[dead] CELSS Greenhouse (TAC Life Support Add-On, Version 0.1, beta)

[Cerebrate]

For various and sundry reasons, this project should now be considered officially dead. Anyone who wants to take the pieces and do something with them later is welcome to 'em.

(Well, technically it's not a plugin yet, as version 0.1 doesn't include the specific functionality that requires it to be one; for the moment it's just piggybacking off of the TAC Life Support plugin. But since version 0.2 and to follow will include plug-in-y bits, I figured it's probably best to call it one from the start.)

First time modder, here, and indeed first time poster, announcing CELSS Greenhouse, a small plug-in part for TAC Life Support. (Some version of it required; tested against 0.8.)

Basically, it's a simple plug-in part designed to simulate a fully closed-ecology life-support system by recycling CarbonDioxide into Oxygen, WasteWater into Water, and Waste into Food with full efficiency, via a notional greenhouse full of all the parts of the cascade you'd need to do something like that. The meta-level intention is to provide a part that saves you all the ongoing resupply trips to long-term space stations and large, long-voyage spacecraft while still (a) keeping some of the need to balance your life-support system around, and ( advantaging the existing TAC Life Support recyclers, etc., on short-term and smaller spacecraft. A full greenhouse is, after all, rather massy.

Download, license, documentation, source, etc., all available here: https://github.com/cerebrate/celss-greenhouse

[NOTE: THE DOWNLOAD YOU WANT AS A USER IS THE TOP ONE ON THIS PAGE: https://github.com/cerebrate/celss-greenhouse/releases

The download button on the front page gets you the source, not the actual mod. Sorry, that's just the way GitHub works.]

And here's a picture of the greenhouses in action in Kerbin orbit, at my own under-construction space station Liberty (four, in the vertical stack at left):

As you can see from the resource window, they're running away happily doing there stuff, and the system is more or less in balance.

Feedback is, of course, welcomed - including, specifically, feedback related to the balancing parameters of the part as well as any bugs, etc.

-c

Edited October 9, 2014 by Cerebrate
project death

[kiwiak]

So, its something less complicated that biomass?

Just simple put waste, get food, waste water -> water, co2 > o2, without taking care of seeds, biomass, germination etc?

I was waiting for such system because long term trips when i hd to put tens of tons worth of supplies were irritating (well yeah for lreal life longer trips we probably will in fact need tons of supplies unless we will have some magical food 3d printers )

One suggestion, include few of these along techtree, with various efficiency, maybe size variation too.

Edited May 11, 2014 by kiwiak

[Ratzap]

The balance question will probably boil down to when is it actually more efficient to use a greenhouse and how many Kerbals can one keep supplied. Variables like light levels (plants will have a hard time out by Jool) and darkness times (while orbiting a body) will make it less useful. Rather than make them overly complex, you could differentiate between tech levels with the efficiency. 100% efficient is rather 'magic tech' land as such systems will always produce some waste which is vented outboard. Meaning you would still need resupply visits, just less of them. Venting is beneficial btw, rather than retain waste which you can't process, you lose mass - everyone loves being lighter

As a yard stick you can look up human LSS efforts. A quick google gets me a wiki on space physiology which quotes usage per human per day at roughly 9kg (including cleaning water, you can no doubt find the page and see the breakdown yourself). Any mission where a greenhouse weighs more than 9kg x Kerbals x length of mission in days is less efficient. So lets plug numbers in for a 2 man Duna return visit with a reasonably close window and call the round trip 500 days - 9 tonnes of supplies. Now the big question is, how much mass does the equivalent journey in TACLS need - that will tell you how light your part *must* be before anyone who actually thinks when designing will take it along.

I was looking these things up lately as I was considering having a go at extending TACLS myself but I'm a terrible modeller, hence why this stuff is fresh in my mind. Main thing is, don't make the process too complex like seeds, harvest, process etc - keep it gamey enough to use and not a full sim.

[Cerebrate]

To answer a few points and give some idea of my roadmap, at least pre-feedback:

I was indeed going for extremely simple, assuming that most of the finicky plant, harvest, process details are handled within the greenhouse - so from an external perspective, it is just a simple CarbonDioxide, WasteWater, Waste (and ElectricCharge) goes in, Oxygen, Water, Food comes out module. (I'm also making the simplifying assumption that it cycles all the time, pretty much, and that the waste going in and resources coming out represent amounts over-and-above it just keeping itself ticking over.) They're also tied together such that you need all three to make it work: without CO2, fr'example, all the nitrogenous Waste in the world isn't going to help you make plants grow.

I have some notions for complications to the resource cycle I might add in the future, but they're all by way of optional features: on the output side, being able to get some sort of Biomass resource instead of Food which you could then reprocess into other resources to make use of spare capacity; or on the input side, being able to use in situ resources, processed or not, in place of the usual inputs to generate more life support resources. But I don't see that as changing the simplicity of the basic greenhouse, and it's a ways down the line anyway.

Size variations are definitely on the roadmap, coming soon. Efficiency variations I'm not so sure about (the existing recyclers in TAC Life Support for CarbonDioxide and WasteWater are already 90% efficient, so there's not all that much room to play with, there), but I'll have a think about possible variants for further down the tree increasing their mass-efficiency somewhat.

On the light-level question: light levels are pretty much the next thing on the list to handle. At the moment in 0.1-beta, it shamelessly ignores the issue, but that's never been the plan, belike.

What I'm thinking of in this respect is that the greenhouse comes with both windows and internal illumination. The first part of the lighting plan, thus, is to implement a change in the ElectricCharge used depending on the access to sunlight the greenhouse has (and whether or not the shutters are open), between a baseline level to reflect the cost of running the pumps, etc., that it uses all the time, and a higher level reflecting how hard the lights have to work to make up the 'light debt'. Putting a scaling factor on there for distance from Kerbol that makes the greenhouse more expensive to run in ElectricCharge terms, therefore, as available light intensities drop also seems rather appropriate.

Thoughts?

(And I'll freely admit that the nominal 100% efficiency - or at least 99-point-lost-in-the-statistical-noise-percent efficiency - is very much "magic tech". That's for metagame reasons - at least for me, working out how to balance life support systems for missions == fun, but running a bunch of repetitive resupply flights to installations pretty much == not fun. So I'd like to keep that around at least at the highest tech levels, although I'm certainly open to variety at various points.)

- c

[ArcFurnace]

Ah, yes. I've wanted to run the numbers for something like this, but never actually did because I didn't have a model.

A handy (if rather technical) document for realistic values is NASA's Advanced Life Support Baseline Values and Assumptions Document. Math time!

Looking through the report, it mentions an earlier study showing that if approximately 50% or more of the required food (by dry mass) is grown on-site, the plants used will be capable of providing all of your air and water regeneration. Since we're going for 100% of food grown on-site, that means we just need to figure out how much growing area we need to support a Kerbal with food, and that should be more than enough to also provide water and oxygen.

Some relevant values I've pulled out of it are the system mass and volume per growing area, which come out to be 101.5 kg/m² and 1.03 m³/m². These round nicely to 100 kg (0.1 t, in the typical Kerbal mass unit of metric tons) and 1 m³Ã¢â‚¬â€¹ per square meter of growing area. Note: these values do include mass and volume for artificial lighting, in addition to the structure and the crops themselves, but do not include extra volume for crew members to actually walk around in the greenhouse and collect the crops. An increase of 25%-50% of the growing area is suggested for crew access, although the extra volume won't mass as much due to not including all the plant-growing equipment. Perhaps 0.125 t and 1.5 m³ per square meter of growing area for a properly accessible greenhouse.

Looking at the definitions for the TAC-LS resources, one Kerbal uses 0.0003166535 t of food per day, or 0.317 kg (317 g)/day. This looks like it's meant to represent dehydrated food for minimum shipping weight (the NASA document's values for the dry weight of one day of food for a human on IVA is 0.67 kg; presumably Kerbals use less food because they are smaller). The NASA document has a table of dry weight productivity (grams/day per m²Ã¢â‚¬â€¹) for various crops. Using those values, for a diet of 30% beans, 30% rice, 30% salad (tomatoes/carrots/cabbage/green onions/lettuce/radishes/spinach), and 10% strawberries (percentages by growing area) we get an overall yield of 8.78 grams of dry weight per day per square meter. That puts us at just over 36 square meters to support one Kerbal; using the above figures for mass and volume, that's 4.5 tons and 54 cubic meters. If we size it to fit with 2.5m diameter parts, the cylinder will be 11 meters long. How big is that compared to stock parts? Orange-tank sized*? Quite bulky for one Kerbal, although that helps push you towards packaged goods for any non-permanent installation, which may be desirable. Higher-tech versions with advanced automation and better packing could increase the density (you don't need room for Kerbals to walk around picking the vegetables if you have little robots to do it), say up to a maximum of 6x less volume than that initial module. That would leave the module at 4.5 tons / 9 cubic meters for an overall density of 0.5 t/m³, which seems about right for a maximum (plants and substrate are mostly water and have a density around ~1, but you need some airspace as well, so that pushes the density lower).

Power requirements are listed as 2.6 kW/m², mostly for the grow lights and a little for ventilation, which with our figure of 36 square meters means we need 93.6 kW per Kerbal. I often go by KSP Interstellar's definition of 1 Electric Charge/s = 1 kW; rounding leaves us at 90 EC/s per Kerbal. That would leave the module very power-hungry compared to most stock parts, and even the stock TAC-LS converters (they use 1000 EC per conversion, but only do 8 conversions per day, or 8.4 on the large converters). If you're actually using KSPI, it's not a problem, as the reactors in that mod produce more than enough power, but we might need to tone it down a little for people who aren't. Using filtered sunlight instead of grow lights would also substantially reduce the power requirements. It appears to be that 2.2 kW out of the 2.6 kW are for the lights, so the remnant is 0.4 kW per square meter, or 14.4 kW per Kerbal, so long as sufficiently intense sunlight is available.

How does it sound so far?

*A check shows that the height of a Jumbo-64 tank is just about 3 times the diameter of other 2.5m parts, so it's about 7.5m tall. This is supported by the part.cfg, which has the stack nodes at ±3.75. Therefore this 11m long module is actually just under 50% longer than an orange tank, or just about the size of three Rockomax X200-32 tanks. Using the 6x density version would let you have a greenhouse the size of a X200-32 tank, which looks to be about the size of your current model, massing 9 tons and capable of supporting two Kerbals indefinitely. Natural sunlight plus two Gigantor solar panels would supply its energy needs in Kerbin orbit or closer to the sun. Trips further out are going to need a non-solar power source, preferably something heftier than the stock RTGs.

For comparison, using the 2.5m TAC-LS parts, 1x Carbon Extractor/1x Water Purifier/1x Life Support Container/3x Food Containers provides 6400 24-hour days of life support at a bit less volume than a X200-32 tank and a mass of 4.7 tons. This is enough to support two Kerbals for 8.8 Earth years, at about the same volume, half the mass and with far less power usage.

Edited May 11, 2014 by ArcFurnace

[Helix935]

dear lord how long did you guys type?

[ArcFurnace]

dear lord how long did you guys type?

The typing doesn't take nearly as long as poking around to find all the numbers, heh heh heh.

[Ratzap]

Good to see people thinking about it. As you say, no CO2 source would mean less output so 1 kerbal in a ship with a 3 kerbal sized greenhouse would still only get 1 kerbals worth out (and a lot of wasted space/mass obviously). I think this would need some sort of dialogue with the TACLS author to standardise on mass/denisty of parts so that your greenhouse can have a purpose without being too wildly divergent from real-ish. TACLS parts are probably too small and efficient right now.

For the sunlight code, you can look at the RSS stuff, they have solar panels working with proper root falloff. Interstellar does something as well.

[Hattivat]

snip

Awesome post, thanks a bunch, I've been investigating this topic for the last two days, but the results of my amateurish searches pale in comparison to the amount of data you've provided in this post. The mass estimates are particularly invaluable, as I couldn't find them anywhere.

Two questions: If I understand it correctly that NASA document assumes a hydroponic system is used (unless they consider aeroponics a subset of hydroponics and use the latter term for simplicity). From what I managed to get out of their website and other sources, it seemed that an aeroponic system would be preferable as it is more weight-efficient. Is there any particular reason why they chose hydroponics?

And another, wouldn't it make sense to make the whole structure inflatable? That should make the whole thing much less bulky on launch. NASA already did some research on this: http://www.nasa.gov/offices/ipp/centers/kennedy/success_stories/Inflatable_Aeroponic_System_BBlinds.html

[ArcFurnace]

Well, I was just running off the document there. They are definitely using a hydroponic design as their base, not aeroponics. Perhaps they simply didn't have as much information on aeroponics at the time it was written? Just from looking at the Wikipedia article it seems that aeroponics might indeed be better, and has specifically been researched for space use (including the inflatable system you mention).

As for inflatable structures, they definitely cover inflatable habitation modules in the report, but don't seem to discuss it in the biomass section, which is what we're dealing with here. I'm not particularly sure why.

I have a slight aversion to using inflatable modules, but that's mostly difficulty-based. Specifically, we'd need animation code and modeling to cover inflating the modules once they're launched. If someone can do that, well, that's not a problem then, but I know I couldn't. I'm still imagining these as stack-mounted, in which case you might not want the animation to change the length of the cylinder, either (I suspect it might cause problems, although InfernalRobotics says it's workable if you're good enough at coding). I suppose you could have a 2.5m diameter initial module that inflates to a 3.75m (or larger) diameter cylinder without changing length, TransHab style. You'd have to be careful if you wanted it to land sideways on a planet, though; it would probably brush the ground if you built it to interface with other 2.5m parts on wheels due to the extra diameter. Landing it vertically would still work.

Hmm ... what would the reference design look like updated with these technologies? The page you linked says that aeroponics can grow crops in half the time of hydroponic systems (is that true? I dunno, but we're going to roll with it because I want a module that's not stupid-huge), so double the number of Kerbals the system can support. Inflating from 2.5m to 3.75m diameter increases our available volume by 2.25x, so that's 4.5x total (although mass also increases by, hmm, let's say 2x to cover the extra plants, with the inflatable structure being a little lighter per volume). Last, a bit more than a third of the mass of the hydroponic system is in the "root zone and nutrients", which would be mostly the mass of the hydroponic nutrient fluid, so cut the mass by 33%.

Basing it on the highest-density design I detailed above, we now have a 2.5m diameter / 3.75m height cylinder (Rockomax X200-32 tank dimensions) that inflates to a 3.75m diameter/3.75m height cylinder. It masses 12 tons, and can support 9 Kerbals indefinitely provided you have sufficient power. Nice! The power usage goes up just as fast, though ... if we scale it by number of Kerbals, at 90 kW per Kerbal we end up at 810 kW for the greenhouse unit. If you stay stock, and inside of Kerbin's orbit, that's 45 Gigantor solar panels, or worse, 1080 RTGs for further out than Kerbin. Yikes! Even with full sunlight instead of grow lights it would require 130 kW just for the ventilation (and that's even more important with aeroponics- got to keep that nutrient solution mist moving), which is still 8 Gigantors- doable, but still a fair few panels. We can get lower power consumption numbers if we only increase the power usage by 2.25x for the extra volume, although with the plants growing twice as fast I'm not sure if that violates thermodynamics or not. That would be 405 kW for grow lights or 65 kW for sunlight+maintenance. Still pretty much need some sort of modded electricity source to handle that much power- KSPI's reactors can do it easily, but I'm not sure how many other things can. I personally use KSPI, so I'd be totally okay with that, but others might not be so happy (it IS a pretty big mod, and the necessary rearrangement of the tech tree makes it incompatible with other tech tree adjustment mods). Plus, now you have a reactor that'll run out of fuel eventually, although if you include one of the extractor refineries most planets have deposits of uranium or thorium that'll keep you running. Of course, if you prioritize reasonable gameplay over strict or semi-strict realism, you could always just tweak the power use numbers around however you like.

A noninflatable version of the above would be a 2.5m dia x 3.75m high cylinder that masses 6 tons and could support 4 Kerbals indefinitely with sufficient power (360 kW for artificial lighting / 60 kW for natural lighting, both at max realism, less if you want to tweak the power values downwards). This version masses 1.5 tons/Kerbal supported, the inflatable version 1.33 tons/Kerbal.

Comparing again to packaged life support, 4.7 tons of that in 2.5m parts gets you 6400 days of life support to spread across all crewmembers. You can also add extra packs of 1 Life Support/3 Food Containers for another 6400 days each, which add 4 tons per pack. The crossover point where a greenhouse module is less mass than prepackaged supplies is between 1600 and 3200 days for 4-Kerbal voyages and the noninflatable module, or between 1600 and 2400 days for 8-Kerbal voyages and the inflatable module. The greenhouse modules always use vastly more power than prepackaged supplies with the current TAC-LS converters.

For reference, the longest Kerbin <-> elsewhere Hohmann transfer orbit is to Eeloo, at 396 Earth days each way, plus however long you have to wait for a transfer window. This is well within the range of prepackaged supplies if you're not planning on hanging around for a really long time. It's looking the like greenhouses mostly make sense for seriously long-duration voyages or (semi)permanent bases, which might be just fine given that said bases were what this was intended for in the first place.

What do you guys think?

Edited May 12, 2014 by ArcFurnace

[Hattivat]

Thank you for a well-thought-out reply.

I have a slight aversion to using inflatable modules, but that's mostly difficulty-based. Specifically, we'd need animation code and modeling to cover inflating the modules once they're launched. If someone can do that, well, that's not a problem then, but I know I couldn't. I'm still imagining these as stack-mounted, in which case you might not want the animation to change the length of the cylinder, either (I suspect it might cause problems, although InfernalRobotics says it's workable if you're good enough at coding). I suppose you could have a 2.5m diameter initial module that inflates to a 3.75m (or larger) diameter cylinder without changing length, TransHab style.

Here is an already made solution, assuming that Porkjet would be interested in cooperating on this or kind enough to let someone else use his model: http://forum.kerbalspaceprogram.com/threads/64442-0-23-5-Habitat-Pack-v0-4

I've already suggested doing something like that further in that thread.

Hmm ... what would the reference design look like updated with these technologies? The page you linked says that aeroponics can grow crops in half the time of hydroponic systems (is that true? I dunno, but we're going to roll with it because I want a module that's not stupid-huge)

I wouldn't know, as I'm no biologist, and plant-related limitations is the part that I don't really understand, all I can say is what sources claim. What I do understand is that from an engineering perspective the existing greenhouses for KSP (from Biomass and other mods) make no sense for a Jool mission (way too heavy, unnecessary windows, too much empty space inside them), which is what I want to use a greenhouse for.

Aeroponics is supposed to make plants grow faster, yes, but the grow time figures quoted in that NASA document are already good results compared to the numbers I've seen elsewhere, so I'm not sure if a 2x improvement could be achieved. What I think could be very plausibly achieved is packing at least slightly more plants into the same amount of space (sources say that since there is no soil or water between the roots, plants do not compete with each other for nutrients).

If I understand it correctly one significant limitation of aeroponics is that the nutrient solution needs to be of very high quality, which would probably mean that recycling kerbal waste directly into it would be extremely hard (if not outright impossible) to achieve. Personally I'm willing to roll with it and assume that kerbals achieved some breakthrough in waste treatment technology (my kerbals already have fusion reactors, duh), that's far preferable to assuming that my kerbonauts are willing to eat nothing but potatoes for two years (given their limited volume, the Biomass mod greenhouses can't really be assumed to provide much more than that) and that my kerbal engineers consider it a good idea to incorporate heavy windows into the design of greenhouses intended for use on Jool's orbit.

I personally use KSPI, so I'd be totally okay with that, but others might not be so happy (it IS a pretty big mod, and the necessary rearrangement of the tech tree makes it incompatible with other tech tree adjustment mods). Plus, now you have a reactor that'll run out of fuel eventually, although if you include one of the extractor refineries most planets have deposits of uranium or thorium that'll keep you running. Of course, if you prioritize reasonable gameplay over strict or semi-strict realism, you could always just tweak the power use numbers around however you like.

I'm using it too so power consumption numbers look fine to me. Furthermore, I think that what we are trying to achieve here actually SHOULD be impossible to achieve with stock parts, they represent mostly 1970s technology after all. There is however a solution if someone is unwilling to use nuclear power - Near Future Propulsion pack includes enormous blanket solar arrays, using a few of them should be enough to run the greenhouse at least in Kerbin orbit. The bigger problem would be packing enough batteries to power the thing during the "night" period of the orbit. Of course tweaking the power consumption down remains a possibility, although I would probably tweak it back up for my personal use.

It's looking the like greenhouses mostly make sense for seriously long-duration voyages or (semi)permanent bases, which might be just fine given that said bases were what this was intended for in the first place.

Personally, I'd be willing to use it mostly to role-play providing a little extra comfort to my kerbals to improve morale (my spaceship designs include at least double the required crew space, including a centrifuge, for the same reason), after all it's more pleasant to eat fresh veggies than food concentrates, and with the late-stage KSPI technology I can easily afford that. Furthermore, my spaceship designs are already single-stage-from-LKO-to-eeloo-orbit-and-back, so I guess they could be considered an indefinite-duration mission, as they are not intended to ever return to Kerbin surface (excepting the unplanned lithobraking scenario).

[CaptainKipard]

Ok seriously, too much text, but the greenhouse is a nice idea. I'll try it.

[Hattivat]

And to avoid being rude: Cerebrate, from the calculations done so far, it seems that the setup you have in that example pic (that's four greenhouse units, right?) should, assuming full automation (which your model's looks would suggest), realistically provide enough food to adequately support two kerbals if it employs hydroponics, and three to four kerbals if an aeroponic system is used. You can of course tweak these numbers up if you don't care about realism as much as I do.

[Ratzap]

And to avoid being rude: Cerebrate, from the calculations done so far, it seems that the setup you have in that example pic (that's four greenhouse units, right?) should, assuming full automation (which your model's looks would suggest), realistically provide enough food to adequately support two kerbals if it employs hydroponics, and three to four kerbals if an aeroponic system is used. You can of course tweak these numbers up if you don't care about realism as much as I do.

KSP is a game sim rather than a hardcore sim so I'd strongly hope that any parts are balanced toward being useful in relation to peer parts. Making things too real tends to mean no-one will use them - which would be a crying shame to waste the time and effort.

[Temeriki]

Yeah, what your not considering with greenhouses is gmo algae. Insert a bunch of genes to code essential amino acids and some extra protien, maybe some avocado fat, a few tweaks to speed up growth, make it thrive on poo. Bamn, fast growing foodstuff.

[Hattivat]

Yeah, what your not considering with greenhouses is gmo algae. Insert a bunch of genes to code essential amino acids and some extra protien, maybe some avocado fat, a few tweaks to speed up growth, make it thrive on poo. Bamn, fast growing foodstuff.

Certainly possible, but very heavy, as water is the heaviest part of any space agriculture system. An aeroponic lettuce only needs to have its roots sprinkled with nutrient-containing mist once every 20 minutes or so, whereas algae have to be constantly surrounded by it.

[ArcFurnace]

Mmm, green slime dinners. Spirulina already has all of the essential amino acids and is ~60% protein when dried (and ~85% digestible biomass total, with the non-protein stuff being carbohydrates, fats, and a little residual water). There's a problem with using it for human food, though, namely that it's high in nucleic acids. Those produce uric acid when metabolized, and if you get too much of that in your blood it crystallizes out into your tissues and causes kidney stones and gout. Very unpleasant. Interestingly, humans and other great apes are one of the very few animals on Earth that lack a functioning gene for the enzyme that breaks down uric acid (we actually still have the gene, but it doesn't work). Perhaps Kerbals still have theirs. Spirulina's an appropriate shade of green, too, which might encourage them to eat up.

Water is indeed heavy, but that depends on how much water you have to cart around. I found a paper on maximizing the output of bioreactors, although I suspect I shouldn't post it because it's copyrighted/supposed to be exclusive access (I got access through my university library)¹. They tested the effects of different light intensities, different light-path lengths (basically different depths of tank), different culture densities, and different levels of agitation (by having the CO2-enriched air bubbled through the reactor at an increased rate). The key to high production per mass is to have lots of algae in very little water, and the key for that turns out to be using very thin bioreactors. With that many suspended particles in the water, light can only penetrate a little way into the solution, so having a foot-deep pond or tank isn't very efficient in terms of production per volume of water. The unit used for light intensity was "photon flux density" or PFD, the number of photons striking a square meter per second. At the highest light intensity with optimal photosynthetic efficiency (2000 µmol photons per m² per s, about the intensity of summer sunlight at noon on Earth's surface), using a 14mm thick flat reactor (two flat plates of glass 14mm apart) continuously illuminated on both sides, at the optimal-for-that-light-level culture density of ~20 grams of biomass per liter of culture, they reached a volumetric dry biomass production rate of 21.6 grams per day per liter. With the reactor 14mm thick, you get 14 liters per square meter, so the areal dry biomass production rate is 302 grams per day per square meter (!). At that rate, using the 85% edible fraction from above, you'd need 1.25 square meters of reactor to provide food for one Kerbal, and the reactor itself is quite thin, although you'll need some extra space for the lights. Also, the reactors are apparently immersed in a secondary water tank to provide temperature control, which will add more volume and mass.

Power usage: The source paper used halogen lamps, but we can do better. LEDs can give up to 1.7 µmol photons/s per watt², and we need 2000 µmol/s per square meter (on both sides, so actually 4000 total). Comes out to 2.4 kW per square meter.

Mass estimate per square meter (justifications for estimates in parentheses, feel free to argue/suggest adjustments):

14 L of culture should weigh ~14kg (water and water-containing algae)

Reactor walls should weigh ~24kg (Plexiglas, density ~1.2 g/cc, make them 10mm thick, 2 walls)

Coolant+tank should weigh ~44kg (Duplicate reactor wall plus 10mm of cooling water, both per side)

Lights should weigh 75kg (based off multiples of this LED light fixture for a total of 2400 W power input)

Additional support equipment should weigh ??kg (pumps for cooling water and CO₂ enriched air, coolant heat exchanger, device to pull algae out of the culture solution, etc)

Say the support equipment mass is equal to the rest of the equipment as a guess.

314 kg per square meter times 1.25 square meters per Kerbal rounds to 0.4 t per Kerbal supported. Going to estimate volume at 1 cubic meter per square meter of growing area; the whole setup besides the lights is pretty thin, so that gives you ~half a meter on either side to add in the lights.

Caveat: Looking at the report, they apparently completely swapped out the nutrient solution every day to keep things fresh (high nutrients/low inhibitory wastes), and were running on a nice fancy growth media. Running it off biological waste products may or may not reduce efficiency. If we triple the growing area/mass/volume/power input to be on the safe side, that puts us at 1.2 t per Kerbal, a little lower than the aeroponic greenhouses, and using vastly less growing space (3.75 m² as opposed to 36 m²), and therefore much less volume as well (3.75 m³ versus 9 to 54 m³). Power input is also much lower at 7.2 kW per Kerbal.

TL;DR: Estimate of 1.2 t, 3.75 m³, and 7.2 kW power input per Kerbal supported by Spirulina. I hope you like green slime.

[1] If you have your own access, the citation is: Hu Qiang, Yair Zarmi & Amos Richmond (1998) "Combined effects of light intensity, light-path and culture density on output rate of Spirulina platensis (Cyanobacteria)", European Journal of Phycology, volume 33, issue 2, pages 165-171

Edited May 13, 2014 by ArcFurnace

[Cerebrate]

I'm still crunching numbers on how I'm going to revise the various parameters of 0.1 for 0.2 - and thanks muchly, everyone, for all the research and pointers - so not too much to say on those points yet. I did, however, want to mention one thing:

Here is an already made solution, assuming that Porkjet would be interested in cooperating on this or kind enough to let someone else use his model: http://forum.kerbalspaceprogram.com/threads/64442-0-23-5-Habitat-Pack-v0-4

I've already suggested doing something like that further in that thread.

I use Porkjet's Habitat Pack myself, so I'd be happy to see that, whether cooperatively, or using his model in my mod, or indeed if he wants to use my configs and upcoming code in his, which of course he's free to do.

I'll stipulate at this point, though, that the reason I'm not using my own greenhouse model for the current version is because my own personal talents when it comes to modeling and/or animation, as I've learnt in the course of various events, are somewhere between zero and negative. So, should anyone wish to make models they feel would better represent this greenhouse or anything else I come up with in the future, I will be pleased and grateful, but it's very unlikely you're going to see any come directly from me.

Speaking of which and while I'm on the subject, is there any sort of handy set of generic freely-usable models for misc parts around somewhere? Thinking about other possible life-support parts, I was wanting to knock up a simple machinery cluster that would let you convert Oxidizer (which is after all just LOX) to Oxygen and vice versa, thus allowing kerbonauts in desperate situations to breathe their fuel, or, going the other way, compress their life-support supplies back into something the engines can use - but the trouble is finding something to hang it off.

- c

[ArcFurnace]

Speaking of which and while I'm on the subject, is there any sort of handy set of generic freely-usable models for misc parts around somewhere? Thinking about other possible life-support parts, I was wanting to knock up a simple machinery cluster that would let you convert Oxidizer (which is after all just LOX) to Oxygen and vice versa, thus allowing kerbonauts in desperate situations to breathe their fuel, or, going the other way, compress their life-support supplies back into something the engines can use - but the trouble is finding something to hang it off.

I just added it as an extra module to the TAC Carbon Extractors. ModuleManager config here:

{
MODULE
{
name = TacGenericConverter
converterName = Transfer Oxidizer In


// Number of units to convert per day (24 hours)
conversionRate = 86400


// A comma separated list of resources to use as inputs.
// For each resource, list the resource name and the amount (which
// is multiplied by the conversionRate)
inputResources = Oxidizer, 1


// A comma separated list of resources to output. Same as above
// but also specify whether it should keep converting if the
// resource is full (generating excess that will be thrown away).
outputResources = Oxygen, 11.65, false
}


MODULE
{
name = TacGenericConverter
converterName = Transfer Oxidizer Out


// Number of units to convert per day (24 hours)
conversionRate = 86400


// A comma separated list of resources to use as inputs.
// For each resource, list the resource name and the amount (which
// is multiplied by the conversionRate)
inputResources = Oxygen, 11.65


// A comma separated list of resources to output. Same as above
// but also specify whether it should keep converting if the
// resource is full (generating excess that will be thrown away).
outputResources = Oxidizer, 1, false
}
}


@PART[TacCarbonExtractorLarge]
{
MODULE
{
name = TacGenericConverter
converterName = Transfer Oxidizer In


// Number of units to convert per day (24 hours)
conversionRate = 86400


// A comma separated list of resources to use as inputs.
// For each resource, list the resource name and the amount (which
// is multiplied by the conversionRate)
inputResources = Oxidizer, 1


// A comma separated list of resources to output. Same as above
// but also specify whether it should keep converting if the
// resource is full (generating excess that will be thrown away).
outputResources = Oxygen, 11.65, false
}


MODULE
{
name = TacGenericConverter
converterName = Transfer Oxidizer Out


// Number of units to convert per day (24 hours)
conversionRate = 86400


// A comma separated list of resources to use as inputs.
// For each resource, list the resource name and the amount (which
// is multiplied by the conversionRate)
inputResources = Oxygen, 11.65


// A comma separated list of resources to output. Same as above
// but also specify whether it should keep converting if the
// resource is full (generating excess that will be thrown away).
outputResources = Oxidizer, 1, false
}
}

@PART[TacCarbonExtractor]

Units are set up to conserve mass.

I also made a config to add a converter from KSPI "LqdWater" to TAC-LS "Water" to the TAC Water Purifiers (the two of these combining to allow KSPI LqdWater to be harvested, used as life-support Water, cracked to produce Oxidizer and converted to life-support Oxygen as a quick-and-dirty KSPI ISRU / TAC-LS integration). Still doesn't get you food, though; you need the greenhouse for that.

{
MODULE
{
name = TacGenericConverter
converterName = ISRU Water Filter


// Number of units to convert per day (24 hours)
conversionRate = 2160


// A comma separated list of resources to use as inputs.
// For each resource, list the resource name and the amount (which
// is multiplied by the conversionRate)
inputResources = LqdWater, 1.8, ElectricCharge, 20


// A comma separated list of resources to output. Same as above
// but also specify whether it should keep converting if the
// resource is full (generating excess that will be thrown away).
outputResources = Water, 1, false
}
}


@PART[TacWaterPurifierLarge]
{
MODULE
{
name = TacGenericConverter
converterName = ISRU Water Filter


// Number of units to convert per day (24 hours)
conversionRate = 2880


// A comma separated list of resources to use as inputs.
// For each resource, list the resource name and the amount (which
// is multiplied by the conversionRate)
inputResources = LqdWater, 1.8, ElectricCharge, 20


// A comma separated list of resources to output. Same as above
// but also specify whether it should keep converting if the
// resource is full (generating excess that will be thrown away).
outputResources = Water, 1, false
}
}

@PART[TacWaterPurifier]

Conversion speed and power usage based off a smallish reverse-osmosis water purifier. Units again adjusted to conserve mass.

(in case anyone wasn't certain, you can do whatever you like with these)

Edited May 13, 2014 by ArcFurnace

[Hattivat]

snip

Excellent as always, I'd give you tons of rep, but forum says I can't :<

It does indeed sound like a superior engineering solution, and a brilliantly simple one too. It never occurred to me that one could make a "flat" bioreactor. Not trying to argue here, just a personal perspective: This is probably the ultimate solution for long-term bases, but for my use (large spaceships) it kind of defeats the whole purpose. I mean, if what the system outputs is "green slime" then it is barely any better than pre-packaged food, and the question of whether or not it makes sense to equip your craft with it depends purely on weight efficiency (which for most missions will be a negative). Whereas actual greenhouses have a psychological value very much like a centrifuge - it might be heavier than the simplest possible solution, but it lets you feel good about providing that little bit of extra comfort to your crew.

Anyway, I'm starting to see a potential for tech tree extension here, something like:

.....hydroponics

.....| .................\

.....v ................. v

biorecycling aeroponics

.....| ........................ |

.....v ........................ v

advanced bior. volumetric efficiency

......\.......................... /

......v....................... v

advanced algaculture

(biorecycling would let you turn waste into hydroponics-grade nutrients (previously you'd need to transport them to orbit); advanced biorecycling does the same to aeroponics; probably not the best names for these techs, I hope ArcFurnace can propose better ones)

Is there something else we could add at the end of the tree? Perhaps liberal application of genetic engineering resulting in some sort of explosively growing fungus-chicken-spinach hybrid straight out of luddite propaganda?

EDIT: Actually now that I think about it, you could provide players with a selection of options with varying levels of palatability to mix and match to individual preferences. That way one could decide what level of "food comfort" he can afford to provide for his kerbonauts. For example:

1. algaculture cultivating spiruline - most effective but least palatable

2. aeroponics bay cultivating veggies - moderately effective, moderately tasty

3. centrifuge for raising plants that would not grow in zero-g conditions, or maybe raising some sort of animals (larvae? insects? chickens?) / aquaculture to cultivate crustaceans or fish - least effective but adds variety to meals

I've found tons of resources. Turns out the codephrase for good google results is "bioregenerative life support".

Watch this, they show a prototype inflatable greenhouse and a brilliant idea of using fiberoptics to transfer concentrated solar light into the greenhouse chamber:

It's designed for static use on the moon, so it is obviously not exactly how a greenhouse for a spacecraft would look like (especially one intended for use beyond Mars orbit, as it makes little sense to have any sort of structure transparency there), but would be a good inspiration for a game model.

In the research paper (from 2012) they claim to achieve an almost 50% reduction in weight (145 tonnes compared to 275) over the previous NASA reference design, and that is while still using a hydroponic system and providing crew access.

Spoilers below contain links to tonnes of relevant research papers I've found, which is either a good or a bad thing, depending on how you feel about reading such things. In any case, you have been warned.

http://www.marshome.org/files2/Hublitz2.pdf The rest of their documents look promising too: http://www.marshome.org/documents.php

the results from that lunar greenhouse prototype: https://www.academia.edu/2194711/Bio-Regenerative_Life_Support_System_Development_for_Lunar_Mars_Habitats (you need an account to download)

http://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20050182966.pdf

http://en.wikipedia.org/wiki/Food_systems_on_space_exploration_missions#Acceptability

http://humanresearchroadmap.nasa.gov/Evidence/reports/Food.pdf

http://www.lss-lab.bme.buaa.edu.cn/download/SCI/SCI/12%E8%83%A1%E6%81%A9%E6%9F%B1-Conceptual%20design%20of%20a%20bioregenerative%20life%20support%20system%20containing%20crops%20and%20silkworms_.pdf

http://forum.nasaspaceflight.com/index.php?topic=33699.0

https://www.researchgate.net/publication/7342241_Microbial_ecology_of_the_closed_artificial_ecosystem_MELiSSA_%28Micro-Ecological_Life_Support_System_Alternative%29_reinventing_and_compartmentalizing_the_Earth%27s_food_and_oxygen_regeneration_system_for_long-haul_space_exploration_missions (you need an account to access)

http://spacearchitect.org/pubs/AIAA-2013-3517.pdf

http://nia-cms.nianet.org/RASCAL/images/University-of-Colorado,-Boulder-Final-Paper-2013.aspx

https://www.researchgate.net/publication/225024211_Advanced_Greenhouse_Modules_for_use_within_Planetary_Habitats

http://globalecotechnics.com/wp-content/uploads/2011/08/Handbook-Envt-Engineering-Closed-system-chapter.pdf http://ajcn.nutrition.org/content/60/5/820S.full.pdf

http://www.ncspacegrant.org/bls/files/Crop%20Roadmap%20Document%20for%20NASA%20B_JC%20Edit.pdf

http://www.transorbital.net/Library/D101_S01.html

http://moonmars.com/sites/moonmars.com/files/items/documents/CEBLSS.pdf

http://science.ksc.nasa.gov/biomed/marsdome/

also of interest: http://www.leonarddavid.com/spin-control-greenhouses-headed-for-orbit/

this is hardly a research paper, but may be useful since unlike most other sources it discusses the downsides of aeroponics: http://www.gardenandgreenhouse.net/index.php/past-issues-mainmenu-18/37-2009-gg/march-2009/498-hydroponics-101

(probably of interest, but I can't get access:) http://www.ncbi.nlm.nih.gov/pubmed/16439102

http://www.sciencedirect.com/science/article/pii/S0273117707007065

http://arc.aiaa.org/doi/abs/10.2514/6.2010-6202

Edited May 14, 2014 by Hattivat

[Cerebrate]

So, I've been thinking about my ongoing development game plan.

Currently, I have in mind three greenhouse models, introduced at different points along the tech tree. The first of them, the "Basic Greenhouse", I propose to introduce from a new node, call it "'Ponics" (not to be specific as to which), attached after Large Electrics, about the time stations are starting to be practical. It's the lower-efficiency (i.e., not closed-cycle) greenhouse for beginners.

Later on, at "Advanced Science Tech" you get the actual "CELSS Greenhouse", the one in which (with associated mass and power tradeoffs) enough of that Advanced Science Tech has been crammed into the thing to reduce its inefficiencies down into the statistical noise level and it will carry on supporting life indefinitely. (Which stretches plausibility a bit, I know, but for the sake of playability/convenience I am just assuming that the topping-up has been reduced to a level which can be ignored in the same way as all rest of the routine maintenance that our kerbonauts keep busy with when we're not watching.)

And as a development alternative, either under "Large Scale 'Ponics" attached after Very Heavy Rocketry or simply under Very Heavy Rocketry itself - although, really, you shouldn't get this without also getting 'Ponics, hence the suggestion of a dependent node - the "Agricultural Greenhouse", which is similar to the Basic Greenhouse, but is much, much bigger (a tall 3.75m part or better yet an inflatable) for use on those stations that plan on growing more food than they need and resupplying other ships. Presumably in exchange for their CarbonDioxide, WasteWater and Waste, although I'm also thinking about ways to get those from primary sources.

So, continuing, then, I did some thinking along these lines using ArcFurnace's numbers:

I'd like to avoid multiplying models too much, especially since I don't _have_ any more models, so while I might ideally like the Basic Greenhouse to be somewhat bulkier, let's say both it and the CELSS greenhouse are the 6x density version (the former just having a _really_ cramped layout. The Basic masses 9 tons. The CELSS can support two kerbals indefinitely due to all that Advanced Science Tech they crammed into it reducing its inefficiencies to statistical noise; the Basic runs at the same 90% efficiency as the TAC converters. (But it doesn't produce its wastage as Waste; we disappear it, assuming that the 10% lossage is outgassing, leakage, and as-yet-unrecyclable waste products dumped over the side.)

The tradeoff for using the CELSS, once you have it, is that with all that AST in there, it masses a bit more (call it, say, 11 tons) and has a higher basic power budget to run the automation.

Basic power budget is that 14.4 kW (call it 15 EC/s per Kerbal; 30.0 EC/s all-in); we'll go so far as to x1.5 it for the CELSS greenhouse to represent the cost of all its fancy automation, so 22 EC/s per, 44 EC/s all in. But that's just to run the pumps, etc. And that's also all you need to provide it with (a) with the shutters open, and ( in Kerbin orbit or sunwards of there. With the shutters closed or in the far outer system (past Jool, say - haven't run those numbers yet), you need to cough up up to another 79.2 kW (call it 79 EC/s per Kerbal; 158 EC/s all-in). In between, it has to be topped up proportionately.

Let's see. IIRC TAC uses 24-hour days right now, so that's: hm, yes. With both TAC converters running, you're paying 250 EC/day for 0.9 units each 277 EC for a Kerbal-day of Oxygen and Water. With the CELSS you're paying 1,900,800 EC for the same plus a Kerbal-day of food, so call it 1,267,200 EC/day equivalent. And that's assuming that you're getting a full load of natural light.

Wow. That's really unfavorable. And one had better be running KSPI, 'cause those numbers - unless I've made some hideous math error, which would be nice right about now - are the sort of thing that ought to come out of MegaJoules, not ElectricCharge.

Also, some really nice power efficiencies on those TAC converters, huh?

Even assuming the later numbers for the airponics, it still doesn't come out looking exactly pretty.

…I have no idea how to dress that up into something anyone would ever actually use. (Well, I mean, I'm a KSPI user and I might consider using these and reactors to power them out of realism-fetish, but it's also hard to see why my Kerbal design engineers would make this choice.) At least until we have a mod that simulates kerbals going ape**** and throwing themselves out the airlock, having been driven stark raving mad by eight solid years of the freeze-dried ration bar diet…

Tomorrow I shall consider the spirulina option.

-c

[Hattivat]

^Sounds reasonable. As far as conversion efficiency is concerned, that depends first and foremost on the efficiency of your bioreactor (waste treatment), which I guess we assume is integrated into the greenhouse, much more than on the greenhouse itself. Basically the more toxic/resistant waste your bacteria/fungi/algae/whatever can decompose, the less "leakage" your system has. Right now it is assumed that part of the human feces and all the things like used clothes and food packaging are non-recyclable in space, so they are assumed to be either stored indefinitely or airlocked, hence the mass conversion losses. If you could achieve 100% recycle-bility (probably by developing some heavily genetically modified organisms and having all of your kerbals use biodegradable materials whenever possible (toothpaste, clothes, etc.)), then efficiency of close to 100% for the whole system is not that difficult to imagine.

If you are concerned about power consumption, you should definitely consider the spirulina option, ArcFurnace's calulations suggest it would be much more energy-efficient. If on the other hand you wanted some sort of a luxury option, one of the papers I linked (the Chinese one) discusses raising silkworms for food in space, and it seems reasonably efficient. (I know eating worms doesn't exactly sound too well either, but after a year of eating nothing but green goo and veggies it may sound like salvation to your astronauts )

[Temeriki]

Not gonna lie, you've defiantly piqued my interest. The lack of greenhouse in tac makes me sadface. I know you are all gung ho about making the greenhouse closed cycle, but I have a simple request. Now we all know 100% efficiency is impossible with damn entropy, but hey its a game/simulator and people are gonna play how they want. but not gonna lie I would like a close too 100% efficiency greenhouse so heres my idea. How about a "minerals" resource, if you're in space you need a small supply of minerals for the greenhouse, landed tho you automatically harvest them (any planet/moon has unlimited supply, use the landed tag). This would shut up the physicists and make teh mod still easy to use. Dont bother with a drill or anything like that, maybe make an option in the greenhouse to harvest minerals or something. Either way its your mod and keep up the good work.

[marce]

Cerebrate: interesting, it seems we started working on exactly the same thing at exactly the same time

Anyway, since it doesn't make sense to do the same work twice: I don't see a plugin on GitHub, so if you want a specialized generator/converter I can help you out, that part is finished. But what i wasn't able to accomplish (and I also didn't get an answer on my request for help) is real sun tracking. Adjusting the light requirement based on available sunlight is working, but the whole thing doesn't rotate. So if you know how to accomplish that (my research showed that maybe fixed meshes would be required) you could put your time on it and don't waste it on backend programming.

If you are interested drop me a line.

[Cerebrate]

Not gonna lie, you've defiantly piqued my interest. The lack of greenhouse in tac makes me sadface. I know you are all gung ho about making the greenhouse closed cycle, but I have a simple request. Now we all know 100% efficiency is impossible with damn entropy, but hey its a game/simulator and people are gonna play how they want. but not gonna lie I would like a close too 100% efficiency greenhouse so heres my idea. How about a "minerals" resource, if you're in space you need a small supply of minerals for the greenhouse, landed tho you automatically harvest them (any planet/moon has unlimited supply, use the landed tag). This would shut up the physicists and make teh mod still easy to use. Dont bother with a drill or anything like that, maybe make an option in the greenhouse to harvest minerals or something. Either way its your mod and keep up the good work.

Actually, that's kinda-sorta-maybe possible under my current game plan, just by avoiding the CELSS version of the greenhouse. The Basic would have 90% efficiency and slowly run down, which is something you could cure by shipping in more Oxygen/Water/Food from outside, or by finding somehow some CarbonDioxide/WasteWater/Waste to add to the system. And since the Agricultural Greenhouse's whole shtick is growing more than you need, I've already been thinking in terms of the latter as an ISRU problem...

- c