Still & Mineral Water Machine Selection Guide — Capacity, Bottle Size and Valve Type

1. Three axes decide a water machine — and they interact

Most still water enquiries arrive as a single number: bottles per hour. That number on its own cannot be quoted. Two plants asking for the same 12,000 BPH can need machines that share almost no parts, because output is only one of three axes that define a water filling machine.

1. Capacity band — the sustained hourly output at a named reference bottle, not the peak figure on a datasheet.
2. Bottle format — small format 0.2–2 L and large format 3–16 L are two different machine families, not two settings of one machine.
3. Filling principle — gravity, mechanical, semi-electric, volumetric or load cell, each with a different accuracy, cost and changeover behaviour.

The three axes are not independent in practice. Format constrains output far more than output constrains format: a carousel rated at 24,000 BPH on 0.5 L will not produce 24,000 5 L bottles per hour, and not because it is slow — the mould, the clamping unit and the bottle gripper are different hardware. Similarly, the valve principle limits how fast you can change between formats, which decides whether a multi-size plant runs profitably or spends its week on changeovers.

This guide works through each axis, then crosses them into one selection table, then resolves the two remaining questions every water project has to answer: combiblock or separate machines, and what water treatment sits upstream. Configuration and price tiers are compared separately in still and mineral water machine configurations compared.

2. Axis 1 — capacity band

Nameplate output is quoted against a reference bottle, normally 0.5 L still water at a defined fill height with the downstream line running without stoppage. Two numbers matter more than the nameplate: the output on your largest bottle, and the output the line holds over a full shift including changeovers, cap jams and downstream stops. Ask for both, in writing, before comparing quotations.

2.1 Where the valve and cavity counts come from

Valve count follows from throughput per valve. A still water filling valve delivers roughly 600–900 bottles per hour at 0.5 L, falling as fill volume rises because the liquid simply takes longer to enter the bottle. Blowing cavity count follows the same logic: 1,500–2,000 bottles per hour per cavity at 0.5 L, dropping to around 900–1,300 at 1.5 L as cooling time increases. Both figures are indicative and are confirmed once preform and bottle drawings are reviewed — they are given here so you can sanity-check a quotation, not so you can specify from them.

2.2 Sizing against the market rather than the ambition

The most expensive sizing error in bottled water is buying two bands too high. An over-sized line idles, and idle capacity is not free: it carries the same installed power, the same footprint, a larger spare part inventory and higher unit prices on wear parts that you consume slowly. Size for the volume you can sell within roughly two years, and confirm the machine family can be extended — adding valves or cavities inside an existing carousel is rarely possible, but adding a second line to a layout planned for it is straightforward.

3. Axis 2 — bottle format

The single largest fork in a water project is small format against large format. These are not two configurations of one machine; they are two machine families with different physics.

Two consequences follow. First, a plant that intends to sell both single-serve and home-and-office formats is buying two lines, and should plan the building for both from day one rather than discovering later that the large-format line has nowhere to go. Second, the large-format line is where filling accuracy starts to earn its cost: a 1% overfill on 0.5 L is 5 ml, but on 15 L it is 150 ml given away on every container.

Large-format machines are covered in detail in combiblock for 5 L and large-format bottles; small-format block configuration is in combiblock for still and mineral water.

4. Axis 3 — filling principle

Water is the product family where the cheapest valve is most often the correct one. The question is not which valve is best in the abstract, but which error the valve has to eliminate for your product and your market.

4.1 What each principle is actually blind to

• Gravity and mechanical are blind to bottle volume scatter. Preform weight variation and blowing variation change the internal volume, so a constant level is not a constant quantity.
• Semi-electric improves the repeatability of the stop signal but does not measure quantity, so it inherits the same bottle sensitivity in a reduced form.
• Volumetric is blind to density. For water at stable temperature this is irrelevant; it becomes relevant only if you later run the same line on juice or syrup.
• Load cell measures the delivered quantity directly and is blind to nothing, which is why it costs the most and why it is hard to justify below 5 L on commodity margins.

The full cross-product comparison, including how to calculate whether accuracy pays for itself, is in load cell vs volumetric vs mechanical filling valve.

5. The three axes crossed

Reading the three axes together gives one table. Find your output band on the left and your bottle family across the top; the cell is the configuration that normally applies.

The blank cells are not a catalogue gap. Filling a 10 L bottle takes tens of seconds, so 24,000 of them per hour would require a carousel with several hundred valves. The economic answer at that volume is multiple large-format lines, which is how high-volume home-and-office producers are configured.

6. Combiblock or separate machines

Once capacity, format and valve are fixed, one architectural decision remains: a blow-fill-cap block, or a separate blow moulder feeding a rinser-filler-capper across an air conveyor.

6.1 Where the dividing line actually falls

Two variables move the line, and they move it in the same direction. The higher the output, the more resin saving and operator saving the block returns per year. The more standard the bottle, the fewer format parts and moulds it needs, so integration costs less to exercise. A plant at 24,000 BPH running two standard formats is a clear block case. A plant at 6,000 BPH running six shaped bottles for different customers is a clear separate-machine case. Between those, run the arithmetic on your own resin price and payroll rather than accepting a general claim.

The detailed trade-off, including the resilience argument for separate machines, is in combiblock vs separate blowing and filling line.

7. Water treatment upstream — and why mineral water is different

The filling machine is the visible half of a water plant. The treatment train upstream decides product taste, regulatory compliance and how long your valve seats and rinser nozzles last. It is almost always a separate scope, priced from a laboratory analysis of your source water, and it is the item most often missing from a first-time budget.

7.1 Purified water and natural mineral water are regulated differently

Purified or prepared drinking water may be treated by any method that produces safe, consistent water. Natural mineral water, in most regulatory systems, must reach the bottle with the mineral composition it had at source. That single rule removes several treatment options and reshapes the plant.

7.2 Ozone and UV in practice

Ozone is dosed into a contact tank and carried as a residual into the filler so the bottle and headspace are disinfected after capping; the residual then decays over the following hours. Typical practice is a low residual at the filling valve with several minutes of contact time upstream. Two engineering consequences follow. Ozonated water attacks ordinary elastomers, so seals, gaskets and O-rings in the ozone contact path must be specified for it, and stainless steel grade matters more than on a plain water line. And if the source water contains bromide, ozonation can form bromate, which is regulated in most markets — this must be checked against the water analysis before the dose is set, not after the first laboratory result comes back.

UV at 254 nm is the other common final barrier. It leaves no residual, so it disinfects the water passing through the chamber but does nothing for the bottle headspace. Most water plants run both: UV as a continuous barrier and ozone where a residual is wanted at the point of fill.

8. Decision matrix and the mistakes that repeat

Working the axes in order, these conditions settle most water projects:

1. If your bottles are 3 L or larger, then you are in the large-format family: 2,000–12,000 BPH, volumetric or load cell filling, and a separate line from any small-format production.
2. If you run 0.2–2 L below roughly 8,000 BPH, then separate blowing plus a rinser-filler-capper is normally the better buy, with gravity or semi-electric filling.
3. If you run 0.2–2 L at 12,000 BPH or above with two or three stable formats, then a combiblock returns its premium through bottle weight, floor area and operator count.
4. If you change format more than twice a week, then choose volumetric filling regardless of accuracy needs — changeover time alone justifies it.
5. If you sell empty bottles or blow for a third party, then do not buy a block; the block fills what it blows.
6. If your product is natural mineral water, then remove RO, softening and remineralisation from the treatment scope and build the barrier around filtration plus permitted disinfection.
7. If your source water is a borehole with variable quality, then budget for a treatment train sized on the worst analysis of the year, not the average.
8. If the plant will run one shift for the first year, then size one band lower and plan the layout for a second line rather than buying idle capacity now.

8.1 Five mistakes that repeat on water projects

• Quoting one number. Asking six suppliers for "12,000 BPH" without stating bottle size, fill volume and format mix produces six quotations that cannot be compared.
• Buying accuracy the product does not need. Commodity water sold by nominal volume rarely repays load cell filling. The money is better spent on treatment or on air drying.
• Leaving treatment, CIP and compressed air out of the comparison. These three items move a water plant budget substantially, and each is easy to omit from a filler quotation.
• Economising on compressed air quality. Blowing air must be dry and oil-free at the machine tie-in point. Wet air produces bottle wall defects that then get blamed on the blower.
• Ignoring the format roadmap. The sizes you plan to add in year three decide the mould and format part budget now. Declaring them at quotation stage costs nothing; adding them later costs a changeover programme.

Delivered water lines in this range include C002 Taza, running four 12,000 BPH combiblock lines on 1.5 L bottles in Kazakhstan, and C005 Rauan, with eight bottling and canning lines between 12,000 and 24,000 BPH. Configuration tiers and their cost drivers are compared in water machine configurations compared, and the shorter operational questions are answered in the still and mineral water FAQ.

Frequently asked questions