Back to Bigger Picture

BIGGER PICTURE Energy & climate

Giving wastewater bacteria the air they actually need

Cuxhaven’s treatment plant used a virtual model to adjust aeration—even where a real-time sensor was missing.

AI-generated illustration: Conceptual wastewater tank cutaway with controlled aeration and oxygen sensors.
Wastewater tank cutaway with controlled aeration and oxygen sensors.AI-generated illustration of the idea, not a photograph of the actual people, equipment or event.
Reported reduction
30% of aeration electricity
Controlled process
Five parallel biological tanks
Constraint
Effluent limits still met, according to the supplier

How the idea works

  1. Estimate incoming load

    Virtual sensors infer nutrient and carbon concentrations.

  2. Model oxygen demand

    The treatment model estimates the air needed for elimination.

  3. Adjust aerators

    Controls match aeration to changing load and effluent targets.

Clean water has an electricity bill

Biological wastewater treatment relies on processes that need oxygen. Supplying it with blowers and aerators consumes substantial electricity, so over-aeration is expensive. At Cuxhaven, EWE wanted to reduce that demand while keeping treated water within required limits. The aim was not simply to turn the air down, but to match it to the work the treatment process had to do.

Sources: [1][2]

A sensor made from relationships in the data

The plant lacked online measurements for some incoming pollutant concentrations. Xylem therefore built virtual sensors: models estimating carbon, nitrogen and phosphorus loads from available plant data. Those estimates fed a digital model of the treatment processes. This is a useful twist on AI: it can infer a missing process variable rather than requiring every quantity to be directly measured.

Sources: [1][2][3]

Five tanks became a control experiment

After a period of manual operation, optimized control began in August 2017 for five parallel biological treatment tanks. The system calculated aerator settings appropriate to the changing load. Xylem’s case accounts report 30% less aeration energy while maintaining effluent compliance. The comparison concerned the treatment process, not every electricity consumer on the site.

Sources: [1][2]

Keep the denominator visible

The accounts come from the technology supplier, rather than a randomized independent trial. They also give different annual kilowatt-hour equivalents in different versions, so the consistent 30% aeration figure is the clearest headline. For another utility, incoming wastewater, equipment and discharge requirements would need their own model and validation before assuming similar savings.

Sources: [1][2][3]

The original sources

  1. XylemWastewater treatment plant uses AI to reduce aeration energy use by 30%
  2. XylemCuxhaven reduces aeration energy use by 30% while ensuring effluent water quality compliance
  3. XylemWastewater management software and process decision support

What to explore next

A steadier kiln can burn less fuelAn ECG alert trial ended with fewer deaths after 90 days
Explore all 99 applications