixed by adding a capacity slider, so you can change it from 0 - 100% load even with ore available. The EC demand could then be set as a constant (representing base operation) plus so much EC per converted ore. 

 

The one thing I would like is the ability to mount radially attached parts onto an ISRU - maybe have a toggleable framework surrounding the part. I always am annoyed when I need to stick radiators one part away instead of attaching them directly to the heat source.

[Sharpy]

20 hours ago, Veeltch said:

 Would an IRL ISRU unit be more efficient if it was upscaled?
 

Definitely so.

Some technologies can only be miniaturized so much. There is no pocket-sized nuclear reactor. There are RTG batteries the size of coins, but the smallest nuclear reactor barely fits on large bomber airplane. And yet that nuclear reactor can provide 1000x the power equivalent mass of RTG batteries could.

Let's imagine our ISRU is manufacturing RP-1 kerosene, the quality fuel, using carbon and hydrogen compounds found in soil.

All compounds are cracked using extreme temperatures (and a lot of energy); they assemble into random hydrocarbons with a huge amount of impurities. An advanced filtering system selects hydrocarbons that comprise RP-1, vents the rest.

Or - after cracking, the plasma passes through a set of advanced filters that remove most of impurities. Hydrogen and carbon are allowed to coalesce into hydrocarbons; the ones that comprise RP-1 are distilled, the rest is redirected for re-cracking. Vastly less raw material is needed, about the same amount of energy is used, but the advanced filtration system occupies several times the size and mass of the rest.

 

[Red Iron Crown]

22 hours ago, Veeltch said:

And please don't argue that 'it's a gaem!!!!'. I'm already kind of sick of all these comments using that as a valid point. Let's have a serious discussion instead.

But that is a serious answer, and a valid one whether you like it or not. Realistic ISRU equipment would be far smaller/lighter and slower; the ingame versions are made heavier to make it a real choice to include ISRU rather than a no-brainer, and faster. 

Same with the difference between the small and large ISRU processors, the larger one gives a mass penalty in exchange for quicker processing and higher efficency (the latter being critical for asteroid mining). 

[Magzimum]

General comment on scale up / scale down

There are multiple reasons that smaller processes are less efficient than larger ones.

1. Thermal efficiency

Firstly, efficiency takes a beating when you downscale, because of increased heat losses (volume goes down by 3rd power, outside area only by 2nd power, so smaller tanks/reactors have a relatively larger outside surface area). That's an inherent issue with downscaling processes that operate at temperatures other than ambient (although I am not sure there even is such a thing as ambient temperature in space).

2. Higher net costs lead to different design choices

The costs of smaller processes are higher than the larger processes in terms of costs per mass of product. A simple rule of thumb for chemical processes is the 0.6 power rule: "changing the size will change the capital cost by the 0.6 power of the capacity ratio" (text from wiki). I want to note that this 0.6 power rule actually applies for chemical factories, and does not really apply for processes as small as our ISRUs, but there are few relevant simplified rules for small scale stuff, so I thought I would mention it anyway. The general idea that downscaling makes stuff more expensive still applies.

Also, when downscaling, components become relatively more expensive. All your control systems (e.g. the electronics that measure temperature, and open/close valves) are not scaled at all, and will be just as costly on a larger rig as on the small scale one. Complex items where the machining costs of the manufacturing process are a major factor (and where the material costs are only a small factor) tend to not become much cheaper when scaled down. Think of the valves, pumps, for example. So, in terms of costs per ton of product, the smaller processes become much more expensive.

Because of the higher costs per ton of product at small scale processes, it may not always pay off to have complex separation/purification systems to use/recycle off-spec streams or waste streams. The choices are to either discard these (we can assume that the small Convert-o-tron does this) or accept the higher costs. Discarding waste streams instead of recycling/upgrading them will obviously affect the overall efficiency.

The increased cost leads to different design choices, and may lead to lower efficiency.

tl;dr, Yes, the lower efficiency of the small ISRU compared to the big one makes sense.

Comment on ISRU differences, fuel cell efficiency and ISRU efficiency

But...

Comparing the 125 and 250 Convert-o-trons, we have to be looking at two completely different processes that happen to have the same feedstock and products. The difference in efficiency is simply too large to originate from the same process. The difference in efficiency is about a factor 6, I think. I think that it's a bit much.

Having said that, I think it is weird that we can power the ISRU with a fuel cell. Especially the large ISRU can easily generate a lot more fuel from ore than its necessary fuelcells consume to power it. If the ore consists of the oxidized version of the fuel, then that's completely unreasonable. Using fuel cells to power the ISRU should never have a net positive fuel production. Instead, it should drain the tanks.

The only explanation how the fuel production can cost so little electricity is that the ore is only partially oxidized. Because it is possible to turn the ore into fuel and oxidizer using quite little energy, some of the chemical energy should already be present inside the ore! Therefore, we should be able to react that ore (with itself) to a lower energy state. We should be able to use ore directly as a low-energy fuel, without any ISRU.

I can live with this mismatch, because it's only a game... The design choices are interesting from a gaming perspective. But if you insist on realistic simulation, then I think you should never use fuel cells in combination with the ISRU, regardless whether you use the big or the small one.

[Wjolcz]

A lot of valid arguments here. So I guess the main topic here should be more about new power sources (like wind turbines) in KSP. But we've had a thread like this in the recent past. Sooo... Close the thread?

[Red Iron Crown]

Closed on request.