A Sector Moving from Reaction to Delivery

It was a pleasure to chair the Smart Wastewater and Sewage Systems Conference on 30 September and 1 October 2025. What stood out most was the clear sense of momentum across the sector. After years of media headlines flagging spills, challenges, and failures, this event showcased tangible progress — utilities translating intent into measurable improvement. The tone was positive, the discussion grounded, and the examples refreshingly solution-led. 

The conference opened on a strong commercial note with Anglian Water outlining the innovations it plans to scale between 2026 and 2027 to deliver meaningful AMP8 performance gains.

Defining What ‘Good’ Looks Like by 2027

Alex Rosenbaum, PR24 Portfolio Strategy Manager, framed progress around four “engines”: the regulatory context, data and telemetry, delivery mechanisms, and digital operations.

Alex shared a shortlist of scalable smart-sewer technologies — a mix of visionary and implementable ideas — ranging from machine-learning-driven alerts to adaptive sewage storage and automated gating. The presentation’s true strength was in defining what “good” looks like for Anglian Water by 2027:

• Telemetry coverage across 90 % of priority CSOs and high-risk sewers
• Blockage prediction and control-room capability delivering 15–30 % spill reduction
• Storm-overflow spill-rate decreases tied directly to measurable ODI outcomes

Crucially, Alex emphasised that AI integration must connect to regulatory reporting — not operate as data for data’s sake. Industrialising pilots through catchment-level standardisation frameworks is now the key to scaling effectively.

Blending Models to Predict the Unpredictable

Next came Jonathan Stokes, Head of Storm Overflows Delivery at Yorkshire Water, joined by Jo Kelsey, Technical Director of the Yorkshire Storm Alliance. Jonathan began with a pointed reminder: “Once upon a time, we verified networks with flow surveys and spreadsheets.” That mindset, he argued, can no longer keep pace with AMP8 — or the even broader AMP9 horizons.

Their new hybrid storm-overflow hydraulic-planning model blends statistical models and deterministic, physics-based rules with fast, adaptive, data-rich inputs — creating a predictive capability able to model overflow behaviour under variable rainfall and groundwater conditions. Traditional hydraulic models can’t deliver 2,000 storm-overflow assessments quickly enough; this hybrid approach can.

For utilities shaping AMP9 plans, the takeaway was clear: design AI-ready architectures now, so deterministic and statistical frameworks can coexist without constant rebuilds.

Using Predictive Analytics to Understand Phosphorus

Joe Vertikan, Head of Digital Data and Innovation, and Philip Ashton-Briggs from @one Alliance / Anglian Water discussed applying predictive analytics to anticipate phosphorus levels in wastewater. Their central question: Can machine learning forecast future phosphorus inflows to optimise treatment operations?

Through collaborations with Cranfield University and detailed site visits — at Halstead, Foston, Colby and Bridstalk — the team proved that it can. Forecasting phosphorus enables optimisation of energy and chemical use, ensuring assets are built only where truly needed, driving both cost efficiency and environmental performance.

Intelligent Catchment Management for Cleaner Rivers

Ally Potts, Head of Technical for Clean Rivers and Seas Task Force at Southern Water, presented the utility’s Clean Rivers and Seas Plan — a holistic redesign of how the entire sewage system operates. Southern Water’s approach blends collaboration across local authorities, councils, the Environment Agency, Defra, and private owners to deliver systemic improvement.

Ally explained the utility’s three-stage network-optimisation model, progressing from:

1. Single-asset intelligence
2. Autonomous-asset control
3. Catchment-wide decision-support tools

Machine learning is already improving traditional modelling. Non-modellers can now test future scenarios through user-friendly catchment tools, with results feeding into deterministic models for validation. Southern Water has installed 30,000 sewer-level monitors, using ML feedback to target interventions precisely where needed.

Automating Optioneering to Close the Workforce Gap

Adam Anyszewski, Co-Founder of Continuum Industries, opened with a reality check: around 20 % of engineers will retire within five years, just as the industry faces a £44 billion AMP8 programme and a 36 % workforce expansion required by 2030.

That backdrop set the stage for Optioneer, a GIS-based optimisation platform used by Severn Trent Water, Southern Water, Northumbrian Water, and the Netherlands’ Thames project. The software models sites, pumping stations, treatment works, and depots to screen and rank infrastructure options in minutes — cutting early-design and feasibility timelines from weeks to hours.

The key message was that automation isn’t about replacing engineers but amplifying their strategic impact — freeing teams to focus on value rather than repetition. In a market under skills pressure, that difference could determine AMP8 delivery success.

Re-Engineering Legacy Systems for Catchment-Level Control

Wez Little of Schneider Electric reframed an old truth for modern times: the legacy combined-sewer system, once adequate, is now unfit for purpose. Population growth and extreme rainfall have outstripped its capacity.

Schneider Electric’s solutions complement digital-twin and storm-harvesting systems, helping utilities treat wastewater with the same precision as clean water. Wez argued that regulatory compliance and resilience will soon require a shift from site-by-site linear control to catchment-level active management — where assets are standardised, visualised, and orchestrated in real time across the network.

Section 82 in Focus: The Cost, the Risk, and the Realism

Rev’d Dave Walker, Global Trends Director at Detronic, offered a timely briefing on Section 82 delivery. He explained that a Parliamentary review later in October 2025 will revisit the legislation’s feasibility, acknowledging that parts were drafted before viable instruments existed — for example, ammonia-sensor requirements that pre-dated reliable technology.

Walker reminded delegates that a typical Section 82 kioskcan cost £30–40 k per discharge point once concrete, solar, and sampling hardware are included. Across 6,000–9,000 CSOs, national spend could reach hundreds of millions — before reliability is proven.

By contrast, Section 81 monitoring (EDM, pump control, energy and carbon optimisation) is already certified and deployed nationwide. His argument to policymakers: redirect near-term investment toward proven Section 81 capabilities while continuing targeted Section 82 pilots through mid-AMP8 — a pragmatic balance between ambition and deliverability.

Technology Readiness and the Case for Pragmatic Adoption

During the subsequent panel discussion, Schneider Electric noted that although Section 82 technologies — smart pumps, variable-speed drives, and energy-efficient controls — exist today, adoption remains immature: only about 10 % of assets currently use modern control systems.

Walker demonstrated a fibre-optic sensing prototypethat could eventually provide reliable, scalable monitoring. Panellists agreed that, until these technologies reach full commercial proof, utilities must prioritise architecture readiness, integration capability, and pilot-to-scale pathways.

This exchange captured the moment the industry is in: technology potential is no longer the issue — execution maturity is.

Alex Shore, Water Network Spills Lead at Severn Trent, expanded the Section 82 debate with a simple question: what triggered the pivot to “buying differently”?

The answer lay in experience. Earlier drives for speed had fragmented programmes into siloed workstreams — infiltration, wear adjustments, asset renewals — that met short-term KPIs but lost cross-catchment context.

Today, Severn Trent is reframing procurement and partnerships around regional outcomes, empowering regional managers as improvement integrators who combine risk, ecology, and delivery intelligence. The shift has exposed a vital truth: quick wins deliver KPIs but can miss high-risk sites, particularly dry-day spills that occur in plain sight.

The company now focuses on high-spilling and environmentally sensitive water bodies as its organising principle — an evolution born not from a single tipping point but from gradual recognition that metrics were driving the wrong behaviours.

High-Frequency Monitoring: Seeing Rivers in Real Time

Mike Bowes, Nutrient Hydrochemist at the UK Centre for Ecology & Hydrology, delivered a compelling case for high-frequency water-quality monitoring as a tool for predicting nutrient blooms and targeting interventions.

His findings were unequivocal: sampling frequency and context matter. Data should be interpreted against flow, not just time, and blooms are often weather-gated rather than seasonal.

Key insights included:

• Ammonium concentration is one of the most robust indicators of CSO events      during rainfall.
• AI-enhanced sampling can sharply reduce monitoring costs while improving accuracy.
• High-frequency analytics create actionable intelligence for both regulators and      operators, enabling earlier warnings and more precise interventions.

Digitising Everything: The Aquafin Transformation

Ronny Goossens, Head of Digital at Aquafin (Belgium), showcased a whole-company digital transformation designed to make wastewater operations greener, smarter, and affordable under tightening EU regulation. His core message: technology and organisational culture must evolve together.

Aquafin has merged IT and OT teams into a single digital department using agile, product-based structures within a scaled-agile framework. Its “Blue Data Lake” on Microsoft Azure integrates field, historical, and external datasets — including rainfall forecasts and day-ahead electricity prices — to drive automated decisions across the network.

Practical outcomes included:

• Anticipatory basin control, lowering reservoir levels before storms.
• Biogas-engine optimisation, scheduling generation when power prices peak.
• Real-time overflow monitoring via 2,000 IoT sensors, sharing data externally under      EU Space Data standards.
• Storm-point logic reducing untreated discharge by 69,000 m² per year through      predictive delay.

For other markets, the strategic message was clear: begin IT/OT convergence early. Cultural and architectural integration takes years, but the payoff is system-wide agility and regulatory confidence.

Integrating Culture, Data, and Guardianship in New Zealand

David Moore, Smart Systems Manager at Watercare (Auckland, New Zealand), began not with charts but with a kākāia — a Māori blessing connecting land, water, and people. It set the tone for a presentation demonstrating that technology and cultural guardianship can strengthen each other.

Watercare’s mission is to make Auckland’s wastewater system more resilient, intelligent, and respectful of its environment. Moore unveiled a risk model combining asset data, performance metrics, and environmental variables — forming the backbone of a major digital rollout:

• 5,000  new level sensors installed, with 500 more added each month
• Machine-learning  and AI analytics through “Storm Officer” to predict blockages and infiltration
• Real-time data integration to prioritise CCTV and rehabilitation programmes

The global insight was simple: digital transformation succeeds when rooted in place-based responsibility, where technology amplifies rather than replaces environmental care.

Denmark’s Fourth-Stage Leap in Pollution Control

Sophie Thorgaard, Project Manager at RS Lands (Denmark), guided delegates through the company’s flagship ozonation project at the Egå Wastewater Treatment Plant — a European benchmark in adopting a fourth treatment stage for micropollutant removal.

Her presentation highlighted Denmark’s leadership in combining operational efficiency with environmental precision. Ozonation not only removes pharmaceuticals and microplastics but also reduces downstream treatment loads, cutting both chemical use and energy demand.

UKWIR and the Chemistry of Continuous Improvement

Jennifer Hughes, Strategic Programme Manager at UKWIR (UK Water Industry Research), reminded delegates that this non-profit consortium of 18 UK water companies has delivered more than 1,000 collaborative projects since 1993. Among them, the Chemical Investigations Programme (CIP)stands apart as one of the most complex and consequential.

Initiated in 2009 to meet the EU Water Framework Directive’s demand for evidence of cost-effective chemical management, CIP has evolved through successive phases:

• CIP3  (2020–2025) refined the questions — antimicrobial resistance,   microplastics, sludge behaviour, removal mechanisms, and long-term trends — proving that source bans work and not every emerging contaminant needs a billion-pound fix.
• CIP4 (2025–2030) is the new frontier, tackling PFAS/PFOS, microplastics, soil residues, estuarine dynamics, wetlands integration, biosolids, and non-target screening.

Jennifer  also addressed data integrity, noting that “you can’t make billion-pound decisions on shaky science.” CIP now operates with standardised sampling specs, UKAS-participating labs, and centralised interpretation to ensure comparability.

Her closing reflection captured the essence of progress: “CIP has taught us that one size doesn’t fit all — but shared data and consistency do.” With CIP5 already scoped for 2030, the work continues: the next unknown is always just around the corner.

Northumbrian Water: Optimise Before You Pour

Steve Blanks, Emerging Technology Product Manager & Innovation Lead at Northumbrian Water, set the tone with a candid scene-setter: climate change, urban growth, hard surfaces, and reputational pressure are converging — and “more concrete” can’t be the default answer.

Northumbrian Water’s “upside-down pyramid” strategy puts asset optimisation and smart network management first, followed by surface-water separation/SuDS and nature-based solutions, with storage as the last resort.

Its operating model follows three disciplined verbs: observe → predict → act.

• Observe: four sensor types; hard-wired control points (1-min cadence into SCADA)      plus ~800 tactical field sensors — right-sized, not over-deployed.
• Predict: hyper-local ensemble weather (forecasts every 15 minutes), continuous      hydraulic modelling in an Azure-fed digital twin.
• Act: a real-time decision engine proposes control strategies (e.g., closing a penstock to redirect flow), always with a human in the loop and an air-gap to plant. Each action is logged for learning and ML tuning.

Key principles emerged:

1. Optimise  before you pour.
2. Right-sized sensing beats sensor sprawl.
3. Ensembled weather + digital twin = fewer spills.
4. Human-in-the-loop builds trust.
5. Own the data; design for OPEX and scale.

During the Day 2 roundtables, participants compared practical experiences and debated where digital investment should concentrate next. One standout observation: modularity and timing matter. Adopting flexible, interoperable systems that let operators “use the best tech when ready” simultaneously reduces vendor lock-in and cyber-risk.

Reflections on a Sector Ready to Deliver

The conference closed on an energising note. Across all sessions, what emerged was not just innovation but operational seriousness— an industry increasingly capable of matching ambition with delivery.

From AI-ready architectures and predictive analytics to pilot deployment and cultural transformation, every contribution offered transferable lessons others can apply immediately.

As Chair, I was struck by the level of engagement: no “death by PowerPoint,” just focused 15- to 20-minute presentations followed by extended, practical Q&A that doubled the learning value.

I left convinced — as did many in the room — that this sector is not only innovating but ready to deliver the intelligent, resilient wastewater systems that the future demands.

Steve Thomas, Chair, Strategy Engineering Research Group 3 October 2025