[This “MakeShift Challenge” column originally appeared in Make: Volume 03, 2005.]


The Scenario

It is easy to forget that access to potable water is considered a luxury for most of the world. You are reminded of this fact on a trip to a rural village in East Asia. You learn from the locals that their water supply has been contaminated — the cause of recent illnesses that sound a lot like cholera and dysentery. In addition to dirt, sewage, bacteria, and parasites, you suspect other contaminants such as arsenic and benzene from industrial dumping many miles upriver. Ideally, nobody should drink this water, but the villagers are unwilling to relocate.

The Challenge

Create a makeshift solution to filter and purify the water. The solution should be permanent and able to provide drinkable water for 20 to 30 people. Tools and materials at your disposal include what can be reasonably extracted from the environment and items on your supply list. You have 48 hours.

Supply List

  • (2) barrels
  • (1) bicycle with flat tires
  • (1) car battery
  • (6) 1-liter plastic bottles of water
  • Various lengths of bamboo tubes (1″ to 3″ diameter)
  • Variety of tools (saw, hammer, pliers, hand drill)
  • Steel wool
  • Endless supply of coconuts
  • $10 in mixed American coins

MakeShift 02: Analysis, Commentary, and Winners

by William Lidwell
August 08, 2005

Tragically, the MakeShift 02 challenge is all too plausible: you visit a rural village in East Asia that has a contaminated water supply. Serious illnesses are cropping up and spreading fast. The villagers have neither the knowledge nor the resources to do anything about it. Now the real-world challenge: the United Nations estimates that approximately 1.1 billion people in the world are forced to drink from unsafe water sources. As a result, millions of people die each year — most of them children. The most common culprits are physical contaminants (e.g., sediment and suspended matter), biological contaminants (e.g., bacteria, viruses, cysts, and parasites), and chemical contaminants (e.g., arsenic and benzene). The bad news is that these contaminants exist in some combination in virtually every water supply on the planet. The good news is that with a little ingenuity and education, most contaminated water can be made potable through the creative use of local materials. That is what this MakeShift Challenge is about: applying creativity to solve an important global problem, and educating others as to how it can be done. Thanks to all the Make: readers who took on this difficult and important challenge.


There are two basic approaches to this challenge: (1) solve the problem in a way that has nothing to do with the obvious parameters of the problem, and (2) find a way to purify the contaminated water.

Few readers opted for the non-obvious approach, which in this scenario would be to forget about purifying the contaminated water and seek alternative sources of potable water. J. Crossen was one of the few, and eloquently describes one such source:

A great water purifier would have self-manufacturing, solar-powered nanotechnology, and also be cheap. A coconut palm! The trees already use a combination of capillary action and semi-permeable membranes to purify the local water, and they even package it in biodegradable bottles. The water inside a healthy young coconut (aged 6 to 9 months) is perfectly sterile and contains a mix of electrolytes and nutrients similar to that of a sports drink. As well as keeping the villagers hydrated, it will keep their muscles working in top form. Each coconut contains about 750ml of water, and each villager will require at least two liters of water per day, so each villager will need about 2.5 coconuts per day. This means the entire village will need about 75 coconuts per day, which is a good yearly yield for a single coconut palm. With a small factor of safety for bad coconuts and dead trees, the village will need an orchard of about 400 palms to supply them with drinking water all year.

C. Granier-Phelps also embraced this “nutty” approach and offered a detailed process for efficiently harvesting the coconuts. I should note that Granier-Phelps is a subscriber from Caracas, Venezuela, who claims to “often throw coconuts at passersby,” so I think we should consider him an expert on the subject. Here is his procedure:

Flip coins to determine who gets to climb the coconut trees each day. Climb coconut tree. Throwing things (sticks, rocks, or the car battery) at the coconuts won’t work… believe me. Use a belt or cut one of the bicycle’s inner tubes and wrap yourself around the coconut tree’s trunk. Move the belt up with both hands and slowly climb the tree with the arches of your feet. Eventually you’ll be able to do it without a belt, but safety first.

Use a saw or a machete to cut the coconuts from the tree. Let them drop on the soft sand; avoid hitting village elders.

Climb down from tree. Avoid the urge to jump.

Using a machete or an edge of the car battery, open each coconut, trying not to spill the coconut water. First remove the outer thick skin and save it in one of the barrels for the night’s campfire; that way, you’ll keep dengue-carrying mosquitoes away. Make a small hole in the hard coconut shell that’s left after removing the skin.

Drink the coconut water or save it in one of the water bottles for later. Coconut water doesn’t hold for very long, so don’t cut any more coconuts than you need.

Open the hard coconut shell and eat the white skin.

Make a nice cup or bowl with the hard shell.

When you get sick of all that coconut water (although, literally speaking, you’ll be a lot healthier), take some time off to think of ways to purify the toxic waste mentioned in the article.

Enjoy your new, healthier body; tan in the tropical sun; learn to surf.

Any excess coconuts that fall from the trees should be dumped into the river in a designated area (demarcated using bamboo tubes and cable made from the steel wool). Ideally they’ll absorb the toxic water, filter the bad stuff using the skin’s fiber, and store clean water inside. Use them if you ever run out of coconut trees.

Assuming the villagers can be persuaded to center their existence around coconuts, this a perfectly viable solution for the short term. Long term, a village of 20-30 people will need about one gallon per person per day to live comfortably, for a total 20-30 gallons per day. And they will need near-potable water for sanitation, etc. So the coconut solution allows us to avert the crisis, but it will not give us the output required to be a total solution. What are the alternatives? A summary of possible approaches was provided by J. Earnest:

As we are attempting to engineer a way to purify contaminated water, a good starting point is to perform some basic research as to how modern systems are used in city water treatment facilities. Most water treatment facilities make use of four main stages of treatment. Filtration is the screening of large particles from the water. Flocculation involves the addition of chemicals which trap contaminants in the form of floc. Rapid sand filters then remove the floc particles. Finally, in the disinfection phase, chlorine gas, ultraviolet light, or ozone is used to kill any microorganisms that have survived the process thus far. Other methods such as reverse osmosis, boiling/distilling, carbon filtering, and ion exchange systems are used by some facilities.

This is a nice list and provides a good starting point. To help organize the alternatives, I offer the following mnemonic: FADD, for Filtration, Adsorption, Disinfection, and Distillation. Smart use of these four methods will render most contaminated water safe to drink. Here is a quick review of each before we look at some of the proposals.

Filtration works by blocking or trapping contaminants. Filters are particularly effective for removing physical contaminants. Examples of filters include sand, cloth, and charcoal. Adsorption works by attaching contaminants to a “sticky” molecular surface. It is generally used for removing chemical contaminants, but works for other types of contaminants as well. Examples of good adsorbents (chemically “sticky” surfaces) include activated charcoal and iron oxide (rust). Disinfection works by killing biological contaminants. You can disinfect water by adding chemicals such as iodine or chlorine or by boiling it. Distillation works by separating the water from contaminants by vaporizing it and then condensing it. The process removes all physical and biological contaminants and chemical contaminants with a boiling point above that of water. Each of these methods has their strengths and weaknesses. The key will be to consider the various approaches in light of the short time frame and limited resources available. Let’s first look at what is often referred to as a sand or layered filter. F. Valica describes one such design:

Get a few villagers to clean the coconuts, saving the meat and water for food. Keep the shells and scrape the hair from the coconut shells.

Start a fire and burn the coconut shells until you get ash/carbon. This may take longer than I’ve allotted for here, but you can build everything except the filter in the meantime.

Clean the inside of a three-foot-long, two-inch-diameter tube. Cut two two-inch rings from the bike tire. Strip the cloth seat from the bicycle. If it has a hard or impermeable seat cover, then find a piece of cloth. Put this cloth on over the end of the bamboo tube and stretch the bicycle tube rubber band around it. Use a larger-diameter bamboo tube for faster flow if the bicycle tube stretches larger.

Fill half of the tube with pea-sized (or smaller) pieces of coconut carbon. Coconut carbon has more micro-pores than traditional wood, coal, or lignite-based carbon.

Clean the change using steel wool and put it in the filter. If you have all pennies, use the steel wool or a file/rock to rub off some copper and expose the zinc. Although we are not dealing with pure alloys, the dissimilar metals should create a redux process and reduce levels of chlorine, attract iron and hydrogen ions, and remove heavy metals. According to the FDA, zinc and copper are good for you anyway, although you may only get a trace amount through this process. Now pack the remaining of the bamboo filter with coconut hair for mechanical filtration. I probably wouldn’t use steel wool here since it would rust. Cover the top of the filter with the cloth and use the second rubber ring as a rubber band, holding it in place. The filter should now be finished:

Clean both barrels using steel wool. You don’t know what’s been living in there. Rinse with coconut water. Cut a hole in the bottom of one to accept the bamboo filter. Make it a little smaller than the filter so the bicycle tube acts as a gasket between the barrel and the bamboo.

Cut off the rear end of the bicycle, and set it on some rocks. Place the second barrel on top. Arrange the barrels so the first barrel will pour into the second (through the filter).

Pour river water into the first barrel, and let it filter through to the second. Then boil the water in the second barrel. Pour boiled water into holding containers and cool to drink.

This is the first stage in Valica’s purification system and is a good example of how a layered filter works, though he kicks it up a notch with the electrochemical layer. This filter should be effective at removing most of the physical contaminants and some of the biological contaminants. His second stage boils the water, which will effectively kill the biological contaminants. As designed, however, it will not do as well with the arsenic and benzene. To address these chemical contaminants, we need to separate them from the water using adsorbents or distillation. A. Thornton used two adsorbent layers in his design:

If the problem is arsenic, then a fairly effective treatment method is to adsorb it with ferric oxide. I bet that bike frame is pretty rusty, and I’m sure it’s steel, not aluminum. File rust off of it. And if it isn’t rusty yet, file some steel from it, and boil that in a pan. Let the filings sit in the sun a while, and generally do whatever you can do to oxidize them quickly. Ferric oxide is going to take most of the arsenic out, but it won’t taste very nice and will discolor the water. However, we can ameliorate that and remove a great many residual nasties by doing a final filtration with activated charcoal. Where are we going to get activated charcoal? Why, both bamboo and coconut shells are excellent sources. Make a mound of coconut husks. Cover it with coconut leaves. Start a fire under it and let it burn; occasionally pour water on it to create steam, but not enough to put it out. After a while, the charcoal at the bottom of the mound will be activated charcoal, because it will have burned in an oxygen-depleted environment. Crush it up and you’re ready to put it in your filter.

Thornton’s one-two punch of using layers of rust and activated charcoal would do a good job of removing the chemical contaminants. Interestingly, mix a lot of rust and sand together and you get better absorbency than activated charcoal — on the order of 25%! Simple and effective. Activated charcoal by contrast, though almost magical in its ability to remove a wide range of contaminants, is not simple to make. As Thornton notes, source material is no problem: coconut shells. You can make plain charcoal by simply setting a bunch of coconut shells (or bamboo, etc.) on fire. Activating the charcoal is the hard part. To activate charcoal, you need to remove all of the tarry residues and non-carbon impurities that clog up its pores. There are two basic ways to do this: (1) soak the charcoal in an acid solution and then cook at high temperatures for a few hours, and (2) immerse the charcoal in super-heated steam (around 1,800°F) for 30 minutes. Method 1 may be possible in primitive conditions. Method 2 would be very difficult to do in two days in primitive conditions. Let’s look at proposals for making activated charcoal using both methods. V. Forgione describes a process of activating the charcoal using acid from the car battery:

CAREFULLY open the vent caps on the battery. The locals should have a plastic container to collect the acid from the battery. CAREFULLY pour the acid into the container. Now, you should have anywhere from 1.8 liters to over 4 liters of acid, depending on the size of the battery. Let’s just say we only need 1 liter of acid, since any more would cost you too much of your drinking water. Battery acid is about 36% sulfuric acid and 64% water. We should use 2 liters of bottled water to get the acid down to 9%. When mixing acid with water, add the acid to the water, NOT WATER TO ACID. HOT ACID WILL SPATTER! Pour 2 liters of water into another plastic container that the locals have provided, and SLOWLY add acid to the water, stirring all the while. You have 3 liters of acid and that should treat enough charcoal for our use. Soak the charcoal in the acid, and then reheat in the charcoal pile. With luck, this will activate enough of the charcoal to get the arsenic and benzene out of the filtered water.

If you get the acid solution right and cook it long enough, you have a good shot at activating a lot of the charcoal. Method 2 involves heating steam to 1,800 °F — no small feat in a primitive environment. Among the best proposals for doing this was M. Kissler’s:

The final ingredient in the second stage of the water purification will be activated carbon. You must construct an apparatus to make the carbon and instruct the villagers on its use.

Take the first large barrel and make five two-inch holes in the bottom using the drill. Make one two-inch hole in the side of the barrel near the top, large enough to snugly fit a bamboo tube of medium diameter. Take the top off the barrel and save it. This barrel will be used to make the activated carbon.

Nail a heavy piece of bark, three to four inches square, by one corner above the hole. This will be used to cover the hole when necessary. Bend the nail over inside the barrel so it doesn’t come out. (If, for some reason, you brought a hammer but no nails, you could just tie the bark around the top of the barrel and slide it down over the hole when necessary.)

Set the barrel up on a few rocks so the bottom is level and high enough off the ground to allow air to enter the holes in the bottom.

Drill an identical hole into the side of the second barrel to allow the bamboo tube to connect the two. Set this barrel over a small pit where a fire can be made and sustained. This barrel will be used for two purposes: it will heat the filtered water and kill remaining bacteria, and the steam produced will be piped into the first barrel to activate the carbon inside.

For now, take the bamboo tube out of the first barrel. Put kindling in the bottom of the barrel and light it. If, for some reason, you do not have access to fire, you could use the car battery to start a fire by *carefully* placing a strand of steel wool between the two terminals to create a spark.

Once you have a good fire going, add the coconut shells to the barrel. Do not pack them tightly: there must be air space between them.

Once the fire is strong, heap up dirt around the base to restrict the air access. Leave about a four-inch gap. Put the lid on the barrel, leaving the hole in the side open for smoke to exit.

A dense white smoke will come out of the barrel for a time. Bang on the side of the barrel as necessary to ensure the shells move and all burn evenly.

When the smoke turns from white to a thin bluish tint, most of the water has been driven off and the charcoal is now burning. Plug the gap in the bottom with soil and plug the hole in the side with the bark covering, filling all gaps with soil to make an airtight seal. The remaining burn will take about four hours.

Let the sealed barrel sit for half a day. Then, stick the bamboo tube in the holes on the sides of the two barrels so they are connected. Put the bottled water into the second barrel, and tightly close the lid. (In the future, the villagers will use their filtered water. You will need to place large rocks on the lids of both the barrels so the pressure from the steam doesn’t push the tops off. In addition, remove the dirt from around the bottom of the first barrel to allow for steam exhaust once it has passed through the charcoal — this will help to ensure that the steam displaces the air in the charcoal barrel.

Light a fire under the second barrel (the one with the water in it). This will heat the water and create steam and pressure. The steam and pressure help to activate the charcoal inside the other barrel. Let this go for at least one hour.

M. Kissler's "Pressure Cooker" for Activating Charcoal

M. Kissler’s “Pressure Cooker” for Activating Charcoal

Could this be made to work in a couple of days? I am skeptical that any of the proposed systems to superheat charcoal could be made to work in two days. It turns out that in Kissler’s case, however, it probably would not matter. The reason is that he cleverly incorporated another filtering material that is every bit as effective as activated charcoal:

…have another group of villagers collect a large quantity of the water hyacinth plants native to the area. Water hyacinth is a weed found in almost every water system on every continent, and is especially prevalent in East Asia. It has been found that its dried and powdered root is an excellent absorbing agent for arsenic in water. According to a report published by the Royal Society of Chemistry, filtering water using the weed reduces the arsenic content of water to below World Health Organization standards. Have the villagers dry and crush the roots.

Considered a bio-scourge to most ecosystems on the planet, the irrepressible water hyacinth is an excellent filter. If it is not in your part of the world yet, it likely will be soon. In any event, it definitely has a presence in East Asia and would be very effective at removing chemical contaminants. Finally, there are approaches that utilize distillation. The problem with distillation pertains to the chemical contaminants. R. Karnesky describes the problem:

While boiling water can be effective at killing bacteria, benzene is more volatile and arsenic less volatile than water. Therefore, neither making a still nor boiling in place will be effective at removing all contaminants. An activated charcoal filter will remove most contaminants in one shot. You don’t happen to have one of those, do you? Well, fortunately, you have plenty of coconuts, the shells of which are a popular way to obtain activated charcoal.

When Karnesky says “more volatile” and “less volatile,” he is referring the boiling points of the chemicals relative to the boiling point of water. The boiling point of water is 212°F. Arsenic is “less volatile” than water because it has a higher boiling point (1137°F). Benzene is “more volatile” than water because it has a lower boiling point (176°F). The implication of this is that boiling a water-benzene-arsenic solution would first vaporize the benzene, then the water, and then the arsenic (assuming you could get it to that high of a temperature). So with all this out of the way, let’s start with B. Doom’s solar still design:

My solution is based on the condensation of the non-potable water. If the water is evaporated and condensed into a clean container, the sewage, industrial pollution, and parasites will be separated from the water. One of the barrels is left in the sun and filled most of the way with water. Solar heat raises the temperature of the water and increases the humidity of the air trapped inside the barrel. The humid air escapes through vent ports cut in the top surface of the barrel, and travels up through bamboo shafts to the upside-down plastic bottles capping the shafts. The bottles are kept slightly cooler, and the water in the humid air condenses there, collecting in the bottom of the bottles. The level of the water must be kept at a height maximizing the surface area of the water/air junction to facilitate evaporation. To this end, a hole is drilled in one end at the given height and plugged with coconut shell sealed with coconut husk.

Figure 3. B. Doom's Solar Still

Figure 3.
B. Doom’s Solar Still

Stills are unique in their ability to simultaneously remove all physical and biological contaminants, as well as all chemical contaminants that are less volatile than water. Stills are also fairly simple to build and start to work quickly, which are important factors when the conditions are primitive and the timeline is short. Doom’s still would remove everything except the benzene. Volatile chemicals not only have lower boiling points than water, they have lower evaporation points. Accordingly, the benzene would evaporate into the bottles first and then condense with the water, giving the villagers benzene-water. Not good. The second problem is with output. Calculating evaporation rates is a messy business, but a quick back-of-the-envelope swag leads me to think that this system would only produce about 3-5 gallons per day. There are ways to deal with the benzene problem. The output problem is more challenging for a system based on evaporation. It is a good design, and with a few tweaks, it would be a winner. Thornton offered two such tweaks. He proposed a more traditional fire-heated still design that addressed the benzene problem using an old chemical engineering (and moonshining) trick:

So now, you heat your barrel over the fire you’ve made, taking care to keep it far enough off the fire that it doesn’t actually burn the barrel. Your water will start boiling; steam will come up the column, recondense on the steel wool, and drip back into the barrel. Eventually, however, steam will start making it out the top and you’ll get a flow out of the tube. You probably want to throw away the first bit of liquid that comes out of there. If you’ve got benzene in there, for instance, it boils at about 80 degrees Celsius. This is, at least in ethanol distillation, called throwing away the heads. Once you get a good stream going, this is mostly potable water, and if the villagers drank it, it probably wouldn’t kill them quickly. Most things biological will have been killed by the boiling, and most chemical contaminants will have sufficiently different boiling points so that by throwing away the heads and not running the still until it’s completely dry, they will generally be left behind as well. If you had to stop here, it’d be a decent stopgap solution.

“Throwing away the head” can be a very effective method of getting rid of the benzene-rich portion of the solution. Additionally, the specific density of benzene is lower than that of water. This means that if you let the condensed water stand in an undisturbed environment, most of the remaining benzene will rise to the top. Since evaporation is a surface phenomenon (unlike boiling, which occurs throughout a liquid), the benzene will evaporate first.

Thus, one approach would be to use a still, throw away the head to get rid of most of the benzene, and let the remaining solution gas out in an open barrel for a day, or boil it to burn off any remaining benzene. Draw water from the bottom of the barrel (just to be safe) and it should be potable. Stills also require far less maintenance than a filter (e.g., no cleaning or replacing of materials) and will work in a reliable, measurable fashion. The two main problems with distillation: (1) it requires a lot of energy as compared to a filter to produce the output required, and (2) it is only about 20% efficient, requiring five gallons of contaminated water to produce one gallon of distilled water. Is there a way to tell how much, if any, benzene remains? Benzene is clear and colorless, but it does have a distinctly sweet odor. Most people can smell benzene in water at two parts per million. Concentrations of benzene in water over five parts per billion are considered unsafe to drink. Suffice it to say that if you can smell benzene at this point, you shouldn’t drink the water, and you should start climbing coconut trees. Now that we are dizzy from the prospect of sniffing benzene, it is a good time to close.


Given a two day time limit with people already getting sick, drawing water from coconuts is the best short-term solution. It will provide pure drinking water with minimal cost and complexity. Long-term, however, a village needs water for sanitation, washing, cooking, and so on — more water than can reasonably be extracted from coconuts. So, short-term let them drink coconut water, but long-term we need a better solution.

Some form of rapid sand or biological filter using a combination of sand, gravel, and carbon is a logical choice. Ideally, you want to activate the carbon. I am skeptical as to whether carbon can be reliably activated in a primitive environment. To the extent that it can, I am skeptical that it can be activated in two days. This is where the water hyacinth proposal is particularly intriguing. The plant is an effective filter in both living and powdered form. So, in addition to being an element in a sediment filter, a basin filled with water hyacinths could be a simple and effective long-term component of an overall water treatment strategy. It is also a good back-up strategy in case you fail to fully activate the carbon. A problem intrinsic to all of the filtration strategies is that there is no easy way to test to what extent the water is potable except for drinking it and seeing if you get sick. This is where distillation carries some distinct advantages. Distillation enables you to visibly and reliably remove all of the contaminants except for volatile chemicals (in this case, benzene). And, as described above, there are ways of dealing with the benzene.

A conservative approach would be to have the villagers use the coconuts for drinking water and use the distilled or filtered water to meet sanitation and other needs. Longer term, I prefer the distillation process described to filtration. It is a simple method that the villagers can apply to achieve consistent results. If you go the filtration route, consider using iron oxide and water hyacinths, if available. They are as effective as activated charcoal and require far less energy to incorporate. That said, activate charcoal if you have the time and kung fu to do it! General factors to consider for all solutions: time to develop, rate of output, effectiveness at removing all types of contaminants, reliability, testability, resources, simplicity, long-term viability, process transferability, and safety.


I would like to again thank everyone who submitted solutions to the MakeShift 02 challenge. The submissions were very creative and well thought-out, and as before, selecting two winners from the batch was no easy task.

Winners receive Make: T-shirts to celebrate and show off their unique brand of genius and the ultimate MakeShift Master tool — the SWISSMEMORY USB Victorinox 512MB — equally useful for hiking and hacking. Honorable mentions get fame and recognition for their excellent contributions. In exchange, we at Make: expect great things. Go forth and solve the world’s problems!

Without further ado, the winners of the MakeShift 02 challenge are:

MakeShift Master — Plausible: Adam Thornton
MakeShift Master — Creative: Jesse Crossen
“Schmutzdecke” — Honorable Mention: Vinny Forgione
“A.A.B. Bussy” — Honorable Mention: Mac Cowell, Nick Cain, Barratt Park, and Brandon Carroll
“Eichhorina Crassipes” — Honorable Mention: Mark Kissler

Congratulations to the MakeShift Masters and the Honorable Mentions (applause…accolades…bowing). You all did a great job of taking on this difficult problem and communicating your solutions in a fun and effective manner. I encourage all readers to study these winning entries and share this link with friends. The first step in solving this global problem is education, so please help get the word out. And until the next MakeShift challenge, happy making!