Guide

How to run a jar test for ferric alum, PAC and polyelectrolyte

Short answer

Dose six identical jars of fresh raw water with different amounts of a coagulant such as ferric alum or PAC, stir fast, add the polyelectrolyte at the same point as in your plant, then stir slowly, settle and check turbidity, colour, pH and alkalinity. IS 3025 (Part 50) uses 1 minute of fast stirring, 15 minutes of slow stirring and 15 minutes of settling, though other sources say to copy your plant's mixing. Find the lowest dose that meets your target (1 mL of a 1% stock in 1 L of sample gives 10 mg/L), run the plant near the middle or lower half of the dose range that works rather than at its edge, and convert it with kg/day = mg/L × MLD, keeping the dose on the same basis (product, metal or Al2O3) as the chemical you buy.

Updated 28 September 2026 · Rise Chemical India, Kirti Nagar Industrial Area, Delhi

What is a jar test, and what are its limits?

A jar test runs coagulation, flocculation and settling on identical samples at different doses. India's method is IS 3025 (Part 50):2023, which draws heavily on ASTM D2035. It finds the optimum coagulant dose, coagulant-aid dose and coagulation pH for removing turbidity and colour. ASTM D2035 also uses the test to compare coagulants and coagulant aids on the same water under the same conditions, so a plant weighing ferric alum against PAC can run both side by side on its own raw water.

Source: Bureau of Indian Standards, 2023; ASTM International, 2019

Turbidity alone does not give the dose; the MoHUA/CPHEEO water treatment manual (2024) says the exact quantity can only be found by jar tests. The test has limits. A lab jar mixes better than a plant, so MoHUA says to scale up with care, and jar settling curves apply only to the plant tested (CPHEEO, 2005). The testing manual of South Africa's Water Research Commission (WRC) says a coagulant is only truly judged at plant scale.

Source: Ministry of Housing and Urban Affairs / CPHEEO, Government of India, 2024; CPHEEO, Ministry of Urban Development, Government of India, 2005; Water Research Commission, South Africa (Freese, Trollip & Nozaic, Umgeni Water), 2003

What equipment do you need for a jar test?

A multiple-paddle stirrer with speed variable from 20 to 150 rpm (IS 3025, Part 50).
Identical jars. IS 3025 (Part 50) and WRC both use 800 mL of sample in tall-form 1 L beakers. The IWA handbook prefers circular 2 L jars and the US EPA prefers square ones, so pick one shape and keep every jar the same.
10 litres of fresh sample, a measuring cylinder, a stopwatch, and tests for turbidity, pH, colour and alkalinity.

Source: Bureau of Indian Standards, 2001; Bureau of Indian Standards, 2023; Water Research Commission, South Africa (Freese, Trollip & Nozaic, Umgeni Water), 2003; IWA Publishing (open access via OAPEN), 2022; U.S. Environmental Protection Agency, 2004

How do you make stock solutions for the coagulant and polyelectrolyte?

The jar test formula is: dose in the jar (mg/L) = stock strength (mg/L) × stock added (mL) ÷ sample volume (mL). A 1% solution holds about 10,000 mg/L, so 1 mL of it in 1 L of sample gives 10,000 × 1 ÷ 1,000 = 10 mg/L (IWA handbook; California State Water Resources Control Board). In the 800 mL sample of IS 3025 (Part 50), the same 1 mL gives 10,000 × 1 ÷ 800 = 12.5 mg/L, and in the US EPA's 2 L jar it gives 5 mg/L.

Source: IWA Publishing (open access via OAPEN), 2022; California State Water Resources Control Board, Division of Drinking Water, 2020; Bureau of Indian Standards, 2001; U.S. Environmental Protection Agency, 2004

No source here gives a recipe for ferric alum or PAC, so the general coagulant rules apply. IS 3025 (Part 50) gives 1% stocks for aluminium sulphate (remade after 1 month) and ferric sulphate (remade after 1 week, or sooner if it turns opalescent), but has no row for ferric alum. The IWA handbook recommends a 1% solution for hydrolysing coagulants, made fresh before each test from the chemical the plant doses. For a ferric alum plant, that means a fresh 1% solution of the ferric alum in the plant's own store; if you test PAC beside it, make the PAC stock the same way. The California guide dilutes most coagulant products to 0.1–1.0% for 1 L jars and, apart from alum and ferric coagulants, counts the product as 100% strength, so doses are stated as product.

Source: IWA Publishing (open access via OAPEN), 2022; California State Water Resources Control Board, Division of Drinking Water, 2020; Bureau of Indian Standards, 2001

Write the basis next to every dose. The IWA handbook notes that a coagulant dose in mg/L can refer to the hydrated or anhydrous salt, or to the metal (Al or Fe), and each basis gives a different number. For ferric alum, record whether a dose is as product or as metal. For PAC, record product or Al2O3; the liquid vs powder PAC guide covers that conversion.

Source: IWA Publishing (open access via OAPEN), 2022

Anionic polyelectrolyte ('poly') is an anionic polyacrylamide flocculant. Add the powder to the water, never water to the powder (IS 3025, Part 50). WRC TT 405/09 says to disperse granules quickly to avoid 'fish eyes' (lumps of undissolved polymer), then stir gently for up to two hours; Hubbard-Hall, a distributor, warns that blenders shear the polymer. Strength and ageing differ by source:

Source: Bureau of Indian Standards, 2001; Water Research Commission, South Africa (Leopold & Freese), 2009; Hubbard-Hall Inc. (chemical distributor), 2025

IS 3025 (Part 50): 0.5% stock, at least 1 day old, remade after 1 week, diluted 10 times just before use.
WRC K5/1184, plant practice: about 0.2%, aged up to 24 hours.
WRC TT 405/09, plant practice: around 0.5%, stirred gently for up to two hours.
US EPA CCP handbook, plant feed: 2% by weight or less.
Hubbard-Hall, vendor lab method for emulsion polymer: 0.1%, aged 30 minutes.

Source: Bureau of Indian Standards, 2001; Bureau of Indian Standards, 2023; Water Research Commission, South Africa (Freese, Trollip & Nozaic, Umgeni Water), 2003; Water Research Commission, South Africa (Leopold & Freese), 2009; U.S. Environmental Protection Agency, 2004; Hubbard-Hall Inc. (chemical distributor), 2025

The IS stock diluted 10 times is 0.05% (500 mg/L), so 1 mL in an 800 mL sample gives 500 × 1 ÷ 800 = 0.625 mg/L. A 0.1 mg/L floc-aid dose (WRC gives doses as low as 0.05–0.1 mg/L) would need 0.1 ÷ 0.625 = 0.16 mL. MoHUA says to make stock solutions at strengths that let the volumes be measured accurately and with ease, so dilute further for doses this small.

Source: Bureau of Indian Standards, 2001; Water Research Commission, South Africa (Freese, Trollip & Nozaic, Umgeni Water), 2003; Ministry of Housing and Urban Affairs / CPHEEO, Government of India, 2024

What is the jar test procedure, step by step?

MoHUA's O&M manual describes six jars at different doses and says to follow IS 3025 (Part 50), whose values are used here.

Source: Ministry of Housing and Urban Affairs / CPHEEO, Government of India, 2024

Take 10 litres of raw water at the plant inlet (IS 3025), before any chemical or returned filter washwater has mixed in (WRC). Test it soon, because biological activity in standing water can change how it coagulates (IS 3025).
Record raw turbidity, colour, pH, alkalinity and temperature, and test at the plant inlet temperature (IWA handbook).
Measure 800 mL into each 1,000 mL beaker, leaving at least 5 mm between paddle and wall.
Set a dose range. For an unfamiliar water, WRC starts alum at 10–60 mg/L and ferric chloride at 5–30 mg/L, and more turbid water may need more; it gives no starting range for ferric alum. If you are also testing PAC, MoHUA reports powdered PAC doses are generally about one third of alum doses in mg/L, a field observation rather than a rule. IS 3025 says to repeat the test with different concentrations, so narrow the range around the jars that work.
If the plant corrects pH before the coagulant, add that chemical to the jars first, in the same order (MoHUA, Part B). Then stir fast for at least 1 minute, add the coagulant (ferric alum, alum or PAC) and start the stopwatch.
1 minute after adding the coagulant, switch to slow stirring and note when pin-point floc appears in each jar.
Add the polyelectrolyte in the same order and after the same delay as in your plant, timed from when the coagulant went in. If your plant doses it within the first minute (MoHUA's weir mixer: about 10 s), add it before the switch to slow stirring in step 6.
After 15 minutes of slow stirring, switch off and let the jars stand for 15 minutes.
Decant 150–200 mL of the clear upper layer (supernatant). If floc floats, siphon from 30–40 mm below the surface.
Measure colour, turbidity, alkalinity and pH within 3 hours.

IS 3025 (Part 50) calls these times illustrative and open to change for site conditions. A note in the same standard gives about 200 rpm for a 30-second flash mix, faster than the 20–150 rpm stirrer it specifies.

Source: Bureau of Indian Standards, 2001; Bureau of Indian Standards, 2023

What mixing speeds and settling times should you use?

Most sources say to copy your plant's mixing and timings. The US EPA CCP handbook sets the jar rapid-mix and flocculation times equal to the plant's detention times, and matches paddle speed to the mixing intensity of the plant's flocculators, measured as the velocity gradient G (in s⁻¹). The IWA handbook and the WRC testing manual also tie mixing to the plant, and a 1995 EPA paper by Lytle says no standard jar test protocol exists.

Source: U.S. Environmental Protection Agency, 2004; IWA Publishing (open access via OAPEN), 2022; Water Research Commission, South Africa (Freese, Trollip & Nozaic, Umgeni Water), 2003; U.S. Environmental Protection Agency (Science Inventory), 1995

Jar test mixing and settling settings by source

Source	Rapid mix	Slow mix	Settling	Jars
IS 3025 (Part 50), procedure	Fast, 1 min	15 min	15 min	800 mL in 1,000 mL beakers
IS 3025 (Part 50), note	About 200 rpm, 30 s	20 rpm, 20 min	Not covered by the note	As above
WRC K5/1184 (2003)	100–300 rpm (plant-like), 2 min	40 rpm, 15 min	Not fixed; settling rate recorded	800 mL in 1 L beakers
California SWRCB (2020)	200 rpm, 20–30 s	20–30 rpm, 5 min	Settled turbidity read after 25 min	1 L round beakers
IWA handbook (2022), if plant G is unknown	G 300–500 s⁻¹ for 30–60 s, then G 100–200 s⁻¹ (plant time)	G 30–60 s⁻¹ (Al) or 50–80 s⁻¹ (Fe), plant time	60 min, or another chosen time	1–2 L circular, 2 L optimal
US EPA CCP handbook	Plant rapid-mix time	Plant flocculation time, rpm matched to plant G	From the basin's surface overflow rate	2 L square
SNF, polymer maker (vendor)	100–125 rpm, 15–60 s (up to 5 min for some coagulants)	40 rpm, twice the rapid-mix time	Typically 10–15 min	Square

Source: Bureau of Indian Standards, 2001; Bureau of Indian Standards, 2023; Water Research Commission, South Africa (Freese, Trollip & Nozaic, Umgeni Water), 2003; California State Water Resources Control Board, Division of Drinking Water, 2020; IWA Publishing (open access via OAPEN), 2022; U.S. Environmental Protection Agency, 2004; SNF (polymer manufacturer), hosted by Michigan Water Environment Association, 2015

Settling time differs most, from 10–15 minutes (SNF) to 60 minutes (IWA handbook), and the EPA handbook says an arbitrary 30–60 minute settle has no relationship to a full-scale plant and should not be done. If you do not know your plant's mixing and settling times, use the IS 3025 (Part 50) times, which MoHUA's O&M manual points to, and keep them the same in every run so that only the dose changes (MoHUA, Part B).

Source: U.S. Environmental Protection Agency, 2004; SNF (polymer manufacturer), hosted by Michigan Water Environment Association, 2015; IWA Publishing (open access via OAPEN), 2022; Bureau of Indian Standards, 2001; Ministry of Housing and Urban Affairs / CPHEEO, Government of India, 2024; CPHEEO, Ministry of Urban Development, Government of India, 2005

When should you add polyelectrolyte after the coagulant?

The usual order is pH correction, then the metal coagulant, then the flocculant aid (MoHUA, Part B). Anionic polyacrylamide cannot form floc on its own. WRC TT 405/09 pairs it with a primary coagulant (alum, ferric or a polymeric coagulant) and adds it after that coagulant, in a low-energy zone, to build larger floc that settles faster.

Source: Ministry of Housing and Urban Affairs / CPHEEO, Government of India, 2024; Water Research Commission, South Africa (Leopold & Freese), 2009

Copy the plant's sequence and delays. If polymer enters a pipeline about 30 seconds downstream of the flash mix, the EPA handbook says to use the same sequence in the jars. MoHUA's design manual doses it about 10 s after the coagulant in a weir-type mixer, as turbulence dies down, and the IWA handbook suggests the end of rapid mixing. Excess turbulence shears polymer chains and makes them less effective (EPA). MoHUA adds that other orders, including polymer first, are sometimes more effective, and jar tests should decide.

Source: U.S. Environmental Protection Agency, 2004; Ministry of Housing and Urban Affairs / CPHEEO, Government of India, 2024; IWA Publishing (open access via OAPEN), 2022

Poly doses in drinking-water treatment are small. WRC gives floc-aid doses as low as 0.05–0.1 mg/L and warns that excess polymer can quickly block rapid gravity filters. The EPA handbook says flocculants are generally fed below 1 mg/L, and WHO notes the maximum authorised polymer dose in drinking water is generally 1 mg/L. None of these limits is written for ETP or STP effluent, so find that range by jar test.

Source: Water Research Commission, South Africa (Freese, Trollip & Nozaic, Umgeni Water), 2003; U.S. Environmental Protection Agency, 2004; World Health Organization, 2011

How do you pick the right dose from the jars?

Score each jar on settled turbidity, colour, pH and alkalinity (IS 3025, Part 50:2023), residual aluminium or iron (IWA handbook) and the time until floc first appears (CPHEEO). CPHEEO's guide to examining alum floc in the plant says tiny floc may mean too little coagulant, 'popcorn flake' floc is what you want, and a milky or bluish tint means the dose is probably too high. Sources define the right dose differently:

Source: Bureau of Indian Standards, 2023; IWA Publishing (open access via OAPEN), 2022; CPHEEO, Ministry of Urban Development, Government of India, 2005

IS 3025 (Part 50):2023 and MoHUA: the lowest dose that gives satisfactory settling.
WRC testing manual: the lowest dose that brings turbidity, colour or TOC below its specified maximum.
IWA handbook: the most removal with the lowest residual aluminium or iron. Then run near the middle or in the lower half of that optimum range, not at its edge, where a small change in raw water can stop coagulation.

Source: Bureau of Indian Standards, 2023; Ministry of Housing and Urban Affairs / CPHEEO, Government of India, 2024; Water Research Commission, South Africa (Freese, Trollip & Nozaic, Umgeni Water), 2003; IWA Publishing (open access via OAPEN), 2022

In jar tests on 37 source waters, California's State Water Resources Control Board found that the jar with the lowest settled turbidity did not always give the lowest filtrate turbidity. Overdosing and underdosing can both reduce solids removal (CPHEEO). Coagulated water at pH 5.4 or below is corrosive and needs a lime dose (IS 3025, Part 50:2023). CPHEEO lists weaker colour removal as a disadvantage of PAC, so if you run PAC beside ferric alum, compare colour in every jar; the guide to ferric alum, alum and PAC compares them on pH, alkalinity and sludge.

Source: California State Water Resources Control Board, Division of Drinking Water, 2020; CPHEEO, Ministry of Urban Development, Government of India, 2005; Bureau of Indian Standards, 2023

If ferric alum and PAC both meet the target, compare cost, not dose, with both doses stated as product. 1 mg/L is 1 kg per megalitre (WRC TT 405/09), which is 1 g per cubic metre, so cost per cubic metre = product dose (mg/L) × price per kg ÷ 1,000. Do the same sum for any lime or other pH-correction chemical a jar needed and add it to that jar's cost.

Source: Water Research Commission, South Africa (Leopold & Freese), 2009; Bureau of Indian Standards, 2023; Ministry of Housing and Urban Affairs / CPHEEO, Government of India, 2024

How do you convert the jar dose to kg per day?

A milligram is a millionth of a kilogram and a litre is a millionth of a megalitre, so 1 mg/L is 1 kg per megalitre (WRC TT 405/09). That gives kg/day = dose (mg/L) × flow (MLD), and 20 mg/L is 20 g per 1,000 litres, not 20 kg.

Source: Water Research Commission, South Africa (Leopold & Freese), 2009

State the dose the way you buy the chemical: as product or as active chemical. A stock made from the product as supplied gives a product dose. A dose stated as active chemical must be divided by the product's strength: WRC turns 15 mg/L as FeCl3 from a 43% solution into 15 ÷ 0.43 = 34.9 mg/L of solution.

Source: Water Research Commission, South Africa (Leopold & Freese), 2009; California State Water Resources Control Board, Division of Drinking Water, 2020

Worked example. WRC's 20 MLD plant at 7 mg/L uses 7 × 20 = 140 kg of chemical a day. That is the product to order when the 7 mg/L is a product dose, as it is when the jar stock was made from ferric alum or PAC as supplied. If the 7 mg/L had been recorded as Al2O3 and the plant used a liquid PAC at 10% Al2O3, the product dose would be 7 ÷ 0.10 = 70 mg/L, or 70 × 20 = 1,400 kg a day. Use the strength on your lot's certificate of analysis.

Source: Water Research Commission, South Africa (Leopold & Freese), 2009

Treat the result as a starting dose: the WRC testing manual recommends plant trials of at least three weeks, preferably six, for each coagulant before judging it. The EPA handbook also says alum and PAC are not compatible: a change from feeding alum to PAC needs a complete cleaning of the storage tanks and feed equipment.

Source: Water Research Commission, South Africa (Freese, Trollip & Nozaic, Umgeni Water), 2003; U.S. Environmental Protection Agency, 2004

What mistakes make jar test results misleading?

Old samples, or samples taken after chemicals or washwater have mixed in.
Stock made from a lab-shelf chemical instead of fresh solution of the plant's own product when the test is meant to match the plant (IWA handbook). IS 3025 (Part 50) itself calls for pure chemicals, so record which you used.
Polymer stock made by adding water to powder, or used before it has aged.
Doses logged without their basis (product, metal or Al2O3).
Comparing ferric alum and PAC on different samples or under different mixing, instead of on the same water under the same conditions.
Changing more than one variable per run.
Judging on settled turbidity alone.

Source: Bureau of Indian Standards, 2001; Water Research Commission, South Africa (Freese, Trollip & Nozaic, Umgeni Water), 2003; IWA Publishing (open access via OAPEN), 2022; CPHEEO, Ministry of Urban Development, Government of India, 2005; California State Water Resources Control Board, Division of Drinking Water, 2020; ASTM International, 2019

Which coagulants and polyelectrolytes does Rise Chemical India make or supply?

Rise Chemical India makes ferric alum, its main product, at its own factory in Bharatpur, Rajasthan. Ferric alum is sold loose by weight, as solid slabs of about 20 kg, with a 5 tonne minimum order. The same factory makes anionic polyelectrolyte, as powder or granules in 25 kg HDPE bags with a 250 kg minimum order, and cationic polyelectrolyte; for cationic grades, packing and minimum order, ask the office.

The company also supplies poly aluminium chloride, which it buys in and does not make. PAC powder is 28–32% Al2O3 at 70–85% basicity, in 25 kg or 50 kg bags, with a 250 kg minimum order; liquid PAC is 10% Al2O3, in 35–50 kg carboys, 200–250 kg drums or tankers. The office in Kirti Nagar, Delhi handles every enquiry, quote and order, and the company holds ISO 9001:2015. It sells in India and exports, and the price depends on quantity and destination. For a quote, use the contact page.

Questions buyers ask

What is the formula for a jar test?

Dose in the jar (mg/L) = stock strength (mg/L) × stock added (mL) ÷ sample volume (mL). A 1% stock holds about 10,000 mg/L, so 1 mL added to 1 L of sample gives 10 mg/L, and in the 800 mL sample used by IS 3025 (Part 50) the same 1 mL gives 12.5 mg/L. To turn the chosen jar dose into plant use, kg/day = dose in mg/L × flow in MLD, with the dose on the same basis (product or active) as the chemical you buy.

Why is the jar test used?

A jar test finds the lowest coagulant and polyelectrolyte dose that treats your own water satisfactorily. Raw-water turbidity only roughly indicates the dose, and India's MoHUA/CPHEEO water treatment manual says the exact quantity can only be found by jar tests. The test also compares coagulants such as ferric alum and PAC, coagulation pH and the order of chemical addition on the same water under the same conditions, so the plant dose comes from a measurement instead of a guess.

How often should a jar test be done?

IS 3025 (Part 50) says a treatment unit should repeat it as often as changes in wastewater composition require, and the MoHUA operations manual asks for daily jar tests while raw-water quality is changing significantly. The CPHEEO O&M manual also calls for one at plant start-up, using fresh chemical from the plant's own feeders.

How much ferric alum or other coagulant is needed per 1,000 litres of water?

There is no fixed figure; the dose of ferric alum, alum or PAC comes from a jar test on your own water. Once you have it in mg/L the conversion is simple: 1 mg/L is 1 g per 1,000 litres, so a 20 mg/L dose is 20 g of coagulant per 1,000 litres, not 20 kg. Record the basis of that dose (as product, as metal or, for PAC, as Al2O3), because the same number on a different basis is a different amount of product.

Can you compare ferric alum and PAC in the same jar test?

Yes. ASTM D2035, the practice the 2023 revision of IS 3025 (Part 50) draws on, uses the jar test to compare coagulants on the same water under the same conditions. Make fresh 1% stocks of both products as supplied, run them on one raw-water sample with the same mixing and settling times, and log every dose as product. Compare settled turbidity, colour, pH, alkalinity and residual aluminium or iron, then the cost per cubic metre at the dose each one needs, including any lime or other pH-correction chemical its jars needed.

How long should polyelectrolyte solution age before a jar test?

IS 3025 (Part 50), India's jar test standard, uses a 0.5% polyacrylamide stock that is at least 1 day old, remade after 1 week and diluted 10 times just before use. Other sources differ: South Africa's Water Research Commission describes plant solutions aged up to 24 hours, and Hubbard-Hall, a chemical distributor, ages a 0.1% emulsion-polymer stock for 30 minutes. Follow the polymer maker's instructions and make the jar stock the way your plant makes its dosing solution.

Will the plant need the same dose as the jar test?

Not always. India's MoHUA manual warns that a plant cannot match the mixing in a laboratory jar, and the testing manual of South Africa's Water Research Commission says a coagulant is only truly judged at plant scale and recommends plant trials of at least three weeks, preferably six, for each coagulant. Start near the middle or lower half of the jar test's optimum range and adjust from the plant's settled and filtered water results.

Products in this guide

Ask for a quote: send the product, quantity and your current dose

We make ferric alum, sodium hypochlorite and anionic and cationic polyelectrolyte at our factory in Bharatpur, Rajasthan. Our office in Kirti Nagar, Delhi usually replies within an hour.

Ask on WhatsApp Request a quote

Sources

Sources were checked on 28 September 2026.