Water Test to Filter Spec: Hardness, Chlorine & LSI | XZH

Short answer: A commercial water filter cannot be specified from the equipment model alone. Six numbers from a water test decide it: total hardness, free and total chlorine, TDS, total alkalinity, pH, and the service temperature. Hardness and chlorine set how long the cartridge actually lasts, alkalinity and pH decide whether the water scales or corrodes, and temperature is the reason a water profile that looks harmless in a cold line destroys an espresso boiler. Every rated capacity on a cartridge label — ours included — is measured at a fixed test condition, so the first job is to derate that number to your water.

We manufacture commercial filter cartridges in Ningbo for importers, foodservice dealers and private-label brands. This guide is the procedure our engineering team runs on every water report a customer sends, written so your team can run it before you ask us for a quote.

---

The Six Numbers That Decide a Cartridge Spec

Most sourcing enquiries arrive as “we need a filter for a Hoshizaki ice machine” or “quote me an Everpure H-300 equivalent”. That fixes the fitting and the flow rate. It does not fix the media volume, the carbon grade, or the replacement interval printed on your box — those come from the water.

Parameter	Typical unit	Cheap test method	What it decides
Total hardness	mg/L as CaCO3, or gpg	Titration kit or strip, $10-30	Scale risk; life of any scale-control or softening cartridge
Calcium hardness	mg/L as CaCO3	Lab panel	The calcium term in the LSI calculation
Free chlorine	mg/L as Cl2	DPD Free tablet/strip	Carbon block life; taste and odour performance
Total chlorine	mg/L as Cl2	DPD Total tablet/strip	Total minus free = chloramine; decides standard vs catalytic carbon
TDS	mg/L (or µS/cm)	TDS meter, $15	Whether carbon is enough or you need RO or UF
Total alkalinity	mg/L as CaCO3	Titration kit	Buffering; the LSI alkalinity term; taste in coffee
pH	—	Meter or strip	LSI; corrosion risk on copper and stainless
Service temperature	°C at the equipment	Equipment datasheet	The single biggest swing factor in scale formation

The two that buyers most often skip are total chlorine and alkalinity, and both of them change the bill of materials. Skipping total chlorine is how a coconut-shell carbon block ends up in a chloraminated city and gets returned for “bad taste”.

Hardness: “10,000 Gallons” Is Not a Capacity Until You Say at What Hardness

Hardness is reported in at least four units depending on who wrote the report, and half the specification errors we see start as a unit mix-up. One grain per gallon equals 17.1 mg/L as CaCO3.

Hardness	mg/L as CaCO3	Grains per gallon	USGS classification
Soft	0-60	0-3.5	Soft
Moderate	61-120	3.6-7.0	Moderately hard
Hard	121-180	7.1-10.5	Hard
Very hard	over 180	over 10.5	Very hard

Classification per the U.S. Geological Survey Water Science School. Conversion: 1 gpg = 17.1 mg/L as CaCO3.

Now apply it. A cartridge label that says “10,000 gallons” is only meaningful next to the condition it was measured at. For an ion-exchange cartridge the arithmetic is exact, because the resin holds a fixed number of hardness grains:

Water hardness	mg/L as CaCO3	Gallons from a 10,000-grain cartridge
3 gpg	51	3,333
5 gpg	86	2,000
7 gpg	120	1,429
10 gpg	171	1,000
15 gpg	256	667
20 gpg	342	500
25 gpg	428	400

A distributor selling the same SKU into Dallas and into Seattle is selling a cartridge with a five-fold difference in service life. If the replacement interval on the carton is the same in both markets, one of those markets is getting warranty claims.

Chlorine vs Chloramine: The Test Almost Every Buyer Skips

Free chlorine and chloramine are both measured “as Cl2” and both show up on a DPD test, but they need different carbon. The test is simple: measure free chlorine and total chlorine. The difference is combined chlorine — chloramine.

Result	Reading	What it means	Cartridge consequence
Free ≈ Total	e.g. free 0.9, total 1.0	Free-chlorine supply	Standard carbon block is fine
Total much higher than free	e.g. free 0.2, total 2.4	Chloraminated supply	Needs catalytic carbon, longer contact time, lower rated flow
Free above 2.0 mg/L	e.g. 3.5	High residual, common at plant-adjacent sites	Derate carbon life; consider larger carbon volume
Free near 0, total near 0	—	Well water or long dead-leg	Chlorine is not the design driver; sediment and hardness are

Two numbers anchor this section, and they are worth writing on the spec sheet:

• NSF/ANSI 42 chlorine reduction is tested at an influent challenge of 2.0 mg/L ± 10%, with a minimum required reduction of 50%. That is the condition behind essentially every “rated capacity” on a taste-and-odour cartridge. (NSF/ANSI 42 particulate reduction is a different claim with its own classes — Class I is challenged with a minimum of 10,000 particles/mL and requires at least 85% reduction.)
• The US EPA maximum residual disinfectant level for chlorine and for chloramines is 4.0 mg/L as Cl2, under 40 CFR 141.65. A utility is allowed to run twice the NSF test condition and still be fully compliant.

So a cartridge tested at 2.0 mg/L, installed on a legal 4.0 mg/L supply, is being asked to absorb twice the chlorine per gallon. Nothing is defective. The label just described a different water than yours.

Sourcing a compatible line and not sure which carbon grade your market needs? Send us the DPD free and total numbers for two or three cities in your territory. We will tell you whether one SKU covers the territory or whether you need a catalytic version for the chloraminated cities — before you commit to tooling or artwork. Talk to our engineering team.

Capacity Derating: The Math We Run Before We Quote a Rated Life

This is the part that separates a factory quote from a catalogue number. Different media are limited by different things, so they derate differently. Treating every cartridge as “gallons” is the most common error in the trade.

Cartridge type	What actually limits life	Derating rule	Confidence
Carbon block, taste & odour	Chlorine mass absorbed	Life ≈ rated × (2.0 ÷ actual free Cl in mg/L)	First-order estimate — good for planning, conservative at high residuals
Catalytic carbon, chloramine	Contact time (EBCT)	Rated flow must be reduced, not just the gallons	Design change, not a derate
Ion exchange / softening cartridge	Hardness grains held	Gallons = grain capacity ÷ gpg	Exact arithmetic
Polyphosphate scale feed	Media dissolution — dose per volume	Roughly volume-limited; hardness decides whether it works at all	Reliable; see scale-control options
PP sediment	Pressure drop rise from solids load	Turbidity-driven; gallons ratings are nearly meaningless	Use ΔP, not a calendar
RO membrane	Fouling and scaling in the concentrate	Sized on concentrate chemistry, not feed	See the recovery section below

Worked example, carbon block rated 6,000 gallons at the NSF challenge condition:

Actual free chlorine	Planning life	vs label
0.5 mg/L	~24,000 gal	4×
1.0 mg/L	~12,000 gal	2×
2.0 mg/L (test condition)	6,000 gal	1×
3.0 mg/L	~4,000 gal	0.67×
4.0 mg/L (EPA MRDL)	~3,000 gal	0.5×

Be honest about what this model is. Carbon dechlorination is a surface-catalysed reaction, so capacity is closer to constant in mass of chlorine than in volume of water — which is why the inverse rule works as a planning estimate. At the high end it is optimistic, because contact time starts to limit the reaction before the carbon is exhausted. We publish it as a derating floor for interval setting, not as a performance guarantee, and we say so on the datasheet.

LSI: Why the Same Water Is Harmless Cold and Destroys an Espresso Boiler

The Langelier Saturation Index is the difference between measured pH and the saturation pH of calcium carbonate:

LSI = pH − pHs, where pHs = (9.3 + A + B) − (C + D), with A = (log₁₀[TDS] − 1) ÷ 10, B = 34.55 − 13.12 × log₁₀(°C + 273.15), C = log₁₀[Ca hardness as CaCO3] − 0.4, D = log₁₀[alkalinity as CaCO3].

Positive LSI means the water tends to deposit calcium carbonate. Negative means it tends to dissolve it and can be aggressive to metal. The term that matters for foodservice is B, the temperature term — it falls from 2.18 at 20 °C to 0.81 at 100 °C, which pushes LSI up by about 1.37 units for the same water simply by heating it.

Three real-world profiles, computed with the formula above:

Water profile	20 °C (cold line)	60 °C (dishwasher)	90 °C (espresso boiler)	100 °C (steam)
A — typical municipal (pH 7.6, TDS 250, Ca 180, Alk 140)	−0.02 balanced	+0.71 scaling	+1.20 heavy scaling	+1.35 heavy scaling
B — soft coastal (pH 7.2, TDS 90, Ca 45, Alk 35)	−1.58 aggressive	−0.85 aggressive	−0.36 mildly aggressive	−0.21 near balance
C — hard Gulf / Middle East (pH 7.9, TDS 620, Ca 340, Alk 260)	+0.79 scaling	+1.51 severe	+2.01 severe	+2.16 severe

Ca and Alk in mg/L as CaCO3. Computed from the standard LSI relationship; values rounded to two decimals.

Read Water A again. It is the water that passes every consumer “is my water hard?” test as merely average — and it is strongly scale-forming the moment it enters a 90 °C group head. This is why a filter that keeps a cold drinking tap perfect gets blamed when the espresso machine scales up in eight months: the cartridge was specified against the cold-water profile.

Two limits of LSI we tell customers about, because most blogs do not:

1. LSI is a tendency, not a rate. An LSI of +2.0 does not mean twice the scale of +1.0. Use it to rank risk and to decide the treatment tier, then set the replacement interval from capacity math.
2. At boiler temperatures the real situation is worse than the index. Heating drives off dissolved CO2, which raises pH beyond the value you measured cold, and calcium carbonate’s solubility falls with temperature. LSI computed from a cold-water sample is therefore the optimistic case for any steam or boiler position — the combi oven and steamer positions are always the first to fail in a kitchen for exactly this reason.

RO Systems Are Sized on the Concentrate, Not the Feed

If the answer is a membrane, the number that matters is not the feed water chemistry — it is the chemistry of what is left behind on the membrane surface. At recovery r, the concentration factor is 1 ÷ (1 − r).

Recovery	Concentration factor	Concentrate Ca (mg/L)	Concentrate alkalinity	Concentrate LSI at 25 °C
Feed (no recovery)	1.0×	180	140	+0.08
50%	2.0×	360	280	+0.65
75%	4.0×	720	560	+1.22
80%	5.0×	900	700	+1.40

Water A feed profile; pH held constant, which is the conservative assumption.

A system running 75% recovery on ordinary municipal water is scaling its membrane at a concentrate LSI that most operators would associate with Middle East groundwater. That is why recovery, softening pre-treatment and flush frequency are specification decisions, not installer preferences — and why the recovery rate belongs on the datasheet of any tankless RO system you put your brand on.

Municipal Report vs Real Lab Test: What the Annual Report Hides

Customers frequently send us a utility’s annual report instead of a test. It is useful — free, and it tells you the disinfectant chemistry — but it has three structural blind spots.

What you get from a CCR	What it hides
Regulated contaminants, detected levels	Hardness is not a regulated contaminant, so many CCRs omit it entirely (US CCR content is set by 40 CFR 141.153)
Annual average or range	Seasonal swings; a summer high can be double a winter low
System-wide figures	Chlorine residual at the end of a long distribution leg can be a fraction of the plant value
Treatment plant chemistry	Anything the building’s own plumbing adds — rust, copper, sediment after a main repair

Practical rule: use the CCR to establish chlorine vs chloramine and the general mineral picture, then spend $30 on a point-of-use panel for hardness, free/total chlorine and TDS at the actual installation. For a multi-site chain, test three sites at the extremes of the distribution network rather than all of them.

From Water Profile to Cartridge Build

This is the matrix our engineers work from. It answers “what goes in the cartridge”, which is a different question from “which scale technology” — for that decision, see the scale control comparison.

Measured profile	Sediment stage	Carbon stage	Scale stage	Membrane
Hardness <7 gpg, free Cl ≤2, TDS <250	5 µm PP	Standard carbon block	Not required	No
Hardness <7 gpg, chloramine present	5 µm PP	Catalytic carbon, derate flow	Not required	No
Hardness 7-15 gpg, any steam equipment	5 µm PP	Carbon block	Polyphosphate feed, fixed 6-month interval	No
Hardness 15-25 gpg	1-5 µm PP	Carbon block	Ion exchange or softener ahead of boilers	Consider for espresso
Hardness >25 gpg or TDS >500	1-5 µm PP	Carbon block as pre-treatment	—	RO, blended back for coffee
Visible sediment, rust, old mains	20 µm + 5 µm staged	Carbon block	Per hardness	Per TDS
Ice production, any hardness	5 µm PP	Carbon block	Scale feed sized to ice machine flow	Only for clear-ice programs

For coffee specifically there is a published target to aim at rather than a guess. The SCAA standard Water for Brewing Specialty Coffee (rev. 21 Nov 2009) specifies total chlorine 0 mg/L, TDS 150 mg/L (acceptable 75-250), calcium hardness 4 grains / 68 mg/L (acceptable 1-5 grains, 17-85 mg/L), total alkalinity at or near 40 mg/L, pH 7.0 (6.5-7.5) and sodium at or near 10 mg/L. Note what that means for RO: strip a hard water to near-zero TDS and you are below the standard, which is why permeate for espresso is blended back rather than used neat. The coffee machine filter line is built around hitting that window, not around maximum removal.

What This Means for Your Warranty Claims

From inside a cartridge factory, the pattern in warranty correspondence is remarkably consistent. Roughly speaking, the complaints that reach us divide into three kinds:

1. “The filter died early.” Almost always a rating-condition mismatch — hard water, high chlorine residual, or a sediment load that raised ΔP long before the carbon was exhausted. No defect exists, but the buyer printed the test-condition interval on the carton and inherited the argument.
2. “The water tastes wrong.” Usually a chloraminated supply with standard carbon, or an RO output used neat where the SCAA window called for blending.
3. “The equipment scaled anyway.” Usually a scale feed past its interval — polyphosphate depletes silently with no visible symptom — or a cold-water spec applied to a boiler position.

All three are prevented at the specification stage by numbers the buyer already could have had for $30. If you are building a private-label programme, the single highest-value thing you can do is print an interval derated for your market rather than copying the branded original, and state the rating condition next to the capacity on your own datasheet. It costs nothing and removes the argument. The same discipline applies whether you are sourcing our compatible line or any other factory’s.

What to Send Us for a Free Spec

Specification work is free and does not require an order. Send:

1. Water test: hardness, free and total chlorine, TDS, alkalinity, pH (a lab panel or clear photos of test strips both work).
2. Equipment list and positions: ice machine, espresso, combi oven, dishwasher, fountain — with model numbers where you have them.
3. Peak flow rate at each position, or the equipment datasheets.
4. Target market and interval you want to print on the carton.

We return a cartridge build per position, a derated interval for that water, and an FOB quotation. Standard terms: MOQ 500 pcs per SKU, 15-20 days after sample approval and deposit, FOB Ningbo, T/T 30% deposit and 70% before shipment. Full sourcing process in our China cartridge sourcing guide, and certification questions are covered in the NSF, FDA and SGS guide.

Frequently Asked Questions

What water test results does a filter manufacturer actually need?

Six numbers cover almost every commercial job: total hardness, calcium hardness where available, free and total chlorine, TDS, total alkalinity and pH — plus the service temperature at the equipment. A $30 lab panel gives you all of them. Send those with the equipment list and peak flow and we can spec and quote without a site visit.

How do I convert grains per gallon to mg/L of hardness?

One grain per gallon equals 17.1 mg/L as CaCO3. So 7 gpg ≈ 120 mg/L, 10.5 gpg ≈ 180 mg/L, 20 gpg ≈ 342 mg/L. US datasheets usually use grains, Chinese and European lab reports usually use mg/L — convert both to one unit before comparing two cartridge ratings.

Why does my filter cartridge die long before its rated gallons?

Because the rating was measured on test water. NSF/ANSI 42 chlorine reduction is challenged at 2.0 mg/L ± 10%; the EPA maximum residual disinfectant level is 4.0 mg/L, so a legal supply can be twice the test condition and roughly halve the throughput. Ion-exchange cartridges are stricter still: gallons = grain capacity ÷ gpg, exactly.

Is a municipal water quality report enough to spec a filter?

It is a starting point. Hardness is not a regulated contaminant, so many annual reports omit the number that most controls cartridge life, and the values are system-wide averages. Use the report to identify chlorine vs chloramine, then test at the actual site for hardness, chlorine and TDS.

What is the Langelier Saturation Index and how do I use it?

LSI is measured pH minus calcium carbonate saturation pH: positive scales, negative is aggressive. Its value here is the temperature term. A typical municipal water (pH 7.6, TDS 250, Ca 180, alkalinity 140) computes to about −0.02 at 20 °C and +1.20 at 90 °C — balanced cold, strongly scaling in an espresso boiler. Use it to rank risk and pick a treatment tier, not to predict a scaling rate.

Do I need a water test if I am only buying compatible replacement cartridges?

You need it if the replacement interval on your carton is going to be true. A drop-in compatible inherits the branded cartridge’s rating conditions, so in a hard or highly chlorinated market you should print a derated interval rather than copying the original’s number. See our materials guide for what changes inside the cartridge.

What is your MOQ and lead time once the spec is settled?

MOQ 500 pcs per SKU, 15-20 days after sample approval and deposit, FOB Ningbo on T/T 30% deposit and 70% before shipment. Specification work is free. A new mould for a non-standard housing or fitting adds tooling cost and time, so confirm the interface before artwork.

Related Guides

• Water Filter Cartridge Materials Explained: PP vs Carbon Block vs RO vs UF
• Scale Control Compared: Polyphosphate vs Softeners vs RO
• Chloramine Removal: Catalytic vs Standard Carbon
• Ice Machine Water Filter Sizing: Flow Rate and Capacity
• Commercial Dishwasher and Glasswasher Water Filtration
• Coffee Machine Water Filter Manufacturers: OEM Sourcing Guide
• Water Filter Replacement Schedule for Restaurants

---

Send us your water report. We will return a cartridge build for every position, an interval derated for that water, and a factory-direct FOB quotation — no order required. Request a spec and quotation, or see what we build for OEM and private-label programmes.

Sources: U.S. Geological Survey, Hardness of Water (hardness classification); 40 CFR 141.65 (maximum residual disinfectant levels); 40 CFR 141.153 (Consumer Confidence Report content); NSF/ANSI 42 certified performance data (chlorine influent challenge 2.0 mg/L ± 10%, ≥50% reduction; Particulate Class I ≥85% at min. 10,000 particles/mL); SCAA Standard, Water for Brewing Specialty Coffee, rev. 21 Nov 2009.