Aqutron Insights · Insight ·
The pool is almost empty. Why are the pumps running as if it were full?
Aqutron relates what happens in the pool to the work of the plant. In one pool that adopted the system: −63% electricity for filtration, −30% chemicals and −27% make-up water.
A plant has to be sized for moments of peak use. But maximum capacity is a possibility, not a permanent condition.
A pool's day is made of peaks, pauses, quiet periods and hours without bathers. Yet fixed-speed operation can stay identical even when the pool is used at only a small fraction of its capacity.
The question is concrete: if attendance falls, water parameters are within limits and plant conditions allow it, how much of the pumps' work is still necessary?
That question is where Aqutron's intelligent management starts: adapting treatment to the pool's real and expected needs while continuing to meet quality and circulation requirements.
From running on a timetable to knowing the pool
In one pool that adopted the Aqutron system, comparing the average monthly consumption for the 12 months before and the 3 months after the intervention showed a 63% reduction in filtration electricity, 30% in monitored chemicals and 27% in make-up water.
Managing the attractions brought a further result: 39% less electricity on the attractions circuit.
Applying those two electricity reductions to the daily baselines, at EUR 0.35/kWh, the benefit is EUR 40.13 per equivalent day. Projected over 365 days under the same conditions it is worth roughly EUR 14,648 a year, before any value is even placed on water and chemicals.
The plant was already there. The data were not connected.
Before the intervention, both filtration pumps ran from 07:00 to 22:00. Overnight one switched off and the other kept running. There were no inverters: we installed them.
Filtration had a stated total power of 8 kW and a reference consumption of 156 kWh per day. The attractions added 42 kWh per day, with a total power of 6 kW.
The water analyser was already in place. Aqutron connected to it, building on the existing investment.
Alongside the water measurements we added bather counting, zone usage, visual clarity analysis, historical data and environmental information. The inverters made pump speed adjustable; Aqutron tied that capability to what is actually happening in the pool.
A plant becomes intelligent when it observes, relates the data and adapts how it runs. Knowing only that it is 9 pm is not enough: what counts is how many people are present, how heavily the pool has been used, how the water is responding and what demand is expected in the hours ahead.
The cameras also watch the water
Aqutron cameras do more than count bathers. The system also produces a visual turbidity estimate through AI analysis of reference marks on the pool floor.
The principle is to observe how distinguishable those reference marks are in the images and how their legibility changes over time. Their visibility gives information on water transparency along the path the camera observes.
This sits alongside the analyser readings and pool occupancy: the question “what values is the analyser reading?” is joined by “how does the water look in the observed areas?”.
Chemistry, vision and occupancy become complementary information about the same pool. The visual estimate is not a direct NTU measurement and does not attest to microbiological quality; lighting, reflections and occlusions can affect it. Its value is to add an observation, not to replace other checks.
Aqutron Chronos: anticipating, not just correcting
Aqutron Chronos is the forecasting layer integrated with Aqutron Vision. Its job is to turn observed trends into an estimate of the needs that are about to arise.
How it works can be summed up in four steps:
- Observe over time. Water measurements and pool-usage data are read as sequences: it is not only the current value that counts, but how it was reached. The platform sets the visual observation of the water alongside these data.
- Recognise recurrences. Attendance, timetables, plant response and environmental conditions help recognise situations that have occurred before and useful relationships in the historical data.
- Estimate what comes next. The system produces a short-term forecast of treatment demand, instead of waiting only for a parameter to drift from its reference.
- Update the forecast. As new measurements arrive, the estimate is revised. Control keeps checking itself against what is actually happening in the pool.
One example of the principle: if occupancy rises and the history points to a subsequent increase in disinfectant demand, the forecast lets the control system prepare rather than merely chase a chlorine drop that has already happened.
The reverse holds too. A pool emptying towards closing time may call for different management from the same pool just before a busy period. The overnight window becomes part of the planning: not because the water recovers on its own, but because treatment can be spread over time when plant conditions and requirements allow it.
The facility's data also showed an association between outdoor temperature, humidity and the estimated treatment demand per bather. That is not the same as measuring an individual's microbiological load: it is additional information to refine the forecast.
Forecasting does not mean dosing blind. The forecast feeds the control logic together with real measurements; flow rates, treatment limits and plant protections remain constraints. The model replaces neither the analyser nor the operator's checks.
Filtration: use the power when it is needed
Inverters make pump speed adjustable. Aqutron relates that capability to water conditions and pool usage, within the operating limits defined for the plant.
Ten per cent occupancy does not automatically mean ten per cent filtration. The aim is to identify the available modulation margin, taking into account the accumulated load, the required turnover and the operation of the treatment equipment. Technical guidance on variable speed likewise provides for conditional reductions during quiet or overnight periods.
For centrifugal pumps, lowering the speed can cut the power demand more sharply than the speed reduction itself. The actual result, however, depends on hydraulics, filters and efficiencies: the saving stated here comes from measured electricity consumption, not from a theoretical curve.
In the observed comparison, the average monthly filtration consumption fell by 63%. The result reflects the combination of inverters, control and data use: we do not attribute the whole reduction to the forecasting model alone.
Chemistry follows the bather load
The free-chlorine setpoint indicated for the case is 0.8 ppm, distinct from the applicable lower reference value of 0.7 ppm in the pool water. These values refer to this installation and its treatment, not to a universal setting for every pool.
The control logic aims to limit dips, late corrections and overdosing. The configured reference does not replace verification: actual measurements, calibration and checks at points around the pool remain decisive in establishing water quality.
Monitored chemicals fell by 30%. The comparison covers the intervention as a whole: dosing, pool usage, make-up water and backwashes. It is not a percentage derived solely from the difference between two chlorine setpoints.
Water: count the bathers, know the filters
Counting bathers helps coordinate water replacement. To schedule backwashes, the differential pressure across the filters is read as well, interpreted under comparable flow conditions: pump speed affects pressure too.
Water-quality requirements and maintenance prescriptions continue to take priority. The goal is to avoid superfluous operations, not to postpone necessary ones.
The make-up meter recorded 27% less water added. Limiting avoidable discharge also makes it possible to retain water that has already been treated and, where heated, the heat it holds. These economic effects are not included in the electricity figures reported.
Attractions that follow the bathers
Dividing the pool into zones makes it possible to define custom rules for fountains, hydromassage features and other attractions. Activation and availability can follow the presence of people and the service modes the facility chooses.
In the case described, higher overall use of the attractions and more rotation among users were observed: two indicators of a more dynamic use of the spaces.
The reduction indicated for the attractions' electricity consumption is 39%, separate from that of filtration.
What the avoided energy is worth
| Circuit | Before: kWh/day | After: equivalent kWh/day | Avoided value: EUR/day |
|---|---|---|---|
| Filtration | 156.00 | 57.72 | 34.40 |
| Attractions | 42.00 | 25.62 | 5.73 |
| Total | 198.00 | 83.34 | 40.13 |
Post-intervention values and amounts calculated by applying the 63% and 39% reductions to the daily baselines, at EUR 0.35/kWh. Amounts are rounded only at the end.
Taking the two circuits together, 114.66 kWh a day are avoided, a weighted reduction of about 58%. The daily value of electricity goes from EUR 69.30 to EUR 29.17.
Over 30 equivalent days: EUR 1,203.93. Over 365 equivalent days: EUR 14,647.82.
Over those 365 equivalent days that is 41,850.90 kWh avoided: EUR 12,555.27 for filtration and EUR 2,092.55 for the attractions.
Everything visible, from poolside to the front desk
SPACE AIR completes the picture by monitoring temperature, humidity, CO2, particulate matter, and VOC and NOx indices through a digital multi-sensor module. The gas indices help observe variations: they are not a selective measurement of chloramines.
The information is available on a 7-inch poolside panel, at the front desk and, optionally, in the cloud. Water parameters, visual observations, air, occupancy and plant status can be viewed together.
In this installation SPACE AIR monitors the air: it does not control the air handling unit and does not generate the stated electricity saving.
An efficient pool is not the one that always runs less. It is the one that can tell when more is needed and when it can consume less.
Aqutron Vision. The pool changes. Now the plant knows.
How to read this case study
For filtration, chemicals and make-up water, the comparison is between the monthly averages of the 12 months before and the 3 months after the intervention. It is a descriptive comparison, not normalised for seasonality or attendance; it does not isolate the causal contribution of individual technologies.
The percentages come from data reported by the operator. The charts show aggregates or explicit calculations, not reconstructions of hourly readings. The results concern this installation and are not a universal guarantee of savings.
Sources