Backcountry water treatment: matching the method to the threat

Water treatment is a matching problem, not a shopping problem. Three classes of organism live in surface water, they differ in size by a factor of several hundred, and no single method covers all three quickly, lightly and cheaply. Choose the method that covers the threat where you actually are, then run a routine that stops treated water being recontaminated on its way to your mouth.

Three classes of organism, three different problems#

Everything here follows from size and from chemical resistance. For scale, a human hair is about 70 microns across, so all three classes are invisible, and so are the differences between them.

Protozoa, 1 to 15 microns. Giardia cysts and Cryptosporidium oocysts. Large enough that any competent filter catches them, and wrapped in a tough wall that makes them the most chemically resistant of the three. Cryptosporidium in particular shrugs off halogen doses that destroy bacteria in minutes.

Bacteria, 0.2 to 5 microns. E. coli, Campylobacter, Salmonella. Filters catch them and every chemical treatment kills them quickly. The easy case.

Viruses, 0.02 to 0.4 microns. Norovirus, hepatitis A, rotavirus. One to two orders of magnitude smaller than protozoa, so they pass straight through a 0.1 micron membrane, and they are inactivated easily by chemicals and by ultraviolet. Whether they are plausibly present is the single question that decides if a filter alone is defensible.

The regulatory thresholds follow the same shape. The EPA purifier protocol asks a device to demonstrate 6 log reduction of bacteria (99.9999 percent), 4 log of viruses and 3 log of cysts. A device meeting all three is a purifier; one meeting only the bacteria and cyst thresholds is a filter. The word on the box is doing real work.

What each method covers, and what it does not#

Two rows deserve emphasis. Activated carbon is a taste and chemistry stage, not a microbiology stage, and it is often bundled inside a filter housing in a way that blurs the two. Iodine is the historical default that current guidance has moved away from, because Cryptosporidium is common in livestock country and iodine does not reliably deal with it.

How long each method takes, and what that costs#

Speed is a compliance question, not a convenience one: the slow methods are the ones people short-cut when they are thirsty.

Filters deliver on demand. A new hollow fibre cartridge typically flows in the region of 1 to 2 litres per minute by squeeze or gravity, falling steadily as the membrane loads with sediment. Backflushing recovers most of it.

Ultraviolet is roughly 90 seconds per litre in clear water, fast enough that nobody cheats, with batteries as the only real constraint.

Chemicals are the slow case: about 30 minutes for bacteria and viruses, and up to four hours against Cryptosporidium in cold water. The four hour figure is the one that gets ignored. Plan around it by treating tomorrow's water tonight.

Boiling costs fuel rather than time. Heating one litre from 10 degrees C to 100 degrees C takes 4.186 kilojoules per kilogram per kelvin times 90 kelvin, about 377 kilojoules. Canister fuel carries roughly 45.8 megajoules per kilogram, so at a realistic 45 percent system efficiency that litre costs 377 divided by 0.45 divided by 45,800, or about 18 grams of fuel. Four litres a day is around 72 grams of fuel a day, a large line item. Our stove fuel planning guide runs the same arithmetic for cooking, and stove types covers which burners hold that efficiency in wind.

Choosing by where you are, not by what is popular#

Local conditions genuinely change the answer here, so treat the following as a starting frame and check what the local public health authority says about the specific catchment.

• Upland water above human and livestock activity, in countries with functioning sanitation. The realistic threats are protozoa and bacteria. A 0.1 or 0.2 micron hollow fibre filter is a defensible sole method, and the fastest one.
• Downstream of settlement, a campground, a busy trail or grazing land. Viruses become plausible as soon as human waste enters the catchment. Use a purifier, or a filter followed by chlorine dioxide, or boil.
• Regions with widespread sanitation deficits, and travel where tap water is not potable. Assume viruses are present.
• Livestock country and agricultural runoff. Cryptosporidium is the design case: filter or boil, or use chlorine dioxide with the full four hour contact time. This specifically rules out iodine as a sole method.
• Glacial, desert and post-storm water. Turbidity is the controlling problem before any of the above applies.
• Freezing conditions. A wetted hollow fibre filter is destroyed by one freeze, so chemicals or melting and boiling snow are the workable options. Melted snow is not sterile: it carries whatever was on the surface it fell on.

Turbidity: the variable that breaks two methods#

Silt is not itself a health threat, but it disables the treatment that is. Ultraviolet works by delivering a dose, and NSF/ANSI 55 Class A specifies 40 millijoules per square centimetre delivered through water of defined clarity; suspended particles absorb the light and shield organisms behind them. Chemicals suffer a related problem, because organic load consumes the oxidant, so the dose reaching the organisms is lower than the dose you added. Filters simply block.

Settling is free and works better than people expect. A silt grain of about 10 microns, at a typical mineral density, settles in still water at roughly 0.09 millimetres per second by Stokes' law, which is about 32 centimetres per hour. A 20 centimetre deep container therefore clears most 10 micron silt in under an hour of standing still. A 1 micron particle settles a hundred times slower, about 3 millimetres per hour, so the finest glacial flour will not settle in any useful time and needs a physical pre-filter.

The sequence in dirty water is: scoop from just below the surface and away from the bed, settle for an hour if you can, pour through a bandana or a coffee filter, then treat. This roughly doubles or triples the working life of a cartridge in silty conditions, and it is the difference between an ultraviolet unit working and merely appearing to work.

The field routine that stops you undoing the treatment#

Good equipment used carelessly produces contaminated water. The recontamination pathways are boring and repeatable, which is why a fixed routine beats good intentions.

1. Choose the source. Moving water beats standing water. Get above human camps, trail crossings, grazing and agricultural runoff. A side stream entering a main river is often cleaner than the river.
2. Declare a dirty bottle and a clean bottle, and never swap them. Mark them. The dirty one goes to the water; the clean one holds treated water and nothing else.
3. Treat the threads and the cap. The classic chemical failure. Untreated water sits in the bottle threads and on the cap lip. Fill, dose, cap loosely, then at about five minutes invert and let treated water run over the threads before tightening.
4. Keep the filter output clean. The clean side of a filter and the inside of the clean cap must never touch stream water, wet hands or the ground.
5. Hands. Faecal-oral transfer from hands to food and bottle mouths is widely argued in the wilderness medicine literature to be a large contributor to backcountry gastrointestinal illness. We cannot quantify it, but hand sanitiser weighs 30 grams.
6. Write down the ready time. Guessing at "about half an hour" is how a four hour Cryptosporidium contact time becomes forty minutes.
7. Store treated water cold and dark. Chemical residual protects stored water; filtered water has no residual and can regrow bacteria in warm conditions.

Groups make this harder: if one person treats and everyone drinks from a shared container, one dirty hand contaminates the group. The same discipline matters more with small children, and the wider hygiene routine sits in our camping with children guide.

Check your setup before you leave home#

1. Baseline the flow rate. Time one litre of clean tap water through the filter and write it on the housing. In the field, when the time has roughly doubled, backflush; when backflushing no longer restores it, the cartridge is done.
2. Run the manufacturer's integrity test. Most hollow fibre cartridges have one, usually a back-pressure or bubble check on a wetted element. Learn it at home, and repeat it after any drop or suspected freeze.
3. Audit for freezing. Write the date of first wetting on the filter. Any wetted filter that has since been below 0 degrees C, including in a car, a garage or a checked bag, is retired. Store dried filters warm, as covered in gear storage.
4. Time an ultraviolet cycle and check the cells. Confirm the completion signal and carry the spare batteries, not the intention to buy them.
5. Weigh the options. Put each candidate system on a kitchen scale with everything a three day trip needs, including spare cells or the full course of tablets. A filter is often 60 to 100 grams heavier than chemicals but removes silt and needs no waiting. That trade lands in your base weight either way.

Common mistakes#

Believing a filter covers viruses. The most consequential misunderstanding on this page. A 0.1 micron pore cannot catch a 0.03 micron particle.

Cutting the Cryptosporidium contact time. Thirty minutes is the bacteria and virus figure. Treating at the source and drinking at the source defeats chemicals.

Using ultraviolet in cloudy water. The unit signals a completed cycle regardless. That is a statement about the lamp, not about the water.

Drinking off contaminated bottle threads. Cheap to prevent, easy to forget, and it undoes the whole exercise.

Assuming remote means clean. Beavers, marmots, deer, sheep and other hikers use the same water. Altitude is not treatment.

Storing a filter wet. Biofilm grows, and the element stays vulnerable to a freeze. Dry it thoroughly before it joins the spares and repair box.

Frequently asked questions#

Do water filters remove viruses?#

Standard backpacking filters do not. A hollow fibre membrane rated at 0.1 or 0.2 microns blocks protozoa (1 to 15 microns) and bacteria (0.2 to 5 microns), but viruses are 0.02 to 0.4 microns and pass through. Covering viruses needs a device sold as a purifier, or chemical treatment or boiling after filtering. Where human waste can enter the catchment, that second stage is not optional.

How long do I really need to wait with chlorine dioxide?#

About 30 minutes covers bacteria and viruses in reasonably clear water at normal temperatures. Cryptosporidium is the outlier and can require up to four hours, with cold water at the long end. Follow the contact times printed on the product, and if livestock or agriculture are anywhere in the catchment, plan on the long wait rather than the short one.

Is boiling still the most reliable method?#

Yes, for coverage. Boiling inactivates protozoa, bacteria and viruses without depending on pore size, water clarity or contact time. Public health guidance is a rolling boil held for one minute, and three minutes above about 2,000 metres, since water boils cooler as altitude rises. The costs are fuel, roughly 18 grams per litre in a typical canister system, and waiting for it to cool.

Does freezing damage a water filter?#

Once a hollow fibre element has been wetted, a single freeze can rupture the fibres from the inside. The damage is invisible and the filter usually still flows, sometimes faster than before, which is the worrying part. No reliable home test clears a suspect filter. If a used filter has been below 0 degrees C, replace the cartridge.

Can I skip treatment for a fast-flowing mountain stream?#

Moving water above human and livestock activity is genuinely lower risk than standing water below a campground, which is why source selection is worth doing carefully. It is not a guarantee. Wildlife shed Giardia and Cryptosporidium into clean-looking headwaters and you cannot see the difference. Treat it, and use source selection to reduce the load rather than to replace the method.

How do I deal with silty or glacial water?#

Settle, pre-filter, then treat. Let the water stand for an hour if time allows, since 10 micron silt drops around 32 centimetres per hour in still water, then pour it through a bandana or coffee filter. Turbidity blocks ultraviolet entirely, consumes chemical dose and clogs a membrane fast, so this step protects whichever method you use.

Is ultraviolet treatment worth carrying?#

It is fast at roughly 90 seconds per litre, covers all three classes of organism, and adds no taste. The trade-offs are that it needs clear water and working batteries, treats one container at a time, and leaves no residual protection in stored water. It suits clear sources and trips where recharging is possible, and pairs badly with silty rivers.