Reverse Osmosis — the Itehil, and the One Thing It Genuinely Fixes

Introduction

Every previous volume in this dive ends at the same wall: a hollow-fibre filter makes water microbiologically safe and does nothing about what is dissolved in it. Carbon helps at the margins and is a 500-litre consumable. Boiling concentrates the problem. The conclusion each time was avoid the source.

⭐ Reverse osmosis is the one portable technology that changes that, and the ITEHIL ITS100G-HW is the version of it built for exactly this use. It deserves a proper look, because it is the only thing on offer that engages with the Michigan chemical problem head-on.

🔴 It is also three to six times slower than a gravity filter and costs $299, which means it cannot be the bulk system — and that turns out to point at the right architecture rather than a reason to dismiss it.

7.1 What it is

Table 1 — What it is

Model	ITEHIL ITS100G-HW portable RO system
Membrane	Reverse osmosis, nominal 0.0001 micron
Stages	Hybrid pre-filter + RO membrane
Output	500 ml/min of product water
Power	Dual 6,000 mAh internal batteries, USB-C; ~9 gallons per charge
Operating pressure	110–150 psi, closed system
Consumables	RO membrane 600–800 gal · hybrid pre-filter 300–400 gal
Price	~$299

Reverse osmosis works differently from everything else in this dive. A hollow-fibre filter is a sieve — it stops what is physically larger than its pores and passes everything smaller, which is why dissolved molecules walk straight through. RO is not a sieve. It uses pressure to force water through a semi-permeable membrane that passes water molecules and rejects almost everything else, including dissolved ions and organic molecules.

⭐ That is why it needs 110–150 psi and a pump, and why it is slow. The pressure is not moving water through a hole; it is overcoming osmotic pressure. A gravity filter runs on 1.7 psi. This runs on roughly a hundred times that.

7.2 What it genuinely fixes

🔴 PFAS. This is the real headline. EPA identifies high-pressure membranes — nanofiltration and reverse osmosis — as typically more than 90 percent effective across a wide range of PFAS, and lists RO alongside activated carbon as an effective treatment. Michigan has extensive PFAS surface-water advisories, and nothing else in this dive touches it. Carbon reduces it; RO removes most of it.

Dissolved metals and salts. The Keweenaw mining legacy, nitrates, hardness — all rejected. Again, a category no hollow-fibre filter addresses at all.

Taste. Not a safety matter, but RO water from a tannin-stained Upper Peninsula stream is a different drink from filtered water out of the same stream.

7.3 ⚠ What is genuinely unclear — and it is the cyanotoxin question

Sources conflict on whether RO removes cyanotoxins, and the conflict matters because that is Michigan’s other chemical threat.

The physics argues it should. Microcystin-LR is a molecule of roughly 995 daltons. RO membranes reject dissolved organics well below that mass, and the peer-reviewed literature on nanofiltration and RO for surface-water contaminants generally reports high rejection of compounds in that range.

⚠ But at least one lake-association treatment guide states flatly that an RO unit “does not remove cyanotoxins or microbiologicals such as viruses.” That claim is hard to reconcile with the membrane physics, and it may be describing a specific consumer appliance or conflating RO with a different stage.

🔴 This is marked Verify. The authority to settle it is EPA’s own guidance on water treatment optimisation for cyanotoxins, not a product page and not this one. ⭐ Until it is settled, the operating rule from vol 2 stands unchanged: do not draw from a bloom. RO may well handle microcystin; it is not a licence to test that theory on yourself.

7.4 ⚠ And a caution that applies to all consumer RO

RO units are generally designed for water that is already microbiologically safe — municipal supply — rather than as backcountry purifiers. Membrane integrity is the reason: a pinhole or a failed seal lets raw water bypass the membrane, and there is no indicator when it does. Municipal RO installations are followed by disinfection for exactly this reason.

⭐ The Itehil addresses this with its hybrid pre-filter stage, which is doing the microbiological work ahead of the membrane — and that is also why the pre-filter, not the RO membrane, is the shorter-lived consumable (300–400 gal against 600–800).

⚠ It is worth confirming with the manufacturer what standard the microbiological claim is tested to, in the same way the LifeSaver’s NSF P231/P248 numbers are stated. A “99.99% filtration efficiency” figure on a product page is not a protocol.

7.5 The arithmetic that decides how to use it

At 500 ml/min it is 3.5 to 6 times slower than a gravity filter.

As the bulk system, filtering everything at this dive’s 18–26 gallons a weekend:

Table 2 — As the bulk system, filtering everything at this dive's 18–26 gallons a weekend:

Weekend volume	Time running	Battery charges
18 gal	2.3 hours	2.0
26 gal	3.3 hours	2.9

🔴 That is not viable. Two to three hours of pumping, two or three recharges, and the 300–400 gallon pre-filter would be exhausted in 11 to 15 weekends.

As the drinking and cooking system only — about 6 gallons a weekend for two people:

Table 3 — As the drinking and cooking system only — about 6 gallons a weekend for two people:

Time running	45 minutes
Battery	0.7 of one charge
Hybrid pre-filter life	50–67 weekends
RO membrane life	100–133 weekends

⭐ Comfortable on every axis. Under an hour, one charge, and consumables measured in years.

7.6 ⭐ The architecture this points to

Two tiers, split by what the water is for:

RO for drinking and cooking — a few gallons a weekend, into the containers that get drunk from. This is the water where PFAS and dissolved contaminants actually matter, because it is the water going inside people.

Gravity for everything else — washing, the shower, dishes, dog. Bulk volume at 1.75–3 L/min, cheap consumables, no power.

🔴 This is a two-path system, which vol 3 argued against — and the objection does not apply here. The reason two paths were rejected was that one of them carried untreated water, so a mix-up meant drinking lake water. ⭐ Here both outputs are microbiologically safe. Drinking gravity-filtered water by mistake means you drank what every other volume in this dive recommends. The consequence of the error is what makes the architecture acceptable, not the tidiness of the diagram.

⚠ Still label them. Different containers, different colours, and the RO output in the ones that go to the table.

7.7 Is it worth $299?

It depends entirely on one thing: whether the sources are chemically suspect.

⭐ Buy it if the plan involves repeated trips to the same inland Michigan lakes, or watersheds carrying PFAS advisories, or the Keweenaw’s mining country. In that case it is the only equipment in this dive that addresses the actual risk, and $299 against a 100-weekend membrane is not expensive.

Skip it if sources are mostly flowing streams well away from agriculture, industry and habitation. There, the chemical risk is low and a gravity filter plus a bleach residual is a complete answer — and the $299 buys nothing the LifeStraw Peak does not already do, more slowly.

⚠ And note what it does not remove from the plan. It does not replace the gravity filter, because it cannot do bulk. It does not replace the chlorine residual, because RO water has none. It does not make a bloom safe. And it is another lithium-battery device on a truck that already has a payload problem.

🔴 The freeze rule applies here too. An RO membrane is wet after use and is damaged by freezing like every other membrane in this dive.

7.8 The recommendation

Buy the gravity system first. It closes the actual capability gap — the inability to fill a container — for a quarter of the price, and it does the bulk work RO cannot.

⭐ Then treat the Itehil as the answer to a specific question: “do I trust this water chemically?” If the honest answer at the sources actually used is no, it is the right purchase and nothing else in this dive substitutes for it. If the answer is yes, the money is better spent on the trailer.

References

• ITEHIL ITS100G-HW — 500 ml/min product water; dual 6,000 mAh batteries with USB-C, ~9 gal per charge; 110–150 psi closed system; RO membrane 600–800 gal, hybrid pre-filter 300–400 gal; ~$299.
• US EPA — Reducing PFAS in Drinking Water with Treatment Technologies: high-pressure membranes (nanofiltration, reverse osmosis) typically more than 90% effective across a wide range of PFAS.
• ⚠ Conflicting sources on cyanotoxins — membrane physics and the nanofiltration/RO literature suggest high rejection of microcystin-class molecules; at least one lake-association guide states RO does not remove cyanotoxins. 🔴 Marked Verify against EPA’s “Water Treatment Optimization for Cyanotoxins.”
• Michigan EGLE — PFAS surface-water advisories; HAB programme at Michigan.gov/HABs.
• Water and Hygiene vols 2, 5 and 6, this project — the chemical gap, the parts-built system, and the off-the-shelf comparison.