Chlorine cannot save water that never reaches the filter

When a pool clouds up, instinct runs to chemistry: more chlorine, a flocculant, an algaecide. But chemistry only works on the water that passes in front of it — and it is the pump that decides how much and how often. The number that governs everything is called turnover: the hours needed for the entire volume to pass through the filter once. Good practice asks 8-12 hours for a residential pool, 4-6 for a public one; and the trap lies in how it is computed: with the rated flow the count almost always works out, with the real flow it almost never does. This calculator uses the second — plate times circuit efficiency — and returns the turnover time, the daily turnovers, the index against the requirement and the plate you actually need.

The formula: plate, circuit and requirement

The model is a three-step chain. The real flow Q_eff = Q·e/100 corrects the plate with the circuit's efficiency: the plate holds at free discharge, but pipes, bends, head and above all a dirtying filter typically return 60-85% of it. The turnover time T = V/Q_eff says in how many hours the whole volume passes through the filter. The final comparison is the index δ = T/T_req: dimensionless, equal to itself under any country's rule — because the required turnover is not hard-wired in the model, it is an input to declare with its source. Below 1 the loop closes in the required time; above 1 what is missing is flow or maintenance, not chlorine.

The volume scale: when the pool grows and the pump does not

The table grows the volume with the pump fixed (plate 8 m³/h, efficiency 75%, requirement 8 hours) — the photograph of the most common error, the pump "inherited" from a smaller pool or picked from a price list:

On the last row the full loop asks for more than a day: even continuous running is not enough, and chemistry fights a hydraulic battle alone. The 50 m³ row is the most instructive: it misses the requirement by four tenths of an hour — exactly the default's case, where the plate was enough and the circuit was not.

A complete numerical example

A 50 m³ in-ground pool (10 × 5 × 1 metre average depth), an 8 m³/h plate pump, a circuit in ordinary health at 75%, the 8-hour residential requirement. On paper: 50 / 8 = 6.25 hours, passed with margin. In reality: the real flow is 8 × 0.75 = 6 m³/h, the full turnover asks 50 / 6 = 8.33 hours and the index is 8.33 / 8 = 1.04 — the loop does not close. To hit the 8 hours with these losses you would need an 8.3 m³/h plate: half a pump more, or — more honestly — a circuit that stops stealing 25%. It is the reveal in one line: the pump is bought on the plate, but the pool turns on the real flow.

How to read the turnover index

The four bands describe four systems. Below 0.7 the turnover runs with ease: the margin should be spent not on speed but on slowness — with a variable-speed pump, running slow and long filters better and consumes a fraction. Between 0.7 and 1.0 the requirement is met, with the exact tie (δ = 1.00) belonging to this band: compliant on the edge, where regular backwashing makes the difference. Between 1.0 and 1.3 the loop does not close: the most frequent case, and almost always the circuit's fault more than the pump's. Beyond 1.3 the pump chases: algae and cloudiness are a matter of time, and the solution is hydraulic before it is chemical.

Where the flow goes: the anatomy of the losses

Between the plate and the return jet there is a circuit that presents its bill. The sand filter is the most variable item: clean it absorbs little, dirty it can halve the flow — which is why the filter's pressure gauge is the cheapest diagnostic tool of the whole system. Undersized pipes and every 90-degree bend add head losses; the static head weighs when the plant room sits above or far from the water level; full skimmer and strainer baskets choke the suction. The calculator's efficiency percentage sums all of this into one number: 85-95% for a new, well-sized system, 70-80% for ordinary health, below 60% when the system steals nearly half the pump.

The requirement: where the 8 hours come from

The required turnover is not a law of nature: it is a health threshold that changes with use and jurisdiction. Residential pools run by good practice between 8 and 12 hours; public ones drop to 4-6 for the bather load, with national rules setting binding values; children's pools and spas demand far faster turnovers still. That is why the calculator treats T_req as a declared input: the same pool with the same pump passes as a family pool and fails as a public one — and whoever manages it must know under which number they are working, and cite it.

International perspective: same hydraulics, different thresholds

The physics of turnover is identical everywhere; the conventions are not. Europe reasons in cubic metres and hours with national rules for public pools; North America reasons in gallons and gallons-per-minute, with typical 6-8 hour turnovers for residential pools and state requirements for public ones; professional operators in every market converge on the same operating rule — better a slow but true full loop than a fast plate on paper. The calculator is parametric and dimensionless precisely for this: volume, plate, losses and requirement are explicit inputs, and the index δ — a pure ratio — compares a pool in Rome, one in Miami and one in Sydney on the same scale.

Limits of the model and correct use

The model solves the hydraulics of turnover, not the chemistry nor the detailed hydraulics: it does not compute head losses pipe by pipe (the efficiency sums them into one number to estimate honestly), it does not cover filter sizing — which has its own optimal filtration velocity — nor the suction safety requirements, which follow their own rules. Turnover is the necessary condition of healthy water, not the sufficient one: for disinfection there is the pool chlorine calculator, the chemical twin of this hydraulic count; and for those who still have to dig the pool, the excavation cost calculator opens the project from the right side.

The efficiency scale: the circuit that eats the pump

The same 8 m³/h plate pump, on the same 50 m³ pool, produces five different verdicts depending on how much the circuit leaves it. The table lowers the efficiency — from the freshly commissioned system to the forgotten filter:

Between the first and last row the pump has not changed by one turn: what changed is what surrounds it. At 45% you would need a nearly 14 m³/h plate to do the work a 6.6 does at 95% — a neglected circuit costs more than a new pump, and it is the item no price list shows. From 60% down the model adds the warning: before any purchase, an afternoon of maintenance.

The requirement scale: passed as private, failed as public

The pool's turnover time does not change — 8.33 hours, the column that does not move — but the verdict does, because the threshold it is compared with changes:

The same pool with the same pump crosses all four bands without a drop moving differently: failed twice as a public pool, on the edge as a strict residential one, passed with margin at the tolerant thresholds. It is the practical demonstration of why the requirement is a declared input — and why whoever turns a private pool into one for public use must redo this count before any other.

Running hours: turnover does not ask for the whole day

The turnover says how long one full loop takes; the running hours say how many loops actually happen. A residential pool with an 8-hour requirement and an index below 1 closes its loop in a daytime run: 8-10 pump hours during the sunny hours — when UV consumes the chlorine and bathers load the water — cover a full turnover plus margin. In summer and with a crowded pool the run stretches towards two loops; in mid-season it shortens; in winter, where the pool stays open, the loop can slow further. The operating rule: running hours are scheduled on the needed turnovers, not on habit — and the calculator's index says exactly how many loops your real flow can deliver in those hours.

Variable speed: the margin is spent on slowness

The variable-speed pump has flipped the turnover strategy. Head losses grow with the square of the flow: halving the speed cuts the losses to a quarter and the electrical consumption to an even smaller fraction — while the filter, crossed more slowly, holds fine particles better. That is why an index below 0.7 is not an invitation to shorten the run but to slow it: the same number of turnovers, spread over more hours at reduced speed, produces clearer water and lighter bills. It is the rare case where the ecological choice and the economic one coincide without compromise — provided you have the hydraulic margin the calculator measures.

Typical mistakes in turnover management

The first mistake is sizing on the pump's plate volume: the count is done on the real flow, and the difference is 15-40%. The second is ignoring the filter gauge: rising pressure is falling flow, and it is the only instrument that warns before the cloudiness does. The third is answering cloudiness with chemistry: if the loop does not close, chlorine chases a target it cannot reach. The fourth is switching the pump off "to save" below the minimum turnovers: the electricity saved is paid back in algaecide, flocculant and backwashing hours. The fifth is the circuit grown in pieces — the enlarged pool, the driveway that lengthened the pipes, the relocated plant room — without ever redoing the count: every hydraulic change moves the efficiency, and with it the index.

Turnover in winter and mid-season

The health requirement holds while the pool is in use; off-season the turnover changes job. In winter, in pools that stay full, the pump runs to protect the system — avoiding freezing in the pipes and stagnation — with loops much slower than the summer turnover; in mid-seasons the organic load drops and the turnovers can thin out following the water temperature, which governs how fast algae and bacteria reproduce. The operators' practical rule: below 15 degrees the water forgives, above 25 it demands — and the required turnover of the warm season becomes the number to respect again at the first heat, not after the first green water.

Frequently asked questions

How do you calculate a pool's turnover?

With T = V/Q_eff: the volume divided by the pump's real flow, that is the plate corrected by the circuit's efficiency. With 50 m³, an 8 m³/h plate and 75% efficiency, the full turnover asks 8.33 hours — against the 6.25 the plate alone promised.

What is the turnover index?

It is the ratio δ = T/T_req between the actual turnover time and the one required by rule or good practice. It is dimensionless: 1.04 means the loop misses the requirement by 4%. Below 0.7 there is margin to spend on slowness; above 1.3 the solution is hydraulic, not chemical.

What is the right turnover for a residential pool?

Good practice points to 8-12 hours for residential pools; public ones drop to 4-6 with binding rules, and children's pools and spas demand even faster loops. The calculator treats the requirement as a declared input: the same pool passes as private and fails as public.

Why is the real flow lower than the plate?

Because the plate holds at free discharge, while the real circuit presents its bill: a dirtying filter, pipes and bends, head, full baskets. A new system returns 85-95% of the plate; ordinary health is 70-80%; below 60% the system steals nearly half the pump.

How many turnovers per day are needed?

It depends on the requirement: with an 8-hour turnover you need 3 full loops in 24 running hours; public pools at 4-6 hours need 4-6. The calculator shows the turnovers possible with your real flow: if they are fewer than needed, longer running hours are the first remedy.

Must the pump run 24 hours a day?

No: it must run enough hours to complete the required turnovers. With the index below 1 the loop closes within the planned running time; the margin can be spent cutting hours or — better — slowing a variable-speed pump, which filters better and consumes far less.

Why is the water cloudy even with chlorine in range?

Often because the turnover does not close: chemistry only works on the water that passes through the filter, and if the full loop asks too many hours part of the volume stays out of reach. Before raising the doses, check the turnover index: stubborn cloudiness is almost always a hydraulic symptom.

How do I estimate my circuit's efficiency?

With three clues: the filter's pressure gauge (high pressure = dirty filter = flow down), the age and diameter of the pipes, the distance and elevation of the plant room. In practice: 85-95% for a new well-built system, 70-80% ordinary, 60% or less with a neglected filter or undersized pipes.

What do I do if the loop does not close?

In order of yield: an afternoon of maintenance (backwash, baskets, seals) that often returns more flow than a new motor; then longer running hours; only at the end a bigger pump — using the calculator's "plate needed", which already incorporates your losses.

Does the calculator work in gallons and GPM?

Yes, by converting: 1 m³ = 264 gallons, 1 m³/h ≈ 4.4 GPM. The index δ does not change with the unit, because it is a ratio between times: just enter volume and flow in the same family of units. American turnover thresholds (6-8 hours residential) are used as inputs as they are.

Does a variable-speed pump change the calculation?

It changes the strategy: for the same turnovers, running slower for longer cuts the head losses (which grow with the square of the flow), filters better and consumes a fraction. The calculation stays the same at every speed: real flow at that speed, time, index.

What does this calculation NOT cover?

The chemistry (chlorine, pH, stabiliser — there is the dedicated calculator), filter sizing with its optimal filtration velocity, pipe-by-pipe head losses and the suction safety requirements. Turnover is the necessary condition of healthy water: sufficiency comes from filter and chemistry together.

Updated July 2026