Tuesday, October 13, 2015


Current hygiene in space leaves much to be desired. Ask an astronaut, or a recovery crewmember who pulls them out of their capsule. We've found a balance point that minimizes the use of consumables without causing significantly harmful effects, but it's a system most people wouldn't be able to stand for ten minutes, let alone years.

 Key to our clean civilization is soap, or more recently detergent. Humans have been making soap for at least 4,200 years, perhaps as much as 4,800 years. For the vast majority of that time the process was much the same: mix ashes, fat or oil and water, apply heat and hope the result doesn't dissolve your skin.

For the TL;DR:
Roughly 3.8kg of detergent per person per year. A mix of traditional soap and modern surfactants in the form of bar soap, liquid hand and dish soap, shampoo and laundry soap. Making these chemicals will cost fats, oils and sodium hydroxide. Making some of the modern compounds will require a bit more chemistry using the same raw material.

First, some background. Soaps and detergents are chemicals called surfactants. They are medium-chain hydrocarbons (10-18 C, averaging 12 C {dodecyl}) that have a polar end (attracted to water) and a nonpolar end (repelled by water). When these molecules are added to water, they tend to bunch together into blobs with the polar ends facing out. This forms a pocket at the nonpolar ends that can trap oily compounds, then the whole blob can be carried away by the water. A variety of chemicals can do this; some are better at removing heavy oils or grease while others are better at breaking up dirt. Most of them cause drying of the skin, and many can cause allergic reactions in some people.

The starting point for these chemicals is fat, specifically triglycerides. The traditional process was to treat the fat with ashes or lye (sodium hydroxide); this breaks the triglycerides into glycerol (glycerine) and fatty acids, which further react to form sodium stearate, sodium palmitate and related chemicals. The results may be purified and cast into bars; traditional soap was always solid.

 Modern industrial process usually breaks this into two steps. First the fats are broken down into fatty acids and glycerol. These products are separated and purified. Next the pure fatty acids are reacted with sodium hydroxide or potassium hydroxide for soaps. (Calcium and lithium can be used instead to make lubricating greases.)

 We actually use very little true soap today. Most surfactants are detergents like sodium lauryl sulphate; even most bar soaps are detergents rather than true soap. The raw material is typically palm kernel oil or coconut oil, broken into bare carbon chains with various functional groups tacked onto either end. The end result is the same: a molecule that aggregates in water to capture nonpolar molecules.

 Most soap or detergent products combine the active ingredient (detergent) with water, water softeners (detergents are disrupted by the 'hard' ions magnesium and calcium, but unaffected by the 'soft' ion sodium), moisturizers, fragrance, preservatives, bleach and/or abrasives. The colony's water will most likely be distilled and therefore very soft, eliminating the need for 'builders' or softening chemicals and in general reducing the amount of soap needed.

 So, our soapmaking process will require two major inputs: fats and lye.

 The hydrocarbon chain can be assembled using entirely synthetic means, but that consumes a lot of power and produces byproducts. If for some reason the colony is already using a Fischer-Tropsch process then appropriate-length chains from that output can be used for making soap. If the colony uses food animals, rendered fat (tallow) can be used as-is. Vegetable fats produce high-quality soaps, but the investment required in an oil crop is significant; some amount could come from animal feed processing, where a seed crop is pressed to expel oil for various uses and the high-protein low-fat meal is incorporated into feed pellets. A fourth option is to raise oil-producing insects on agricultural waste as an oil crop (or to extract oils from animal-feed insects raised on ag waste).

 The lye can be made via the Chloralkali process using sodium chloride, water and electricity as inputs into a membrane cell with sodium hydroxide, hydrogen gas and chlorine gas as products. Hydrogen and chlorine can be combined to form hydrochloric acid if desired, otherwise the hydrogen is useful for fuel and the chlorine is useful for a variety of industrial processes (including making bleach / sodium hypochlorite). If a source of metallic sodium or sodium carbonate is available, some of the chlorine stream can be used to convert that into sodium chloride to be fed back into the process.

 Next question is how much will we need? I will assume that bars are used for hand and body washing in personal quarters, but that all other soap is liquid. This is a pretty subjective question, but a random check of Google search results and forum posts suggests that the average large bar of soap (4oz/113g)  will last 2-4 weeks. Taking the most aggressive value yields just under 3kg of bar soap per person per year. A handwash reportedly consumes 0.35 grams of bar soap vs. 2.3 grams of liquid soap, but I think some of that is the pump dispensers providing more soap than necessary. Regardless, that suggests that a bar of soap can provide over 300 handwashes or around 15-20 showers. I will estimate 2kg of bar soap per person per year and assume the difference is taken up by liquid body wash and hand soap.

 Liquid soap (as in shampoo) averages perhaps 2 tsp / 10ml per use. Specific gravity of liquid soap tends to be just slightly over 1.0, meaning about 10 grams per use. Again a subjective topic, but people tend to shower once per day to once per two days depending on climate. The colony is not likely to experience large temperature swings, so every other day would be reasonable. Still, better to plan for one shower a day which means about 3.7kg of shampoo. (Note that only 10-15% of the mass of this product is detergent, or about 0.6kg.) Here's a link describing ingredients and their purposes for a typical product. Many people prefer to use a liquid body wash, using about 1 T / 15 ml per shower (5.5kg or 0.8kg detergent; after accounting for bar soap users, only 0.3kg is needed). All combined, that's roughly 1kg of detergent per person per year for personal hygeine. Note that toothpaste often contains small amounts of detergents, but the amounts are trivial in comparison.

 Laundry soap uses slightly different ingredients vs. other detergents. The primary ingredient is linear alkylbenzenesulfonate, which starts from the same 12-carbon chain but requires a benzene-based attachment partway down the chain. Average use varies widely by habits, climate, etc., but a rough range seems to be 1-2 loads of laundry per person per week. Cloth diapers, napkins, etc. tends to drive this to the high end, so let's assume 2 loads per week or about 100 loads per year. Many people assume they should fill the cap full of liquid for a load but that's an enormous waste; 1-2 T / 15-30 ml is enough in most cases. As with shampoo the actual detergent content of laundry soap is only about 15%, so the annual demand is about 0.5kg. If the chemistry is too difficult then SLS or soap can be used instead along with sodium carbonate (washing soda) and sodium tetraborate (borax) or salt. As an alternative, dry cleaning can be performed with supercritical carbon dioxide as solvent; this requires central or at least regional facilities.

 Dish soap is essentially the same as hand soap, though with fewer moisturizing ingredients and more water softeners. This is used only for hand-washed dishes; dishwashing machines use a different chemistry. Smaller cooking and serving items can be machine washed, but large items generally won't fit into a washing machine. That means perhaps a sinkful of heavily soiled dishes or pans per day to be handwashed even if a machine is available, 5-10ml of liquid or 0.14kg of detergent per person per year. I will use EnergyStar numbers for the rest of the analysis. That's a ballpark of 6 loads per week for four people, with a load defined as eight place settings. (Yes, 4 people x 3 meals should be 12 settings, but people don't normally sit down to a formal place setting for all three meals. My family uses a single glass each for water throughout the day, for example.)  Each load takes about a tablespoon (15ml) of liquid or about 0.18kg of detergent per person per year.
 A mechanical dishwasher uses a detergent mixture with abrasives, salt and a lot of softeners. It is possible to use liquid dish soap with added ingredients (salt, baking soda, possibly borax) and still get clean dishes. Automatic dishwashers use less detergent than handwashing; we can assume perhaps 0.15kg per person per year in addition to the 0.14kg required for handwashed pans, etc.
 Overall we can estimate about 0.32kg of detergent per person per year for dishes in the home. Machines used in restaurants are more efficient, using high-pressure sprays, mildly corrosive chemicals and higher temperatures to clean more dishes with less detergent. If "corrosive chemicals" sounds bad, consider that vinegar (that most natural of miraculous cleaning agents / food additives / condiments) is a mildly corrosive chemical. Anyway, if the colony serves most of its meals in a restaurant or mess hall setting then the detergent use will drop quite a bit.

 Cleaners for floors and surfaces also use detergents, and typically include a heavy dose of bleach for sanitizing. I'm having a terrible time finding usage estimates for these products, unfortunately, so for now I'm going to have to assume that the amount of detergent is small, within the margin of error of my other estimates and that most of the cleaning power actually comes from ingredients like vinegar, ammonia, bleach, soda, salt, etc. If anyone knows of a source to estimate the average use of these products I'd appreciate it.


  1. Any data comparing moon dust to pumice for industrial hand soap? I'm assuming our colonists will be working on dirty things with the animal and mechanical work and pumice does a great lob of reducing both water and soap consumption in that case.

  2. Lunar dust is made of very sharp fragments. I would be concerned about excessive embedding and tissue damage, plus toxic metal poisoning. It's made primarily of fractured impact remelt, which is essentially tiny slivers of glass and ceramics with shards of nickel-iron for flavor.

    Pumice is quite benign by contrast. Although it is a porous glass and can be sharp-edged, it is fragile and generally breaks down during use into dust.

    It would be possible to make artificial pumice. Start with regolith that has been screened (magnetic separation, volatile bake-out and fracture energy sorted), melt it under pressure, inject a volatile (water for example), then spray the results into vacuum or into a large volume of water. It is only an extra step or two vs. making rockwool, plus pumice is also an acceptable hydroponic media.

    Alternative materials include peanut shell fragments, eggshells or bone. A slightly more radical option would be to use a strigil to scrape off bulk grime before applying a surfactant.