Kore Design
This page covers the key design decisions that you'll encounter with MKS.
- Provide an option to build 100% self-sufficient bases, but provide this at a significant cost, in both time and effort, commensurate with the reward.
- Add the capability to build parts and ships where you need them.
- Build a mechanism that forces part dependencies. For example, to get food the player must build an Agriculture Module, which requires Dirt from the Drill, and many converters require machinery generated by the Assembly Plant to operate.
- Include resource exploitation through a mechanism that forces the player to choose their colonization sites carefully, and situate them with easy access to resources.
- Keep things at a fairly high abstraction level. The sweet spot of MKS is in construction and logistics management over a fairly abstracted system of resources.
- Leverage existing add-ons where possible, and do this with minimal dependencies. The stock resource system is used to handle all generators and harvesters, the Community Resource Pack (CRP) handles shared resources, and the USI Catalog is used for Generators, Life-Support, and Kontainers.
Using the USI life support mod (USI-LS) with MKS can help provide an appropriate challenge where players will weigh the costs, risks, and rewards of supplying missions exclusively with carried life support Supplies, or in making the investment in a permanent colony. That being said, this is not just a single greenhouse part (although food production is part of the mod). Rather, it brings an entire colonization end-game to KSP in a fun and challenging way. Once fully developed, a planetary MKS colony can provide for a self sustaining support loop. Due to the way MKS is designed, orbital colonies will always have some degree of lossiness and need to be supported with periodic supplies from Kerbin or a nearby ground base.
The USI-LS mod also includes the Habitation construct, where Kerbals need room to move around or they get grumpy. They're also fond of Kerbin, and unless serious effort is made to create the kreature comforts of home, known as ColonySupplies, they'll eventually get homesick and need to go back to Kerbin.
For more information, see Life Support
There are four classes of kolonization parts. Each has its purpose in the building of your kolony.
- The smaller Ranger series is intended as the simplest way to deploy the basic modules. They sacrifice small size with the need to deploy large Material Kits. The parts also have a lower efficiency rate than the other two classes.
- The middle Duna series offers better performance based on standardized lander configurations. They can use the Ranger modules either as direct support or as enhancers.
- The Tundra series, has both 2.5 and 3.75 form factor that serves well as either a landed or orbital base.
- The Atlas series, the largest of the four, is for "end game" bases that are built on site as the parts are too large to realistically be flown in.
- The WOLF series provides a behind-the-scenes mechanism to simplify resource harvesting, conversion, and transport.
For the most part, inflatable/deployable parts (like the Ranger habitat and agricultrual modules, as well as the Tundra habitat ring) require a kerbal on EVA to deploy. There is also a cost in MaterialKits, which must be on the vessel at the time of deployment. This is because an inflatable module is more or less just an empty structure. The MaterialKits cost is meant to represent the components needed to flesh it out into a usable module instead of just a rigid balloon.
For more information, see the Parts page.
There are many trade-offs when designing a base. Here are several of the main ones.
USI-LS introduced the concept of Supplies being converted by Kerbals into Mulch just by the fact the Kerbals exist.
MKS introduces two other primary resource conversions. First, everything that does "work" requires Machinery, which slowly wears down into Recyclables. Similarly, there are replenishable nuclear power sources that work by converting EnrichedUranium into DepletedFuel and Xenon gas. Both of these conversions take place whenever the modules are used and are at a fixed rate.
Beginning bases can normally operate for quite some time on shipped-in resources. Each Kerbal requires 10.8 Supplies per 6-hour Kerbin-day. A single Mini-Pak will hold over 9 days worth of Supplies (100 / 10.8), and a single 1.25 flat tank hold enough for over 46 (500 / 10.8) days. However, as you move further out and shipping becomes less convenient, you'll need to start producing your consumables.
For more information, see Functions (Manufacturing).
When evaluating how you'll produce Supplies, there are three paths:
- Convert
Mulchstraight intoSupplieswith the Agroponics converter in an Agriculture module - Use
Water,Fertilizer, andSubstratewith the Cultivate(S) converter in an Agriculture module - Use
Water,Fertilizer, andDirtwith the Cultivate(D) converter in an Agriculture module
This choice is often made based on what local resources you can find. If you find a good location with a good supply of Water and either Substrate or Dirt, the Cultivate converters make more sense as they're more efficient. If not, you can use the Agroponics converter.
MKS provides a means of transporting resources around without having to have parts physically connected. Parts that are within 2000m of each other can be pulled from one part into another. In addition, ISM or Logistics module, resources can be moved into "Planetary Logistics" where they can be retrieved from any other ISM or Logistics module.
Resources are only pulled when necessary by a part that is performing an operation. The resources are pulled in bulk when required, instead of of trickling in.
See Functions (Logistics) for more information.
Note that a new game mechanic, WOLF, added in December 2020, provides a means for extracting, manufacturing, and transferring resources between biomes on a planet, into orbit, and between planets. More information is in the process of being added to the wiki.
Orbital stations can be created from any of the MKS parts, but the Tundra parts are especially designed for that task, as well as the Expandable Habitats. By design, orbital stations can never be 100% self sufficient as they have no source of resources. They also cannot participate in Planetary Logistics, at least not yet.
See Base Construction for more information.
Particular Kerbals provide increases in the efficiencies of particular operations. Also, their experience level impacts the part efficiencies. Initially, you may be able to get by with just Jeb and Bill, but eventually, you'll need to bring additional, less experienced scientists, engineers, and pilots. You can use the Tundra Training Akademy to raise their experience, or train them in the traditional way.
For more information, see Crew Skills Impact on Parts
RoverDude provided the following synopsis on the forum:
Supplies are not just food. Food is in fact a very small amount of that mass (less than 10%). Supplies are 'everything a Kerbal needs to keep alive and reasonably healthy'. 90% of that mass is water. For drinking, as well as sanitation and food prep. These are based on NASA numbers.
'Mulch' is just a nice way of saying waste water / grey water / candy bar wrappers / kerbal poo. The two are mass neutral - i.e. there is no mass loss in any of the processes.
Your simplest conversion is a basic reclamation system / biological filter / algae tank. Take your waste output, recycle it with a bit of fertilizer to grow new stuff. Easy peasy. (that is, 10kg of supplies become 10kg of mulch. 10kg of mulch plus 1kg of Fertilizer becomes 11t of supplies. So as long as you have fertilizer to feed back into the equation, you can keep the algae tanks filled indefinitely.
Cultivation (i.e. making a much larger growth of simple foods - algae, etc.). Takes some medium (Substrate or Dirt), Water, and Fertilizer to make more supplies than you could make purely by recycling mulch.
The next step up beyond that would be complex plants and simple edible animals (fish, etc.). NASA has toyed with this as well. This is covered under Organics in MKS. And for this you need everything for cultivation (just a lot more of it) plus a starting stock of Organics. There are more steps to prepare them (i.e. having a pig is nice, but it takes more work to turn it into wrapped packages of bacon and pork chops), but that's the basics.
Recyclables and Machinery are like 'life support for stuff'. Kerbals consume supplies and produce mulch. Mechanical processes and conversions consume machinery and produce recyclables. So same concept in parallel.
RoverDude provided the following clarification before releasing 0.50.9. It gives a great insight into how he thinks when playing with KSP, so it gives some insight into how he designed this mod. Note that the referenced ratios and amounts are for 0.50.9. They may not be exact for other releases.
There are four types of conversions in the chains. Working backwards:
Assembly is just taking some stuff and turning it into other stuff. Machinery and ColonySupplies do this in a lossless way (that is, toss in five tons of raw materials, you get five tons of stuff). Think of this as (literally) people doing simple construction with generic pre-fab parts, or packaging things up in the case of ColonySupplies. This is space efficient, and no real space to improve the process as it is already lossless.
Fabrication is where we take raw materials (sheets of metal and polymers, bins of chemicals, etc.) and make a finished good. These are also lossless (we assume any scrap is recycled), with the exception of SpecializedParts, where we lose some of our silicon. And even in that wors-case, the 50% loss is taken by the silicon (the rares/exotics are reclaimed).
Since this process as well is also pretty lossless except for SpecParts, but that's a small percentage of our used mass downstream in assembly anyway.
So going one step down... we get to refining. Since this is the step where we're taking rock and getting valuable stuff out of it, the lossiness is significant (to the point where you'd never want to ship around raw ore). The standard here is 25:1 for everything other than dirt, which is 50:1. Resource Lodes (coming in the next update) are the exception, in that they represent almost pure deposits (i.e. you are actively seeking and retrieving these vastly superior chunks vs. passively pulling stuff out of rocks).
Optimistically, this could drop to 10:1 or even 5:1, though below that would get a little sketchy. But you are never going to realistically live in a world where your refinery is the same size as your fabricator with 100% efficiency and throughput on both ends as that would just not make sense.
In real world terms, this number is reasonable. While you can get really good yields out of high quality iron ore (say, a 3:1 ratio), copper ore would be more like a 100:1 ratio. So the refinery to make that iron ingot will be significantly larger than the machine shop that shapes it into rivets and panels if you want said machine shop to operate non-stop at 100% capacity.
In gameplay terms, there are a couple of goals. First, to discourage the whole 'pokemon' base approach where you just include one of each module (and expect it to work with 100% efficiency). Second, to give a valid reason why you would want to shuffle around shipments of refined materials vs ores.
That being said, we've always had the issue of folks picking lousy spots to mine on, then getting upset when they have to drill spam. Active harvesting sorts some of this, but prospecting remains the best approach. But even with that, I can see a yield change from 25:1 to 5:1 being a reasonable tweak.
Let's run through some examples (and science!).
First... the current system.
A refinery, fully tuned for mineral processing and with a level 5 engineer, would take in 0.114375 minerals per second.
On my save, I found a mineral deposit of just over 9% (not enourmous, but pretty respectable). Using a drill completely tuned for mineral extraction and with a level 5 engineer on board, it was able to pull enough minerals to oversaturate the refinery with ease (0.180735 per second)
So let's extrapolate this out.
Assume I have a smallish industrial base with reasonable output levels. I have a level 5 engineer, two refineries and two drills. Based on the numbers above, I can conservatively expect to kick out 0.2 tons of harvested materials per second from the drills, which translates to 0.008 of refined materials per second.
At that low of a volume, I can get away with a single workshop (it can handle up to 0.028 per second).
That lets me crank out a ton of material kits every 34 hours.
Now, let us scale things up a bit (and optimize).
Unless we're on lousy deposits, I can feed a single refinery with a single drill. Worst case I need two. So let's go with five refineries total to match the 2/1/2 ratio of inputs for MaterialKits.
That gives me an effective rate of 0.571875 per second in raw materials, or 0.022875 in refined materials (and material kits!) or, 1 ton of material kits in 12 hours. And still in the range of using the workshop.
So if we extrapolate this out and assume that drills are not the bottleneck (i.e. their function is more about base selection and actually prospecting...) and we assume dedicated refinery/drill pairs per resource, then I should be able to kick out about 15 tons per kerbal month. For perspective, that's the earth equivalent of extracting and refining all of the materials to build a Toyota Camry in a week.
Probably the most I'd go is a 500% boost (i.e. move yield from 4% to 20%). If we did that, then we go from 15 tons per Kerbal Month to 75 tons per month (or to extrapolate our real-world example, we're mining and kicking out Camrys almost every day...)
The only downside is that I will need to deploy more workshops/fabricators but I expect folks can live with that 
TL;DR: Balance reasons and tweaks above. Refineries get a boost. Now onto other stuff!