Wednesday, October 21, 2015


 One of the basic necessities not yet mentioned is clothing.

 Most clothing is made, not surprisingly, of cloth. That in turn is made of woven threads, which are made either of plastic, animal hair or plant fiber.

 I'm going to eliminate animal fibers simply on an efficiency basis. The varieties of animals that are raised for fur (sheep, rabbits, goats, yaks) can also produce meat and milk, but are much less efficient than purpose-bred varieties. Sheep are probably the best all-round performers (~5.4kg wool per animal per year eating about 2kg/day of feed), but we can still do much better than that. Of course any incidental fur, skin, etc. will be used for filling, lining, leather and such, but we cannot assume there will be enough of that material to clothe everyone.

Read on for the rest. This turned out to require a lot more growing space than I would have thought, 10.26m² per person.

 Plastic has many benefits. Some routes to plastic production don't involve living organisms as a bottleneck. Others use byproducts of food production. Depending on the choice of plastic the resulting fabric can be quite strong, durable and chemical-resistant. Dyes can be incorporated directly into the fibers for simpler, more durable coloring. Plastics can also have drawbacks like static buildup or melting, and tend to be degraded by UV exposure.
 On Earth, the main clothing fibers are nylonpolyurethane (as Spandex), acrylicrayon and polyester. Nylon and polyurethane are fully synthetic, typically at the end of a long chain of processes that start with benzene. Acrylic is also fully synthetic, most commonly encountered as cheap knitting yarn. Rayon is a cellulose polymer that typically starts life as wood or bamboo fiber; it is considered semisynthetic even though it is just as heavily processed as nylon. Polyester refers to a family of plastics, some of which are found in nature. Notable polyesters include PET (PET as Dacron, BoPET as Mylar), Vectran and PLA.
 Notice that none of those are my favorite plastic, polyethylene. I suspect PE fabric for clothing would be uncomfortable, with no stretch and a slightly oily feel. Excellent as an outer layer in a protective garment but terrible for underwear.

 Plant fibers also have many benefits. Softness and durability are at the forefront. Natural fibers tend to withstand flexing better than synthetic fibers. These tend to char rather than melt. They also have a softer feel. On the other hand, natural fibers are more difficult to dye and can be more labor-intensive to produce.
 On Earth the main clothing fibers are cotton, flax (linen) and hemp. All three are grown for their fibers and seeds; the seeds can be eaten, milled into flour or pressed for oil. All three are suitable for hydroponic cultivation, mechanical harvesting and mechanical processing.

 Regardless of type, the various fibers are spun into yarns. Often the yarn is composed of several different fibers in order to combine desirable properties. For example, it is very common to include a few percent Spandex with cotton or nylon to produce a yarn with some elastic response.

 From yarn, some types of clothing are knitted directly. Machine-knit sweaters and socks are common. Other types require that the yarn is woven into fabric, which is then cut and sewn. One key variable is the weight of the fabric, typically specified in ounces per yard or grams per square meter. I'll be using GSM / grams per square meter. For example, a thin t-shirt might use 100 gsm material while a resort hotel might offer super-thick 800 gsm bath towels. I've used values below that are in the middle of the range for each fabric type.

 Fabric amounts are handled oddly in the US. People specify some number of yards of length, but often fail to mention that it's not square yards. Fabric is typically sold in widths of 45" or 60" (sometimes 32"). If a pattern calls for 2 yards of 45" fabric, that's actually 2.5 square yards. I've listed values below in square meters. I used a variety of sources; there was one excellent sewing blog with tables for the first three items, then I found a pattern for coveralls. For the bath and bedroom items I used actual dimensions, added an allowance for hems and converted to metric.

Typical fabric requirements:
pants, 1.5-4.2m² (2.5m² avg.) x 250gsm = 625g
shirt, 0.7-3.4m² (2m² avg.) x 150gsm = 300g
dress, 1.3-7.7m² (3m² avg) x 200gsm = 600g
coveralls, 4.3-6m² (5.3m² avg.) x 350gsm = 1855g
sheet (queen): 6m² x 150gsm = 900g
sheet (full): 5m² x 150gsm = 750g
blanket (single-layer fleece): 7m² (90x110"/ 230x280cm) x 350gsm = 2450g
quilt (2-layer without filling): 14m² (90x110"/ 230x280cm) x 200gsm = 2800g
pillowcase: 1.4m² x 150gsm = 210g
napkin: 0.4m² x 200gsm = 80g
towel (face / washcloth): 0.1m² x 350gsm = 35g
towel (hand): 0.4m² x 350gsm = 140g
towel (bath): 0.9m² x 500gsm  = 450g (also worth a read)

(Edit 2015-11-11: I found another link with some citizen-science that provides measured values for clothing. A women's outfit was 0.65kg and a men's outfit was 1.1kg. This suggests my clothing allowance below is high but not unreasonable. I also recommend Rob Cockerham's site on a personal level; lots of interesting content.)

 Let's assume each person has a full set of bedding and bath linens, a blanket, a napkin and one week of clothing. The linens come to 20.2m² of various weights, 4.865kg. Clothing depends a lot on the person's size, preference and occupation; I will add the first four values and divide by three to get an 'average' mass for a day's clothing. This is likely to be high as I doubt that a third of the crew will need heavy denim coveralls for daily work. Still, that comes to 29.9m² or 7.887kg for seven outfits. That's 12.75kg in total; add a bit of leeway for knits and undergarments (~1.5kg), thread, sizing, etc. and call it 15kg per person. Specifics may differ; I'm assuming a full-size bed and all single people. A couple could use the same sheet and bathroom set. A more regimented facility might use Navy-style hot bunks and hot air drying after bathing to cut most of the linens. I suppose wealthy tourists might want more amenities.

 These things wear out. Socks and undergarments typically last perhaps half a year. Outerwear lasts 1-3 years depending on circumstances, so let's use one year. Bedding and bath items can last many years, but let's call it three. Heavy-duty work outfits like denim coveralls might last a decade or might wear out in three months; call it one year. On an annual basis using these numbers each person needs about 12.5kg of replacement fabric. These estimates are a bit conservative; many people replace their clothing far less often than this and only when they are actually worn through. I assume that the 'replaced' garments may not necessarily be discarded and people may simply build up a wardrobe over time. Sufficient space could be available to store 2-3 weeks worth of clothing, while worn or damaged pieces would be used as rags or as fiber sources for filter paper.

 Let's look at yields. I simply don't have enough information available to predict the material requirements for plastics, so this will focus only on natural fibers.

 Cotton can be grown at 1000-1500 lb per acre in open fields. In a hydroponic environment with no pests and tight nutrient control it should be possible to significantly exceed that mark, but let's use 1500lb/ac for this estimate. 4047m² per acre gives us about 0.37lb per m² or about 168g/m². Estimates range from 150-180 days for a growth cycle, so let's use 165 days. That works out to almost exactly 1 gram per square meter per day. Each person needs 34.25 grams per day, which would require 34.25m² per person. If we use a record yield of 6.31 bales per acre, that would be 3,536kg/ha or 354 grams per m² or 2.15g/m² per day > ~16m² per person. Seed production is about 1.62 x cotton production, or 3.48g/m² per day.
  Flax can be grown to yield over 1800kg/ha or 186g/m². Growth cycle is 90-125 days (in Canada), so we will use 108 days. That gives 1.72g/m² per day or 19.89m² per person.
 Hemp has yielded 6 tonnes of fiber per hectare or 600g/m² (again, Canada). Growth cycle is 70-90 days for fiber only (80) or 110-150 days (130) for dual crop. Using the dual crop number we get 4.62g/m² per day or 7.4m² per person. Seed production is about 0.7g/m² per day.

 Based on those numbers it sounds like a mix of 2:1 hemp and cotton would be ideal. That would be 4.93m² per person of hemp and 5.33m² per person of cotton to produce a total of 12.5kg of fiber. The process would co-produce about 1.25kg of hempseed and 6.77kg of cottonseed plus another few kg of plant waste.

It's worth mentioning that the record land yields tend to be twice the high averages, if not more, for most crops; even so, hydroponic methods can often surpass the record yields simply by preventing stress. Crops that have not been optimized for performance (like flax and hemp) have the potential to double or triple their yields given a dedicated breeding project. Have a look at dwarf rice and wheat yields vs. the varieties that existed before the 60's for an example of this in action. In other words, within a decade any number I post here will be obsolete as long as someone is actively developing the potential of these species. I think a combination of longer 'useful life' assumptions for clothing and improved hydroponic yields could cut the required area down below 3m² per person. That research and breeding could be applied to commercial crops on the ground, providing improved incomes for farmers around the world.


  1. I notice you've not included bamboo as a potential fibre source. It is an unusual fibre for clothing, I will grant that, but it is becoming increasingly used for things like socks and undergarments that need to breathe. The trick would be the methods used to process it into cloth, since some require reasonably harsh chemicals that may not be readily available in a colony. (I'm also having trouble finding how much bamboo is required per square metre of cloth.)

    Bamboo is fast growing, doesn't require much cultivation space and shows promise as a building material in other applications relevant to colonising space, so it seems reasonable that it will make up some portion of the textile needs as well. Particularly for the more commonly replaced undergarments.

    1. Bamboo has bast fibers, similar to flax and hemp. Most things referred to as 'bamboo fiber' are actually rayon made with bamboo as a feedstock, but true bamboo fiber is possible. The raw material is high in lignin, so processing is often a combination of mechanical chopping and chemical or enzyme boosted bacterial retting. Lots more details here:

      The tenacity of those fibers is less than half that of cotton. The resulting cloth won't be as durable or tough, but still falling between wool and rayon. That is to say, still useful and with a reportedly pleasant feel.

      The yield of bamboo basts can be as high as 53 tonnes per hectare (5.365kg/m² or 14.7g/m² per day). As long as the net fiber yield is at least 15% of the gross bast mass then it is competitive with cotton. If it is over 31% then it is competitive with hemp.
      yield source:

      If we look back a ways (1909), at least one author noted yields as high as 44 tons per acre (an eye-popping 98.6 tonnes per hectare) and fiber yields as high as 44%.
      Congressional Serial Set: pulp and paper investigation hearings. 1 January 1909. (available free)

      Let's go with the modern optimistic yield (while noting that high-intensity hydroponics can almost certainly double that value) and a fiber yield of, say, 38%. That would give a fiber yield of 5.6g/m² per day or about 6.1m² per person. The plants require a three-year lead time and a third of the stand is harvested each year. Waste from this process would be suitable for paper or fuel alcohol.

      I think the main drawback would be that bamboo grows very tall. A dwarf species could be found that grows to 4m after three years, but it's not certain the yield numbers would still apply. Even so, that's about 25m³ per person. Cotton by contrast grows to perhaps 130cm. Allowing 20cm for lighting and nutrient systems, that same 4m space could house two stacked crops. If a mild dwarf variety of cotton was developed that matured to 110cm or less, three stacked crops could occupy the same space and would become competitive again on a floor-space and volume basis.

      Another drawback is that bamboo requires several years to develop for this purpose. Peak fiber yields occur at three years. Peak structural strength is typically seen around 5 years. For paper pulp or wattle it can be taken at 1-2 years.

      Those drawbacks are certainly opposed by several advantages unique to bamboo. It's a structural material, can be used similar to wood (buttons, flooring, furniture) and resists bacterial and fungal attack. Bamboo resists high-pH environments and can be embedded in concrete; it could serve as the tension member in a reinforced regolith-block construction if metal is scarce.

  2. Beta cloth is also a good candidate for durable outer wear. It's non-flammable, completely recyclable and made from silica fibers, a major component of lunar regolith. The Apollo/Skylab spacesuits used it.

    1. Excellent idea. I'd probably not want to wear that against the skin and there's a potential inhalation hazard from shed fiber fragments, but as an outer layer with heat and chemical resistance it would be very useful for engineering coveralls. Equipment to make silica fibers would also be useful for making rockwool insulation and rooting media.

  3. I neglected to mention children. They need clothes a bit more frequently but they need less material for each set. Should even out.
    (Actually my kids go through clothes at about the same rate as the original post. In the last ten years I've bought 12 work shirts; 8 of them are still serviceable. I used a pretty high estimate to account for a variety of jobs and hobbies.)

  4. Snap! I see dwarf bamboo, genetically bonsai'd to mature at 2 meters high, as a source of all-purpose fabric, and its chips also double as a sawdust replacement when making pykrete.

    Also, my socks and underwear last a lot longer than six months. Just sayin'.

  5. That would be very useful. So much simpler to build something out of bamboo than go to the trouble of making plastic.

    I know; those numbers probably make me seem a bit crazy. They don't reflect my own consumption habits, but that's because I know I'm more conservative than average about buying new clothes. I used such aggressive numbers for two reasons: first, by adding new garments before the old ones wear out you can build up a stock of clothing in various states of wear. It's useful to have 'nice' clothes as well as 'painting' clothes, as it were. Second, I prefer to overshoot so there is enough margin to account for unusual wear for a given occupation and a bit of cushion in general. There was very little real data backing those estimated replacement rates. I'd imagine there are military sources with good data on wear for soldiers, but those numbers are likely to be a lot higher than for civilians. Might be a good fit for the engineering division / maintenance though.

  6. Quilts? We can maintain a constant, comfortable temperature. Why do you need a quilt?

    1. The production numbers in my post don't include an allowance for a quilt; I included it in the list of items for comparison. It is not very clear from the post itself, my apologies.

      I agree that a civilian habitat is likely to have independent climate control for occupants, and that keeping a 'shorts and short sleeves' environment to minimize clothing requirements is probably more effective than maximizing the climate system's efficiency.

      We could take this further and assume the station is kept at a comfortable temp for swimwear (or for nudists). That would cut the requirements down to perhaps 2kg per person, though some occupations would need more for safety reasons.

      I prefer 74 °F and about 40% RH, while my wife prefers about 82 °F and 60-70% RH. She would definitely want a quilt or heavy shawl if the station's temp was kept low. (She would also be extremely happy if we moved to the tropics, while I would be utterly miserable.)

      Central equipment is more efficient than distributed units in most cases. It is also more efficient (up to a point) to allow a larger hysteresis range of temperatures; a station may swing between 65 °F and 85 °F simply to keep the number of start cycles down.

      Someone planning to operate an actual station would have to trade between environmental system efficiency and wear vs. clothing allowances at various points.

  7. This one is good. keep up the good work!..

  8. There is a fashiong clothing store for women