Novo Nordisk is testing whether wastewater drawn directly from pharmaceutical manufacturing can be purified and returned to the production line, with company engineers estimating the approach could halve water use at the source. A pilot using membrane-based treatment was installed in December 2025 at the company's Hillerød facility in partnership with the Technical University of Denmark (DTU), according to a report published 6 September (Interesting Engineering).
Key takeaways
- Novo Nordisk and DTU installed reverse-osmosis-based water purification at the Hillerød site in December 2025, in a project reported 6 September (Interesting Engineering).
- Company modelling suggests recycled production water could cut overall water consumption by 50%, environmental lead Andreas Herfelt said (Interesting Engineering).
- Danish pharmaceutical water use rose 35.8% between 2012 and 2022 while total industrial water demand stayed flat, an analysis by the Danish Academy of Technical Sciences found (Interesting Engineering).
- DTU researchers are running continuous post-membrane water analysis to feed risk assessments for the Danish Medicines Agency, with Professor Anders Baun supporting the work (Interesting Engineering).
- Separate research into ultra-selective "POMbrane" crystalline membranes, published 20 January, has been positioned by its developers as relevant to pharmaceutical separations and water reuse (Phys.org).
What happened: a pharmaceutical-grade reuse pilot at Hillerød
The project, being carried out with researchers at the Technical University of Denmark (DTU), targets one of the more difficult forms of industrial water recycling: producing water pure enough for injectable medicines (Interesting Engineering). Today, Novo Nordisk already recycles some process water in cooling towers, where it evaporates after removing excess heat, but the new study asks whether water generated by the pharmaceutical process itself can be returned to manufacturing (Interesting Engineering).
DTU researcher Ravi Kumar Chhetri installed the membrane-based system at Hillerød in December 2025, using reverse osmosis to push water through a semipermeable membrane and separate it from dissolved substances (Interesting Engineering). Because the recovered water would have to meet the strictest quality standards for injectables, the team is sampling the permeate continuously and feeding the results into risk assessments of any residual substances (Interesting Engineering).
The background pressure on water is rising: an analysis by the Danish Academy of Technical Sciences found that Denmark's overall industrial water consumption was almost unchanged between 2012 and 2022, but pharmaceutical manufacturing's water consumption increased 35.8% in the same window (Interesting Engineering).
"If we can recycle the wastewater streams that come from the processes in the pharmaceutical production itself, we estimate that we can reduce our water consumption by 50 per cent," said Andreas Herfelt, Environmental Lead at Novo Nordisk (Interesting Engineering). "This is because the water used in the production of pharmaceuticals – and especially those that are to be injected – is subject to the strictest quality requirements. This is to prevent unwanted substances from ending up in the patients who need the medicine," Herfelt added (Interesting Engineering).
On the risk side, "In the risk analyses, we assess whether there are substances remaining in the purified water and whether they pose a high or low risk. If the risk is at the high end, we look at how to address it," said Chhetri (Interesting Engineering). "Recycling water in a production company has great potential for water savings compared to other water-saving measures such as optimizing processes with high water consumption," he added (Interesting Engineering).
The work is also aimed at regulators: "It is with DTU's help that we can obtain the necessary data for the Danish Medicines Agency as documentation that recycling the water will not pose any risk to either the quality of our products or patient safety," Herfelt said (Interesting Engineering).
In a separate line of research reported 20 January, an India–Singapore team built "POMbranes" — crystalline polyoxometalate films with permanent 1-nanometre pores — and described the class of material as relevant to drug purification, solvent recovery and industrial water reuse (Phys.org). The developers say the membranes showed almost ten times the separation performance of existing technologies in their tests and remain stable across different pH ranges, suggesting potential relevance to demanding separations in pharmaceutical manufacturing (Phys.org).
Specification read
The Hillerød pilot sits at the tertiary end of a typical pharma wastewater train, with the membrane step positioned downstream of biological treatment and acting as the barrier that decides whether the effluent can be reclassified as production-grade water rather than discharged. Reverse-osmosis systems in this duty are normally sized in cubic metres per day, but the source text does not give a flow figure for the Hillerød skid; readers should treat the project as a single-site pilot rather than a full-plant retrofit until a throughput is published. The 50% reduction claimed by Novo Nordisk is a company estimate tied to its current water use at the site, and Danish pharmaceutical water demand grew 35.8% between 2012 and 2022 even as total industrial water use stayed flat, so any buyer extrapolating the saving to their own plant should benchmark against their own intake baseline before sizing equipment (Interesting Engineering).
For a plant manager evaluating a similar reuse scheme, the train will typically run screening → primary clarification → biological treatment (commonly MBR or MBBR for pharma effluents) → tertiary membrane (reverse osmosis, sometimes preceded by UF) → polishing/disinfection, with the reverse-osmosis permeate then validated against the pharmacopoeia-grade water specification before it can touch the process. The Hillerød work specifically bears on the membrane-plus-risk-assessment stage, where the question is not whether the RO works mechanically but whether any trace solutes in the permeate are acceptable under the relevant injectable-water quality standard. If your plant runs an MBR-fed tertiary line at a few hundred to a few thousand m³/day with effluent COD typically in the low hundreds of mg/L and you are under pressure to cut raw-water intake or strengthen water security, this is the kind of pilot that tells you whether closing the loop to production water is technically defensible. A plant specifying biological-stage equipment for this duty would look at Mbr Membrane Systems and the wider Pharma treatment line; for a comparison of MBR against conventional activated sludge in a chemical/pharma duty, see MBR vs Conventional Activated Sludge for Chemical & Pharma Wastewater in Willow Island (2026 Engineering Guide). For adjacent MBR benchmarking in food and beverage duty, see MBR vs Conventional Activated Sludge for Food & Beverage Wastewater in Green Forest (2026 Guide). Plant-side dust and vapour control around upstream API handling can be specified through Air Dust Treatment.
FAQ
What scale of water saving is the Hillerød pilot targeting?
The company estimate is a 50% cut in water consumption if production-process wastewater can be returned to manufacturing, against a Danish backdrop in which pharmaceutical water use rose 35.8% between 2012 and 2022 (Interesting Engineering).
What is the pilot actually doing at Hillerød?
A reverse-osmosis membrane unit was installed in December 2025 by DTU researcher Ravi Kumar Chhetri, and the team is now continuously analysing the permeate to feed risk assessments for the Danish Medicines Agency (Interesting Engineering).
How long does a pharma-to-pharma water reuse project like this typically take from pilot to approval?
Lead times depend on how many full-scale batches of risk-assessment data a regulator requires; a typical pharmaceutical RO reuse programme, once pilot data is in hand, runs on the order of 12 to 36 months to validation (general industry range, not from the sources). Novo Nordisk says the current project is structured to give the Danish Medicines Agency the documentation it needs (Interesting Engineering).
What should a buyer look at when specifying equipment for a similar reuse scheme?
Focus on the biological stage feeding the RO (commonly MBR or MBBR for pharma effluents) and the membrane stage itself, with risk-based validation of any trace solutes against the relevant injectable-water quality standard (general industry range, not from the sources). The Hillerød pilot specifically stresses the risk-assessment step rather than the membrane mechanics (Interesting Engineering).
