Per capita water consumption is a municipal design figure
Per capita water consumption is the average volume of water one person uses in a day, and it is the first term in the standard municipal chain: a served population multiplied by a per-head consumption gives a water demand; a share of that demand returns to the sewer; infiltration and inflow are added; a peaking factor turns the daily average into a design flow. Published municipal practice works in bands of the order of 100 to 200 litres per person per day for a served population, and those figures should be read as a starting method rather than as a specification, because consumption per head moves with climate, income, metering and the price of water.
The chain is a good piece of engineering. Every term in it can be checked, and for a residential area it produces a flow that is close enough to design on. The difficulty begins when the same chain is applied to a factory, because not one of its assumptions still holds.
What the allowance quietly assumes
- That the water is used by people, at a rate that changes slowly from day to day and from season to season.
- That consumption per head is similar across the area being served, so that an average over the area means something at any point inside it.
- That the share of water drawn that reaches the sewer is stable, the remainder going to irrigation, evaporation, product or loss.
- That the load per person is stable too, so that a flow allowance quietly carries a load allowance with it.
- That the sewer is a sewer: a network that collects everything and delivers it to one place, so that an average and a peak factor between them describe the stream.
Each of those is a fair description of a housing estate. Each of them is false for a factory with a wet process.
An industrial site does not have a population
The unit is wrong before the number is. A factory uses water in proportion to what it makes. The natural units are cubic metres per tonne of product, per head of livestock processed, per thousand containers filled, per square metre of cloth dyed, per barrel of oil refined, per vehicle washed, or per shift worked. None of them converts into persons, and the conversion is not the point: there is no population on the site to count.
Two plants making the same product can differ several-fold in water use per tonne, because one cools with a recirculating loop and the other once through, because one cleans with a hose and the other with a machine, because one reuses rinse water and the other does not. A per-head allowance cannot represent that spread. A production-based unit can, because the differences live in the production process, which is where the unit is defined.
Industrial wastewater flow calculation: find the unit the site has
There are five sources worth building a flow from, and a defensible figure usually comes from two or three of them agreeing. An industrial wastewater flow calculation built on one of them alone is a guess with a decimal point.
- Production-linked units — volume per tonne, per head, per unit produced. Available from the production plan before a single meter is read.
- A water balance — every metered inlet, every metered outlet, and the difference between them. The strongest route where the site has inlet meters and a few weeks of records.
- Outfall metering — a flow meter on the discharge, recorded across a full production cycle and across a shutdown, so that the wash-down days appear in the record rather than in a note.
- Batch and tanker records — loads delivered or collected, which do not appear in a per-head allowance at all and which can be a large share of a small site load.
- Clean-water estimation — the water drawn that is lost to product, evaporation or cooling, subtracted from the meter reading rather than assumed.
Where the five disagree by more than a working tolerance, the disagreement is the finding and it should be chased before the plant is designed, because it usually marks an unmetered connection.
Industrial wastewater treatment plant sizing starts from the load
A biological plant is sized by the mass of material it has to oxidise: kilograms per day, which is flow multiplied by concentration. In industrial wastewater treatment plant sizing, two factories discharging the same hundred cubic metres a day can differ by an order of magnitude in the kilograms of organic load they hand to an aeration tank of the same size, and the second one needs a different plant.
This is where a per-head allowance fails in a way that a flow figure alone does not. The allowance collapses flow and load into a single term, people, and carries a load assumption inside it. An industrial site with an ordinary flow and an extraordinary load is invisible to it.
The arithmetic that follows is simple: measure or estimate the concentration of each parameter of concern, multiply by the flow, and express the result per day. The values come from the site, not from a table. The only judgement is which day to use, and the answer is the day with the heaviest load the plant has to survive, not the day the sampling team happened to be on site.
Design flow for industrial wastewater: the peak comes from a schedule
A design flow for industrial wastewater is a peak as much as an average, and the peak is not a factor looked up from municipal practice. It comes from the way the site works.
- A batch discharge — a tank or a process vessel emptied in twenty minutes, which is a flow rate no daily average describes.
- Clean-in-place and wash-down — a short, warm slug with its own chemistry, arriving at the end of a shift.
- Cooling and boiler blowdown — continuous, warm and low in organic load; a flow and temperature problem rather than a treatment problem, but it still sets the pipe.
- Yard and roof runoff — a rain event can reach the outfall carrying the site behind it, and where the yard is dirty it belongs in the design.
- Tanker deliveries and rejected loads — an entire vehicle emptied into the works in an hour.
- The shift pattern — three shifts finishing at once is a daily peak with a rhythm, and the rhythm is the design case.
A municipal peaking factor is derived from the shape of a domestic day. An industrial peak comes from a schedule, and a schedule is a document to be read rather than a factor to be looked up.
Industrial wastewater characterisation: what to measure first
A flow and a load are the beginning of a design, not the whole of it. Industrial wastewater characterisation is the list below, and every item on it is measured rather than assumed.
- pH and alkalinity, recorded as a range across a shift rather than as one number, because a biological process needs both a tolerable pH and the alkalinity to hold it.
- COD and BOD, with the ratio between them, which says how much of the load the biology can actually reach.
- Suspended solids, and what they consist of, which decides whether they settle, float, or have to be screened out.
- Oil and grease, which decides the pre-treatment stage more often than any other single measurement.
- Total nitrogen and phosphorus, which set the biological duty and any chemical dosing that follows it.
- Heavy metals and specific inhibitors — solvents, biocides, phenols — which decide whether a biological process can be used at all.
- Chloride, sulphate and conductivity, which act on the biology and on the materials the plant is built from.
- The batch pattern, recorded as traces of flow and concentration against the clock rather than as averages.
Sample the worst case as deliberately as the average. The design case is the hour in which the plant is least able to cope, and it is the hour an average never describes.
Where per-capita figures are still the right tool
None of the above makes the per-head allowance a bad number; it makes it an answer to a different question. Domestic sewage on a site, and a residential estate, a campsite or a small community, is exactly the flow the figure was written for. The population to capacity chart is that case, set out as a look-up, and the peaking factor chart is where its daily average becomes a design flow. The mistake is not using the figure; it is using it for the process stream, where the unit it is written in does not exist.
Common mistakes
- Multiplying a headcount by a per-head allowance and calling the product the design flow for a factory with a wet process.
- Sizing on the average day. An average conceals the batch discharge that decides the equalisation volume and the peak that decides the pipe.
- Treating flow and load as one problem. A plant adequate for the flow can be an order of magnitude short of the load.
- Leaving out the water that arrives from outside the process: yard runoff, tanker washing, blowdown, and infiltration into a long site drain.
- Designing for today. Where the production plan contains a second line, the flow and the load in that plan are the ones to size on.
Self-check before the flow is fixed
- Is the flow expressed in a unit the site actually controls — tonnes, head, shifts — rather than in a population the site does not have?
- Is the peak taken from the site schedule, including the batch discharges, rather than from a factor borrowed from municipal practice?
- Has the worst-case day been sampled on purpose, at the point and the time that decides the design?
- Are the domestic flows on the site separated from the process flows, so that each is sized on its own basis?
- Does the design include the production plan and the growth in it, or only the plant as it stands today?
From an allowance to a design
A design flow is a decision about what the plant has to survive, and it is made from the records the site keeps rather than from a table. The plant selection and sizing guide works through what follows from a flow, a load and a site, and the food and beverage processing page is the same reasoning carried through a wet-process industry. The population to capacity chart remains the right starting point for domestic and residential flow. The flow, the load and the peak a plant has to be built around are confirmed with the factory for your process rather than read from a general figure.