Catchment Water Filters and UV Sterilizers, Explained
A catchment treatment train is three different tools doing three different jobs in a fixed order: sediment cartridges strain out particles, carbon takes out taste, odor, and many chemicals, and ultraviolet light inactivates the organisms the filters let through. None of the three substitutes for another, and the specifications that decide whether each stage is doing its job are printed in smaller type than the marketing: a micron rating with the word "nominal" in front of it, a UV flow rate quoted at a dose too low for untreated water, a certification class meant for city water. This page reads the small type, stage by stage, for the system a catchment home actually runs.
Affiliate disclosure: Power & Pantry is reader-supported, and some links on this site are affiliate links that may earn us a commission at no extra cost to you. This article links to the Hawaii Department of Health, to the University of Hawaii CTAHR catchment guidelines, to the text of the federal UV dose table, to Pentair's published cartridge literature, and to our own catchment guide and maintenance schedule. See the About page for how we work.
Why the order is not negotiable
Each stage exists partly to protect the next one. Sediment comes first because particles are what clog carbon, coat UV sleeves, and shadow organisms from ultraviolet light: an organism hidden behind a particle is an organism the light never reaches, which is why CTAHR and every UV manual treat prefiltration as part of the disinfection step rather than an optional extra. Carbon sits in the middle, after the sediment that would plug it and before the lamp it would otherwise starve of clear water. UV goes last for a reason that shapes the whole system: ultraviolet light leaves nothing behind in the water. Chlorine keeps working in the tank and the pipes; UV works only in the chamber, in the seconds the water passes the lamp. Water disinfected by UV and then stored can be recontaminated in storage, so the unit sits at the point of entry, after the tank, treating water on its way to the taps. The complete catchment guide walks the whole system part by part; this page stays on the three treatment stages and what their specifications actually promise.
Stage one: sediment, and what a micron rating really says
A micron is a millionth of a meter, and a cartridge's micron rating names the particle size the rating refers to. What it does not name, unless you read the maker's footnote, is how much of that size gets through. Pentair, whose Pentek cartridges are the standard example in this market, prints the definition on its own spec sheets: its micron ratings are "based on 85% or greater removal of given particle size." A rating like that is a nominal rating. It describes a cartridge that stops most particles at the stated size, which is exactly what a sediment stage is for, and it is not a wall. An absolute rating is the other kind: it names the largest particle that passes, with retention at that size up around 99.9% or better. The same physical cartridge can carry both kinds of number, a fine nominal figure and a coarser absolute one, and the two numbers answer different questions. For sediment duty, nominal is fine and normal. For a barrier you are counting on to stop a specific organism, the word in front of the number is the specification.
CTAHR's staging for catchment runs coarse to fine, typically a 20-micron cartridge ahead of a 5-micron one. The coarse cartridge takes the load of grit, roof sediment, and organic debris so the fine one does not plug in a week; the fine one polishes the water to the clarity the UV stage needs. Cartridges come as spun or wound depth media, which trap particles through the body of the material and are the cheap consumables of the train, and as pleated surface media, which offer more area in the same housing. Which construction matters less than the rhythm: the sediment cartridge is changed about once a month on a typical Hawaii system, and falling pressure at the taps is the standing signal to check it first. The full rhythm, with prices, is on the maintenance schedule page. What matters here is the division of labor: the sediment stage's job is particles and protection, and no sediment cartridge, at any micron number, is a disinfection step.
The fine stage: cysts, and the claim that actually counts
The organisms that make a fine stage necessary are the protozoan cysts, and they are large as microbes go. Per the CDC, Cryptosporidium oocysts run about 4 to 6 microns across and Giardia cysts about 8 to 12. Both shrug off chlorine at the doses a catchment tank carries, which is why the defense against them is physical: a barrier they cannot fit through, or UV, covered below. Size alone suggests a 1-micron filter should stop a 4-micron oocyst, and a true absolute 1-micron cartridge does. A nominal 1-micron cartridge, rated at 85% removal, is a different promise. This is why CTAHR's wording for the cyst stage is specific: an absolute 1-micron cartridge certified for cyst removal. The CDC's guidance lands in the same place, naming absolute 1-micron filters and filters certified to NSF/ANSI 53 for cyst reduction as the greatest assurance against Cryptosporidium.
The NSF/ANSI 53 cyst claim has a number behind it: a system carrying it reduces the concentration of parasitic cysts by at least 99.95%, per the standard's own text. Reading a real cartridge against that bar is instructive. Pentair's Pentek CBC-10, the carbon block the maintenance page prices, is rated at a nominal 0.5 microns and is capable of reducing 99.95% of Cryptosporidium and Giardia cysts in the manufacturer's testing, run on the NSF Standard 53 protocol with particle counts. Its NSF certification, though, is to Standard 42 for material requirements only. Both halves of that sentence matter. The cyst performance was measured against the standard's method; the cartridge is not itself certified to Standard 53 for cyst reduction. A buyer who wants the certified claim looks for the certification listing, not the protocol reference, and a buyer who wants a cyst barrier at all looks for one of the three wordings: absolute 1-micron, an NSF/ANSI 53 cyst-reduction certification, or a maker's cyst figure with its test basis stated, in that order of strength. What does not belong on the list is a plain carbon or sediment cartridge whose only credential is a small nominal micron number.
Stage two: carbon, the chemical stage that is not disinfection
Carbon's job is adsorption: chlorine, the tastes and odors of tank water, and many organic chemicals bind to its surface as water passes. It comes as loose granules (GAC), which flow freely and handle taste and odor, and as a bonded carbon block, which packs the carbon into a porous solid that also strains fine particles and, in rated models like the CBC above, carries the cyst performance. CTAHR puts the carbon change at every 6 to 12 months, and its tell is taste and odor returning rather than pressure falling.
Two limits define the stage. The first is printed by the makers themselves. Pentair's cartridge literature carries the instruction not to use its cartridges with water that is microbiologically unsafe or of unknown quality without adequate disinfection before or after the unit. Carbon is not a barrier against bacteria or viruses, and a carbon stage is never the reason catchment water is safe to drink; it is the reason treated water tastes like water. The second limit is placement. Carbon removes chlorine, which is its virtue in a glass and its complication in a system: a household that maintains a chlorine residual in its tank, as CTAHR describes, loses that residual at the carbon stage. That is usually the intent, chlorine's work is done by then, but it means nothing downstream of the carbon is protected by the tank's chlorine, which is one more reason the UV unit, not the carbon, closes the train.
Placeholder: matched offers
When this site's affiliate programs are approved, this block will show the stages this page explains: sediment and carbon cartridges, absolute-rated cyst cartridges, and whole-house UV units with their Class A flow ratings and lamp part numbers, with dated prices.
No affiliate links are live on this page yet, and nothing here is a product endorsement.
Stage three: ultraviolet, where dose is the whole story
A UV unit removes nothing. Water flows past a lamp emitting light at 254 nanometers, and the light disrupts the DNA of organisms in the stream so they cannot reproduce. What the water receives is a dose: the light's intensity multiplied by the seconds of exposure, measured in millijoules per square centimeter (mJ/cm²). Everything about a UV unit's real performance is in that number and the conditions attached to it, because dose falls as flow rises, falls as the lamp ages, and falls as the water gets less clear. Makers quote flow rates at a stated dose, at a stated water clarity (UVT, the share of UV light the water transmits), and at the end of the lamp's rated life. Change any of the three and the same unit is a different size.
The standard that sorts UV units is NSF/ANSI 55, and it has two classes that are not two grades of the same job. Class A systems deliver at least 40 mJ/cm² and are built to disinfect water that may be contaminated: wells, surface water, and, by the same logic, catchment. They must carry a UV sensor, an alarm, and a flow restrictor so the water cannot outrun the dose. Class B systems deliver 16 mJ/cm² and are designed for supplemental bactericidal treatment of water that has already been disinfected or deemed acceptable by the health agency with jurisdiction. Class B is not a failed Class A; it is a certification for a different starting point, one a catchment tank does not provide. Roof water is untreated water. A UV unit serving as a catchment system's disinfection barrier is doing Class A's job, whatever its certification says.
| NSF/ANSI 55 Class A | NSF/ANSI 55 Class B | |
|---|---|---|
| Minimum UV dose | 40 mJ/cm² | 16 mJ/cm² |
| Intended water | Water that may be contaminated: wells, surface water, catchment | Water already disinfected or deemed acceptable by the health agency |
| Role | Disinfection barrier | Supplemental treatment |
| UV sensor and alarm | Required | Not required |
| Flow restrictor | Required | Required |
What a dose buys, organism by organism, is in the federal government's own table. The EPA's UV dose table for drinking water credit (40 CFR 141.720) was written for public water systems, not home units, and it is printed here for scale, because it shows where UV is strong and where it is not:
| UV dose (mJ/cm²) | Cryptosporidium | Giardia | Viruses |
|---|---|---|---|
| 12 | 3-log (99.9%) | 3-log (99.9%) | No credit |
| 22 | 4-log (99.99%) | 4-log (99.99%) | No credit |
| 40 (Class A floor) | Past 4-log | Past 4-log | About 0.5-log by this table |
| 186 | 4-log (99.99%) |
Read it plainly: the cysts that survive chlorine are the organisms UV handles most easily, inactivated many times over at Class A dose, and bacteria are comparably sensitive. Viruses are the hard case. The EPA's virus column is set by adenovirus, an unusually UV-resistant virus, and by that municipal yardstick even 40 mJ/cm² earns little virus credit. NSF's Class A certification rests on the standard's own test protocol and carries bacteria, virus, and cyst claims for home units; the EPA table is the stricter, utility-scale accounting. The honest summary for a catchment owner is that a Class A unit is a strong barrier against bacteria and cysts, the organisms catchment guidance worries about most, and that no honest page should call any UV unit a guarantee against every virus. Homes that chlorinate as well as running UV, a combination CTAHR describes as common in Hawaii, hold a second barrier whose strength is precisely where UV's is thinnest.
The worked example: one unit, three sizes
Viqua's VH410 is the whole-house unit a Hawaii catchment owner is most likely to meet, so its manual is the honest test of how UV sizing is sold. The manual rates the same unit three ways at 95% UVT, at the end of lamp life: 34 gallons a minute at 16 mJ/cm², 18 gallons a minute at Viqua's standard 30 mJ/cm², and 14 gallons a minute at the 40 mJ/cm² NSF/EPA dose. A listing that calls the VH410 a 34-gallon unit is quoting the Class B dose. Sized the way this page argues untreated water should be sized, at 40 mJ/cm², it is a 14-gallon unit, still ample for a house, since household demand peaks in bursts well below that. And the certification line in the same manual reads: tested and certified by NSF International against NSF/ANSI 55 for Disinfection Performance, Class B. The unit is well made and its maker publishes better data than most. It is also, on paper, certified for the supplemental job, which is the gap between the UV market as sold and the standard as written that a catchment buyer has to close with their own reading. The manual's dose chart carries its own caution, that it is not performance tested or certified by NSF.
The same manual sets the water the unit expects, and it reads like a description of why the filter stages come first: UV transmittance above 75%, hardness under 7 grains per gallon, iron under 0.3 mg/L, tannins under 0.1 mg/L. Tannins are the catchment-specific entry on that list. Leaf-stained roof water, the tea-colored water that follows heavy rain through overhanging vegetation, absorbs ultraviolet light, and water that absorbs the light lowers the dose at any flow. Sediment and carbon stages, kept on their schedule, are what hold the water inside the lamp's assumptions. A UV unit fed cloudy or stained water is not failing; it is being asked to disinfect water its rating never covered.
One cost of the UV stage is electric and belongs in the open. A whole-house lamp runs continuously, not on demand: the VH410 draws 60 watts around the clock. That is 0.06 kW × 24 hours × 365 days = about 526 kWh a year, which at the EIA's August 2026 average prices is roughly $95 a year at the US average of 18.16 cents per kWh and roughly $248 a year at Hawaii's 47.18 cents. The lamp itself is replaced annually on hours, at the prices the maintenance schedule lists. None of this argues against UV; it argues for counting the disinfection stage as a utility, with a running cost, rather than as a one-time purchase.
What the train does not do
Three honest boundaries close the page. First, rat lungworm. Hawaii DOH's prevention guidance rests on exclusion: covered and screened tanks, slugs, snails, and rats kept away from the roof and tank, produce washed. CTAHR is explicit that the parasite's larvae can pass pores smaller than was once recommended, so no cartridge on this page is presented as a rat lungworm barrier, and any seller who rates one that way is ahead of the guidance. Second, metals. Lead and copper in catchment water come from roof and plumbing materials, not from anything a cartridge train is sized to solve, which is why DOH's answer is the annual subsidized test described on the maintenance page rather than another filter stage. Third, chemicals and clarity run in opposite directions through the train: carbon handles many chemicals and UV handles none, UV handles organisms and removes nothing at all. A stage claimed to do another stage's job is a stage to read twice.
Questions people actually ask
What size micron filter do I need for catchment water?
For sediment, CTAHR's staging: about 20 microns ahead of about 5, coarse protecting fine. For the cyst barrier, the number is not the point by itself: an absolute 1-micron cartridge or one carrying a cyst-reduction claim, because a nominal rating, per Pentair's own footnote, promises 85% or greater removal at the stated size.
Is a UV sterilizer enough to make catchment water safe on its own?
Not as the manuals and the standard frame it. UV disinfects only water clear enough for the light to reach the organisms, at a flow the dose covers, which is why prefiltration is part of the step and why the unit belongs after the filters. It also leaves no residual, so it protects nothing downstream of the chamber. A Class A unit behind maintained filters is a disinfection barrier; a lamp behind clogged cartridges is a light bulb.
What is the difference between Class A and Class B UV?
Dose and intended water. Class A delivers at least 40 mJ/cm², carries a sensor and alarm, and is certified to disinfect water that may be contaminated. Class B delivers 16 mJ/cm² and is certified as supplemental treatment for water already deemed acceptable. Catchment water is untreated, so the barrier job is Class A's, and a unit's flow rating should be read at the 40 mJ/cm² dose.
Does UV remove sediment, chemicals, or lead?
None of them. UV inactivates organisms and removes nothing from the water; sediment is the cartridges' job, many chemicals are carbon's, and lead and copper are handled by materials choices and testing, not by any stage in this train.
Why does my UV unit's flow rating change between listings?
Because the rating is a dose at a flow, and sellers quote different doses. The VH410 in Viqua's manual is 34 gallons a minute at 16 mJ/cm², 18 at 30, and 14 at 40. Ask which dose any quoted flow is rated at, and at what water clarity; the manual's figures assume 95% UVT at the end of the lamp's life.
Can I skip the carbon stage?
The train still disinfects without it; carbon is not a safety stage. What you give up is taste, odor, and chemical reduction, and, in a chlorinating home, the removal of the chlorine residual before the tap. What you cannot skip is the sediment stage ahead of the lamp, because the UV rating itself assumes clear water.