A dosing station is the stage that introduces a set amount of reagent into the flow: antiscalant ahead of membranes, alkali for pH correction, coagulant and flocculant ahead of flotation or a clarifier, flocculant ahead of sludge dewatering. It removes nothing itself: the next stage does the removing, and dosing is what makes that removal possible.
Principle
A dosing station consists of a tank for the reagent solution, a metering pump, a suction line with a foot filter and a non-return valve, a discharge line and an injection assembly — an injection valve cut into the treated pipework. On some versions the tank carries a stirrer: suspensions and solutions prone to separating give an uneven strength within hours if they are not stirred. The catalogue offers two tank volumes, 100 and 200 l, and the connection — dn25, dn32 or dn50 — is chosen to match the pipe diameter at the injection point.
The dose is set by calculation and measured in milligrammes of reagent per litre of treated water. Dose and flow give the hourly mass: 5 mg/l at 10 m³/h is 50 g of reagent per hour. The pump feed follows from dividing that mass by the strength of the solution: with a ten per cent solution, that is 100 g of reagent per litre, 0.5 l/h is required. It follows that the pump feed is not a property of the duty but a quantity derived from three others, and one and the same station covers a different range of doses at a different solution strength.
The dosing mode decides how the feed follows the flow. A constant feed suits a steady flow. Proportional dosing is driven by pulses from a water meter: on each pulse the pump makes a calculated number of strokes, and the dose stays constant at any flow. The third option is feedback dosing, where the feed is set by the signal of a pH or redox electrode; it is used where it is the composition of the water that varies rather than the flow.
The injection point and the mixing length belong to the calculation just as the pump does. The injection valve is fitted so that the reagent enters the core of the flow rather than running down the wall, and downstream of it either a length of pipework or a static mixer is provided, sufficient for mixing to finish before the next unit. Injecting immediately before the flange of a unit means part of the flow passes untreated, while the analysis is blamed on the wrong reagent.
Where there is more than one reagent, the order of injection is set by chemistry, not by the convenience of the pipework. pH correction comes before the coagulant, since coagulants work within a narrow band. The coagulant is injected under vigorous mixing: it needs rapid contact with the whole volume. The flocculant follows it and is mixed slowly — the floc that forms is fragile, and hard mixing breaks it faster than it can grow.
Principle
A metering pump delivers volume, not mass. It measures out litres of solution, whereas the duty is stated in milligrammes of substance per litre of water, and what links the two is the strength of the solution — a quantity that does not stay constant. It depends on how the solution was made up, on the room temperature and on the reagent itself: sodium hypochlorite decomposes in storage, and a few weeks later the same pump at the same setting delivers noticeably less active substance. Hence a rule that contradicts the expectation of «set it once»: what is monitored is the result — the water parameter after the stage — and not the setting on the pump.
Overdosing is often worse than underdosing, and that holds for three reagents for three different reasons. Excess coagulant reverses the charge on the particle surface: from negative it becomes positive, the particles repel one another again, the solids are restabilised, and the water leaving is more turbid than it was at a smaller dose. Excess alkali pushes the pH outside the band set for drinking water, that is, creates a new breach in place of the one removed. Excess antiscalant achieves nothing beyond the cost.
Antiscalant does not bind hardness salts and does not take them out of the water. It adsorbs onto crystal nuclei and stops them growing, which is why it works at doses many times smaller than the amount that binding the calcium would need. Two consequences follow, and they are what people ask about most often: doubling the dose does not double the margin of safety, and the feed water loses no hardness at all — the salts leave with the concentrate rather than disappearing.
Practice
One stage covers five different duties, and what they share is not the reagent but the requirement: to introduce a set amount of a substance into the flow accurately and repeatably. Antiscalant ahead of reverse osmosis, alkali ahead of reagent-free filtration for the sake of manganese, coagulant and flocculant ahead of flotation or a lamella clarifier, flocculant ahead of sludge dewatering, pretreatment reagents ahead of ultrafiltration — the assembly is the same one; the reagent, the dose, the injection point and the materials differ.
Dosing does not improve the neighbouring stage so much as, in many cases, make it possible at all. A flotation unit without a matched coagulant passes the effluent through almost unchanged; catalytic media at a pH of about 7 removes manganese only in part; sludge without flocculant gives up no water during dewatering. In those schemes the dosing station is a condition of operation, not an option.
Chemical protection of membranes and softening solve one problem by different means, and the choice between them is settled by the flow. At large flows antiscalant is cheaper to run: there is no salt handling and no regeneration discharge. At small flows the ratio reverses, and on water going to consumers softening also solves the problem of scale in domestic appliances, which antiscalant dosing does not address at all.
A reagent is a piece of housekeeping, not a consumable in a box. It calls for room for the tank and a bund, a heated space, since aqueous solutions freeze, pipework materials compatible with the substance, personal protective equipment, a procedure for making up the solution and someone who follows it; dry flocculants must in addition be wetted correctly and the solution left to mature. On sites where none of that exists, a reagent scheme is not chosen, and that is a sound design decision rather than a compromise.
The station is assembled from standard parts — tank, digitally controlled metering pump, injection assembly. We do not manufacture it: our part of the work is the water analysis, the calculation of the dose, the choice of dosing mode, the selection of tank volume and injection point, supply, commissioning, adjustment against the analysis and subsequent service.
Limits
Practice
The cases are typical examples, not site reports.
Innovations
Scheme
A dosing station is never a stage in its own right. Its place in the scheme is decided not by itself but by the stage the reagent is introduced for: ahead of a membrane unit, ahead of a filter, ahead of a flotation unit, ahead of sludge dewatering. Everything else follows from that, down to the connection diameter.
Upstream of the dosing point sits whatever evens out the flow and the composition: on effluent that is screens and balancing, on natural water a storage tank. The reagent regime is set for a range of flow and composition, and going outside that range means losing effectiveness at exactly the moment of peak load, that is, when it is needed most.
Downstream come the mixing length and the stage that takes the result. The length, or a static mixer, is calculated: mixing must finish before the inlet of the unit. The position of the electrode is chosen on that same length where dosing runs on feedback.
The reagent line is treated as part of the scheme, not as an auxiliary connection. A non-return valve at the injection point and protection against siphoning are essential: where the pressure at the injection point is lower than the head of solution in the tank, the contents of the tank run into the pipework by gravity and reach the water in one slug. The materials of tank, hoses, pump head and seals are chosen for the particular substance, not for «reagents in general».
Scheme
Operation
Energy
The electricity used by the assembly is not a cost item: the metering pump is the only consumer here, it works in pulses, and its rating is small beside any pumping equipment in the scheme. The significant item is the reagent itself, and its consumption is arithmetic — the dose multiplied by the volume of water treated. At 5 mg/l and 10 m³/h that is 50 g of reagent per hour, or 1.2 kg over a day of continuous running, and it is that quantity, rather than the price of the station, that sets the annual cost of ownership. A third item is not expressed in kilowatt-hours at all: the assembly needs a heated room, because aqueous solutions freeze, and a frozen solution stops the stage outright.
Consumables
What goes to drain
A dosing station produces almost no waste of its own: the reagent leaves with the water to the next stage and ends up either in the sludge or in the treated water. Two consequences follow, and both are settled at the scheme stage. First: everything introduced beyond what is needed stays in the water, and on drinking water the doses are capped by a limit rather than by economy alone. Second: the mass of reagent passes into the sludge of the neighbouring stage and increases its volume — coagulant produces a hydroxide sludge of its own, and in dewatering calculations it is added to the mass of retained solids. Its own waste amounts to the rinsing water after washing the tank and the empty reagent containers; how those are handled is set by the rules for the particular substance, not by a general rule for the assembly.
Sizing
What we account for
What a wrong choice costs
Sizing
Questions
It depends on the duty. Reagent-free filtration, mechanical treatment and ion exchange introduce nothing into the flow, and a scheme on natural water is often assembled with no dosing at all. But a flotation unit without coagulant does not work on most parameters, mechanical dewatering without flocculant gives no reduction in sludge volume, and manganese at a pH of about 7 is removed only in part. In those cases the choice is not between «with reagent» and «without», but between a reagent scheme and a different technology entirely.
Because a coagulant works by charge, not by quantity. It removes the negative charge on colloidal particles and they stop repelling one another. If more reagent goes in than neutralisation requires, the charge changes sign: the particles are once again alike in charge, once again repel, and the suspension is restabilised. The dose in such a case is reduced rather than raised, and the optimum is found by a jar test rather than by trial on a running unit.
Because it keeps the dose constant rather than the feed. A pump running on a timer introduces the same amount of reagent per hour regardless of how much water passed in that hour. On a variable flow that means overdosing during quiet hours and underdosing at the peak: over a day the reagent consumption will match the calculation, but the design dose will not be held even once. Proportional dosing is driven by water meter pulses and the feed follows the flow by itself. A timer is justified where the flow is steady.
That is worked out before purchase. The interval is the tank volume divided by the pump feed: at 0.5 l/h a 100 l tank is used up in roughly 200 hours, that is, about eight days of continuous running, and a 200 l tank lasts twice as long. The feed itself depends on the dose, the water flow and the strength of the solution, so the same station gives a different interval from site to site. The tank volume is chosen so that refilling falls within the site's service schedule and needs no separate visit.
For protecting membranes, yes, and at large flows that is the usual solution: no salt handling and no regeneration discharge. For water going to consumers, no. Antiscalant does not take calcium and magnesium out, it only stops them precipitating inside the unit; the hardness of the feed water is unchanged and the salts leave with the concentrate. If scale in domestic appliances or in a boiler has to be dealt with, that calls for ion exchange or desalination, not dosing.
No, for two independent reasons. Chemical: reagents are rarely compatible in one solution — acid and alkali neutralise each other, and coagulant with flocculant forms floc in the tank itself. Process: their injection points differ and so do their mixing conditions — the coagulant needs vigorous mixing, the flocculant immediately after it needs slow mixing. Each reagent therefore gets its own tank, its own pump and its own injection point, and the only thing they share is the control system.
Tell us about your site — our engineer will prepare a process design, a specification and a quotation within one business day.