Table of Contents
Water Filtration Experiment for Kids: What It Misses
A water filtration experiment kids can build from a plastic bottle, plus the tests that prove the clear water coming out is still not safe to drink. Ever.
A water filtration experiment kids build from a plastic bottle produces something genuinely impressive: muddy brown water goes in the top and clear water comes out the bottom. And here is the thing that has to be said before anything else, loudly and more than once. Nobody drinks it. Not a sip. A column of gravel and sand removes the things you can see and almost nothing else. The dissolved salt sails straight through, which you can prove in ten minutes, and any bacteria or viruses pass through as well, which you cannot see at all. That gap between clear and safe is the most important thing in this article, and it is also the real lesson.
Key Takeaways
- You’ll build a layered gravel, sand and activated carbon filter in a cut bottle, run measured dirty water through it, and test what came out.
- It teaches filtration as size-based separation, adsorption by carbon, and the hard limit: dissolved and living contaminants are not removed.
- Best for ages 7–14, with an adult cutting the bottle and making the dirty water from clean tap water and garden soil.
- The filtered water is not drinkable and must never be tasted. The salt test proves why.
- The usual failure: unwashed filter media, which makes your output dirtier than your input.
What You’ll Need
| Item | How much | Where to get it | If you don’t have it |
|---|---|---|---|
| Clear 2 litre plastic bottle | 2 | Recycling | A large plastic jar; an adult cuts it |
| Pebbles or aquarium gravel | 1 cup | Garden centre, pet shop | Washed stones from the garden, scrubbed clean |
| Coarse sand | 1 cup | Builders’ merchant, play sand | Grit from a garden centre |
| Fine sand | 1 cup | Play sand, sieved | Sieve the coarse sand through a kitchen sieve |
| Activated carbon | Half a cup | Aquarium shop, pocket change | A crushed charcoal-based water-filter cartridge, or leave this layer out and note the difference |
| Cotton wool or a coffee filter | 1 | Bathroom, kitchen | A clean cut-up cotton sock |
| Clean jars for samples | 4 | Kitchen | Clear plastic cups. Label every one “DO NOT DRINK” |
| Garden soil | 2 tablespoons | Your garden | Never use water from a pond, stream, puddle or drain |
| Cooking oil, food colouring, salt | 1 tsp, a few drops, 1 tsp | Kitchen | These are the contaminants you add on purpose |
| Measuring jug | 1 | Kitchen | Any marked container |
| Dark plate or saucer | 1 | Kitchen | For the salt evaporation test |
| Soap | For handwashing | Bathroom | Not optional |
Time: prep 25 min, filtering 30 min, salt test 1–2 hours in the sun Cost: activated carbon is the only likely purchase, about the price of a coffee Mess level: do this outside or in a sink
Do the Experiment: A Water Filtration Experiment Kids Can Test Properly
Step 1 — Build the column
An adult cuts one bottle in half and tapes the cut edge, which is sharp. Invert the top half, neck down, into the bottom half so it works as a funnel. Push a plug of cotton wool or a coffee filter into the neck.
Step 2 — Wash every layer before you use it
Rinse the gravel, both sands and the carbon separately under the tap until the water running off is clear. This takes longer than you expect and it is the step people skip. Unwashed sand and carbon dust will come straight out into your sample and make the filter look useless.
Step 3 — Layer it finest at the bottom
On top of the cotton plug, add 3 cm of activated carbon, then 4 cm of fine sand, then 4 cm of coarse sand, then 4 cm of gravel. Fine at the bottom, coarse at the top. Water meets progressively smaller gaps on its way down.
Step 4 — Make the dirty water deliberately, from clean water
In a jug, mix 1 litre of tap water with 2 tablespoons of garden soil, a few drops of food colouring, a teaspoon of cooking oil, a teaspoon of salt and some torn leaves or paper. Stir. Keep 100 ml aside, unfiltered, as your “before” sample. This water is made from clean tap water on purpose, so that nothing dangerous is ever in the room.
Step 5 — Pour 250 ml through slowly and collect it
Pour gently down the side so you do not dig a crater in the top layer. Collect what comes out in a labelled jar. It will take several minutes. Pour a second 250 ml through and keep that separately; the second batch is usually clearer than the first.
Step 6 — Compare before and after, four ways
Hold both jars against a printed page and see which letters you can read through each. Shine a phone torch through both from the side. Look for floating debris and oil film. Note the colour. Write all four observations down for both samples.
Step 7 — The salt test, which is the whole point
Put 20 ml of your filtered water on a dark plate and leave it in the sun or a warm place until it dries completely. Do the same with 20 ml of plain tap water on a second plate. White crystals left behind on the first plate are the salt your filter did not remove. Nobody tastes anything to confirm this.
What you should see: The first 250 ml often comes out cloudy, because the column is still settling and shedding fines. The second batch is usually dramatically better: clear or near-clear, with the floating debris gone, the oil mostly trapped in the top layers, and the colour noticeably paler if you used activated carbon. The letters test is convincing. And then the salt plate ruins the triumph in the best possible way: there are visible crystals on it, and there are none on the tap water plate, which proves that something you added is still in there and that your eyes cannot find it.
What’s Actually Going On
Your column does two different jobs at two different scales.
The gravel and sand layers do mechanical separation. The largest particles are simply too big to pass between the grains, which is straining. Smaller particles that could fit through still get caught, because they collide with grain surfaces on the winding path down and stick there. That second mechanism, called depth filtration, is why a thick sand bed removes far more than a sieve of the same pore size would. It is also why the filter gets better for the first few minutes and then eventually clogs.
The activated carbon does something else entirely: adsorption. Activated carbon is processed to be riddled with pores, giving it an enormous internal surface area, and organic molecules stick to that surface rather than being strained out by size. That is what pulls out some of the food colouring, and it is why carbon cartridges are sold for taste and odour. It does not remove salts, because dissolved ions do not stick that way.
Which brings us to the limit. Filtration separates by size and surface chemistry, and dissolved salt is neither large nor sticky. Sodium and chloride ions are a fraction of a nanometre across; the gaps in your sand are measured in hundreds of micrometres, hundreds of thousands of times larger. Asking this filter to remove salt is like asking a tennis net to stop sand. Your evaporation plate makes that argument visible without anyone having to take your word for it.
The pathogen point matters even more, and the numbers are stark. The CDC’s guidance on water treatment for hiking and travelling states that most portable filters remove parasites but not viruses or all bacteria, that a filter needs an absolute pore size of 1 micrometre or smaller to remove parasites, and 0.3 micrometres or smaller to remove bacteria as well. Your sand column’s gaps are hundreds of times wider than that. The CDC also says, in as many words, that clear water can be unsafe, and that water containing harmful chemicals cannot be made safe by boiling or disinfecting at all.
There is one genuine complication worth telling an older child, because it is interesting and it is not a loophole. Slow sand filters, used in real treatment since the nineteenth century, do remove pathogens, and well. But they do it biologically: a living layer called the schmutzdecke develops on the sand surface over weeks of continuous slow flow, and the microbial community in it consumes and traps pathogens. Muhammad and Hooke characterised that biomass and its effect on filter performance in Environmental Technology in 2003. Your twenty-minute bottle has none of it.
Finally, the comparison that makes the whole thing click. The CDC describes municipal treatment as a sequence: coagulation, flocculation, sedimentation, filtration, and then disinfection, usually with chlorine. Filtration is the fourth step of five, and the step that makes water safe to drink is a separate one that comes afterwards. Your bottle is step four. The EPA’s overview of drinking water treatment technologies sets out the same architecture at scale.
How to Explain It to Your Kid
Ages 5–8: clean-looking is not the same as clean
Let them do the pouring and the comparing, and keep the rule simple and absolute: we never drink anything from an experiment. Explain that the filter is a maze with gaps that get smaller and smaller, so big bits get stuck and small bits keep going. Then show the salt plate and ask where they think the salt went.
Ages 9–12: size, and what is too small to catch
Introduce the idea of scale with numbers they can hold. The gaps in the sand are about the width of a pencil line. A germ is about a thousand times smaller than that. So the germ goes through, easily, and no amount of looking at the water will tell you. This is also the right age to introduce the five steps of municipal treatment and ask which one their bottle is.
Ages 13+: mechanisms, pore sizes and the biological filter
Give them straining, depth filtration and adsorption as three distinct mechanisms, then the actual numbers: 1 micrometre for parasites, 0.3 micrometres for bacteria, reverse osmosis for viruses and salt. Then introduce the schmutzdecke as a case where a filter works for a reason that is not filtration at all. The real investigation: measure flow rate against time and watch the filter clog, which is the trade-off every water engineer manages.
The question to ask: “If you filtered seawater through this and it came out perfectly clear, would it be fresh water? How would you check without tasting it?”
Why Teaching This Early Pays Off
The transferable idea is that absence of visible evidence is not evidence of absence. Clear water looks safe, and that intuition kills people. It also drives a lot of bad consumer decision-making, from filter jugs sold as purifiers to the assumption that a mountain stream must be clean. The CDC says plainly that drinking straight from a clear mountain stream is generally not safe, and a child who has watched salt survive a filter has a physical reason to believe it.
The second payoff is the idea of a treatment train: a sequence of steps where each does a specific job the others cannot. Children, and adults, look for the single thing that solves the problem. Real systems almost never work that way.
The National Research Council’s framework names “constructing explanations” and “engaging in argument from evidence” among the practices of science, and the evaporation plate is a small, complete example of both. Be honest about the scale of the claim, though: this is one afternoon, it will not shift anyone’s grades, and the hands-on learning literature does not support that. What it reliably produces is a child who asks what a filter actually removes before trusting it, which is a better outcome than it sounds.
Before You Start: What Parents Should Know
| Best age | 7–14. The building is fine from 7; the micrometre arithmetic suits 11 and up |
| Supervision | Close. An adult cuts the bottle, tapes the sharp edge, and mixes the dirty water |
| Real hazards | The filtered water is not safe to drink and must never be tasted, by anyone, for any reason. Make the dirty water yourself from clean tap water and garden soil, never from a pond, stream, puddle, drain or water butt. Soil carries bacteria, so wash hands thoroughly with soap afterwards, and keep it away from mouths and cuts. Activated carbon dust irritates airways, so rinse it wet rather than pouring it dry. The cut bottle edge is sharp until taped |
| Don’t | Don’t substitute real environmental water for your made-up dirty water, don’t collect samples in a drinking glass or a sports bottle, and don’t add bleach, chlorine tablets or any disinfectant. Water disinfection is a real procedure with real doses and it is not a children’s activity |
When It Doesn’t Work
- The output is dirtier than the input. Your media was not washed. Rinse the sand and carbon until the rinse water runs clear, rebuild, and try again. This is the single most common failure.
- Nothing comes through. The cotton plug is too tight or the fine sand is too deep. Loosen the plug and reduce the fine sand layer to about 2 cm.
- It drains straight through in seconds and nothing is removed. Not enough fine material, or you poured a hole through the top. Add more fine sand and pour gently against the side of the bottle.
- The oil is still floating on top of the filtered water. Expected. Oil does not dissolve and much of it stays on the surface, so it gets left behind in the jug rather than removed by the filter. Note it as a result.
- The colour did not change at all. Either you skipped the activated carbon or there is not enough of it. Carbon is the only layer that touches dissolved colour, and even then it is partial.
Take It Further
- Change one variable: carbon. Build two identical columns, one with the carbon layer and one without, and run the same dirty water through both. Compare the colour of the two outputs against a white page. This is the cleanest controlled comparison in the activity.
- Measure the clogging. Time how long each successive 100 ml takes to pass through. Plot time against the volume already filtered. The rising curve is the filter loading up, and it is exactly the measurement that tells a water engineer when to backwash.
- Compare with the real thing. Look up your local water utility’s published water quality report, which most are required to produce, and find the list of things they test for. Counting how many of those your bottle addresses is a sobering and excellent exercise.
Our piece on water engineering with dams and filtration projects sets this in the wider context of how water actually reaches a tap. For the separation method that does remove salt, see the solar still, and for a different measurement of water quality that your eyes also cannot do, try testing the pH of rain, tap and pond water.
What to Watch For Over the Next 3 Months
- Week 4: They ask what a filter jug at a friend’s house actually takes out, rather than assuming it makes water safe.
- Month 2: The phrase “that doesn’t mean it’s clean” appears about something unrelated, which is the real transfer.
- Month 3: They distinguish between steps in a process rather than looking for one fix. If they ask which step of water treatment kills the germs, they have the whole idea.
Frequently Asked Questions
Can we drink the filtered water if it looks completely clear?
No. Never. Clear water can carry bacteria, viruses and dissolved chemicals, none of which this filter removes, and the CDC is explicit that clear water can be unsafe. The whole point of the activity is to prove that your eyes are not a safety test. Treat every sample as waste.
Why bother building it if it does not make water drinkable?
Because it does the fourth of the five steps real water treatment uses, and doing one step properly is how you come to understand why the others exist. Removing suspended solids is genuinely necessary: disinfection works poorly in cloudy water, which is why filtration comes before it.
What would actually make the water safe?
In a real emergency, the CDC’s advice is that boiling is the best method, at a rolling boil for one minute, or three minutes at elevations above 6,500 feet. Filtering first and then disinfecting is the next best option. None of that applies here, because your sample also contains soil and is not worth treating. Pour it away.
Can we use water from the pond or the stream instead of making dirty water?
No, and this is the most important substitution to refuse. Environmental water can carry Giardia, Cryptosporidium, E. coli and more, and the activity involves handling it, splashing it and putting it in jars in your kitchen. Garden soil in tap water looks just as dramatic and introduces nothing you did not already have.
About the author
Ricky Flores is the founder of HiWave Makers and an electrical engineer with 15+ years of experience building consumer technology at Apple, Samsung, and Texas Instruments. He writes about how kids learn to build, think, and create in a tech-saturated world. Read more at hiwavemakers.com.
Sources
- Centers for Disease Control and Prevention. “How Water Treatment Works.” https://www.cdc.gov/drinking-water/about/how-water-treatment-works.html
- Centers for Disease Control and Prevention. “About Water Treatment Options When Hiking, Camping, or Traveling.” https://www.cdc.gov/drinking-water/prevention/water-treatment-hiking-camping-traveling.html
- Centers for Disease Control and Prevention. “About Choosing Home Water Filters.” https://www.cdc.gov/drinking-water/prevention/about-choosing-home-water-filters.html
- U.S. Environmental Protection Agency. “Overview of Drinking Water Treatment Technologies.” https://www.epa.gov/sdwa/overview-drinking-water-treatment-technologies
- World Health Organization. “Drinking-water quality guidelines.” https://www.who.int/teams/environment-climate-change-and-health/water-sanitation-and-health/water-safety-and-quality/drinking-water-quality-guidelines
- Muhammad, N., & Hooke, A. M. (2003). “Biomass characterization of slow sand filtration schmutzdecke and its effects on filter performance.” Environmental Technology, 24(1), 43–50. https://doi.org/10.1080/09593330309385534
- National Research Council. (2012). A Framework for K-12 Science Education: Practices, Crosscutting Concepts, and Core Ideas. National Academies Press. https://doi.org/10.17226/13165
