<img height="1" width="1" style="display:none" src="https://www.facebook.com/tr?id=323483448267600&amp;ev=PageView&amp;noscript=1">

Solutions by industry / Chemicals & pharma

Pharma clean utilities: from fixed rules to real capacity

At a global pharmaceutical manufacturer, Wizata is replacing conservative rules along the water-treatment chain with data-driven decisions, using data the site already has.

~40%fewer softener regenerations, on the site's own data
3 unitson one water chain: softeners, distillers, chillers
Existing datahistorian and lab data, no new sensors to start
Industry
PharmaceuticalsClean utilities
Focus
WaterChemicalsEnergy and steamSupply reliability
With Wizata
Digital twinSoft sensorsRecommendationsCondition monitoring

First results

Less water, fewer chemicals, a safe supply

Softener regenerations ~40% fewer

On the site's own data, at a cautious first checkpoint. The data suggests the resin can go further.

01

Water and chemicals saved

Every regeneration avoided saves its rinse water and its regeneration chemicals.

02

Supply secured

Distillers follow demand, so the clean-water buffer stays covered, with less electricity and steam.

03

Maintenance on condition

Chillers serviced when their efficiency drifts, not on a calendar or after a failure.

The challenge

Static setpoints, paid for every day

Before it reaches production, municipal water goes through several purification stages. Each stage runs on rules set long ago, conservative by design: the safe choice when nobody can see the real state of the equipment.

Those margins are paid for every day, in water, regeneration chemicals, electricity and steam, on equipment that is rarely anyone's first priority.

Fixed regeneration volumesDistillers capped below capacityA thin clean-water bufferEfficiency judged by experience

What changes with AI

  1. From a fixed volume to the capacity really left in the resin.
  2. From a static cap to an operating mode that follows live demand.
  3. From fixed capacity to measured efficiency for each unit.
  4. Decisions that are logged and traceable.

One water chain

Go from city supply to pharmaceutical-grade water

Municipal waterIncoming supply, variable hardness
Use case 1SoftenersRemove hardness; the resin is regenerated periodically
Reverse osmosisNot in scope
Use case 2DistillersFinal purification, two units in parallel
Use case 3ChillersCool distilled water for the next steps
ProductionPharmaceutical-grade water

Our approach

A model for each unit, all on one platform

Starting from historian, SCADA and lab data already on site. Recommendations first, approved by the utilities engineers.

1Softeners
Early results

Regenerate on real capacity

A virtual resin-load indicator, computed from conductivity and hardness at inlet and outlet, estimates the capacity really left. Regeneration follows that capacity instead of a fixed volume, with the sanitary safety margin kept.

2Distillers
In progress

Run what demand needs

Two distillers, one decision: run one or both, and at which setpoint. The operating mode follows live water availability and tank demand, so supply is always covered without unnecessary run time.

3Chillers
In progress

See which unit drifts

A coefficient of performance, computed continuously for each chiller, shows which unit leaves its optimal window, before it fails. No control access needed.

Start with utilities

Utilities first, inside your limits

Your existing dataHistorian, SCADA and lab data. No new sensors to start.
No control access to startMonitoring and recommendations first. Your systems stay yours.
Inside your limitsEvery recommendation is checked against your sanitary and quality limits.
Traceable by defaultData used, model version and approval recorded for every decision.

How we deploy AI under 21 CFR Part 11 →

Built on the Wizata platform

Explore the platform →
Data HubHistorian, SCADA and lab data of the whole water chain in one digital twin.
AI LabResin-load and efficiency indicators, and the distiller model, versioned.
Control PanelRecommendations approved by the utilities team, alerts when a unit drifts.

Next step

Which process would you optimise first?

Your contact Jérémy Lambert Pharma lead, Wizata jla@wizata.com