Across the UK water sector, modelling teams are working under an AMP-8 cycle that demands more evidence than any programme before it. Many modellers are running into the same challenges: multiple projects running at once, limited hands on the tools and a growing list of models that need updating, recalibrating or verifying before decisions can be made.
This shift places new attention on the core stages of modelling itself. The speed at which models are created, calibrated and verified now shapes the pace of AMP-8 delivery, and these stages have become the natural focus for teams looking for practical ways to keep projects moving.
Under previous investment periods, modelling activity usually followed a predictable sequence of building the model, calibrating it, running scenarios, verifying outputs, then moving on. AMP-8 requires significant acceleration of this procedure to meet spill reduction targets. Teams are working to consider requirements for different programmes (DWMP, WINEP, growth and resilience programmes) and attempt to improve verification for multiple purposes. Whilst offering potential to reduce the overall workload this increases complexity of verification and solutions development studies.
Modelling teams now work under stronger accountability pressures. Models need to predict factors like spill count over long horizons (10-year periods), capturing seasonal and year-to-year hydrological variation. Many existing models were built primarily to predict short-term responses or individual events rather than sustained long-term behaviour and therefore more complex items such as rainfall dependant and groundwater infiltration require significant improvement.
Teams are required to:
These demands expand the number of catchments modellers need to verify. Spill verification is a major strain because it must be performed across a period of multiple years to often sparse depth-only datasets and any baseline change triggers further checks. This creates a working environment defined by constant movement through model updates, calibration and verification, with sustained pressure from DWMP review, spill-reduction assessments and growth modelling.
Against this backdrop, a practical question is surfacing across utilities and consultancies.
How do we keep up with AMP-8 and work towards AMP-9 when manual modelling processes take longer than the programme allows?
This question is shaping the shift towards modelling approaches that reduce the time spent on repeated tasks and create more room for interpretation and engineering judgement. These approaches hinge on three ideas:
They offer a way to increase modelling throughput while maintaining the accuracy and governance standards the sector relies on.
Across utilities and consultancies, there is increasing curiosity about modelling automation tools that can take on the more repetitive elements of model setup. These approaches give modellers a way to move through early development stages more smoothly by applying consistent rules for model updates and simulation scenario updates.
Instead of repetitively copying and pasting data between multiple systems manually, modellers integrate the systems together in an auditable workflow where they model updates and the submission of simulations. This creates a modelling pace that aligns better with AMP-8, where baselines evolve frequently and projects often overlap. The modeller stays firmly in control, but the work moves forward with less manual interruption.
Verification has traditionally been a point of friction. Modelling teams invest significant time in detailed checks and the full calibration of observed data, yet programme teams often experience slow, conditionally stable simulations that delay progress. This creates tension between the technical discipline of verification and the need to keep programmes moving. Auto-calibration reduces that tension. It improves calibration quality for modellers and accelerates the delivery of results for planning teams.
Auto-calibration strengthens the verification process by accelerating the calibration step and supporting the wider validation workflow. It highlights deficiencies earlier, reduces rework and shortens the route to a fit-for-purpose verified model. This supports the faster review cycles expected in AMP-8 and gives planners earlier visibility of results that guide investment discussions.
Another noticeable shift involves the scale of scenario testing. Manual modelling methods naturally limit the number of variations a team can assess. Optimisation-led approaches provide confidence that the best solution has been achieved in the fewest cases necessary.
For AMP-8 planning, this wider search space is extremely valuable. It helps teams understand performance across a broad range of conditions, including growth pressures, climatic changes, operational constraints and future scenarios that would be impractical to explore manually. With a more complete understanding of the range of options available, modellers can support decision-makers with greater confidence and clarity.
The ideas outlined above are already shaping how many teams approach AMP-8 delivery. Modellers want solutions that become fit for purpose as quickly as possible without force fitting it to match the observed data, support broader scenario testing and produce clearer evidence at the point it is needed.
This is the environment in which HEEDS for Water, supported by STRIDE’s integration and enablement work, has become an important part of modern modelling practice. The platform brings automation, large-scale scenario exploration and auto-calibration into a single, consistent workflow, and it fits naturally alongside the modelling tools already used across the sector. Together, they provide a practical way for teams to increase throughput, maintain auditability and deliver evidence fast enough to match AMP-8 expectations.
For water companies and consultancies preparing for the years ahead, these approaches offer a route to working at the speed the programme demands while supporting the quality and confidence that regulatory processes require.