UK Power System Dispatch
National-scale DC load-flow with security-constrained dispatch, RES curtailment, and congestion pricing.
Energy system models, open & auditable.
We support your decision-making by building transparent, production-ready models of power systems using Python, Pyomo, and PyPSA. Flexible, reproducible, and typically 10× cheaper than legacy black-box tools.
- PyPSA
- Pyomo
- NumPy
- Pandas
- SciPy
10×
Cost reduction
100%
Model transparency
⚡
Rapid iteration
=== Arbitrage + Ancillary Reserve (January 2025)
Battery Pmax (nameplate): 50.00 MW
Chosen reserve (fixed R): 34.00 MW
Usable for arbitrage: 16.00 MW
Reserve rate: €7.00 per MW per hour
Reserve revenue: €177,192.58
---- Arbitrage component ---------------------------
Imported energy: 2,902 MWh @ 50.11 €/MWh
Exported energy: 2,581 MWh @ 82.75 €/MWh
Import cost: €145,456.06
Export revenue: €213,615.97
----------------------------------------------------
TOTAL PROFIT: €245,352.49
Project Portfolio
Demonstrative and client-ready projects. Click a tag to filter.
Battery Sizing Optimizer
Co-optimises power/energy rating with degradation and round-trip losses; NPV-driven sizing.
Capacity Expansion (CE)
Long-run least-cost build plan with policy constraints (CfDs, cap-and-floor, CfD strike limits).
Battery Arbitrage + Ancillary Co-Optimisation
We co-optimise a 50 MW / 200 MWh battery between energy arbitrage and ancillary service commitment.
Monetising Stranded Energy
Modelling the monetization of stranded/curtailed energy using Bitcoin mining as flexible off-take.
EV Charging Scheduler
Stochastic scheduling against day-ahead & imbalance markets with network constraints.
How we work
Collaborative, open, and action-oriented — from discovery to handover (or ongoing partnership).
-
Discover & align
We begin with working sessions to understand your objectives, decision deadlines, constraints, stakeholders, and success metrics. Together we define a focused, testable scope.
Outputs: problem statement, KPIs & constraints, acceptance criteria, roadmap & milestones.
-
Data & parameters
We inventory available data, document quality & gaps, and agree project-specific parameters and assumptions (with ranges). Where data is missing, we propose pragmatic proxies.
Outputs: data catalog, parameter sheet, assumptions log, gap-closure plan.
-
Build the model — collaboratively
We implement in PyPSA/Pyomo with a clean repo, tests, and scenario runner. You see progress early and often via co-working reviews, so the model evolves with your feedback.
Outputs: running model, reproducible repo (tests/CI), docs & notebooks, prioritized backlog.
-
Validate & stress-test
We back-test against historicals, run sensitivities and stress tests, and profile performance. Assumptions are challenged and tuned until the model is fit for decisions.
Outputs: validation report, calibrated parameters, scenario set, risks & mitigations.
-
Explain results & actions
We don’t just hand over charts. We walk you through the drivers, trade-offs, and uncertainties so you have clear, defensible takeaways and next actions.
Outputs: results deck/dashboard, decision memo, action list (what/where/when, with rationale).
-
Deploy & handover — or ongoing partner
Choose full handover (training, docs, CI, packaging) to run in-house long-term, or keep us on as a partner for updates, new scenarios, and model ops as your needs evolve.
Outputs: deployment plan, training sessions, support options (handover or retainer).
Guiding principles
Open, transparent, bespoke — and focused on decisions you can act on.
FAQ
Do you replace tools like PLEXOS or just complement them?
Usually replace for new work. We build open, transparent models that cover your scope without license costs. For transitions, we can export/import data and interoperate alongside PLEXOS during a handover period.
What stack do you use?
Python-first: PyPSA, Pyomo, NumPy, Pandas, SciPy; visualizations in Plotly/Matplotlib; packaging with Poetry; CI with GitHub/GitLab; all open-source. Code, data, and assumptions live in a clean repo with docs and tests.
How do engagements work?
Fixed-fee discovery to align on scope and KPIs → collaborative model build with regular reviews → validation & stress tests → results walkthrough with clear actions. Afterwards: full handover and training, or an ongoing support retainer.
Who owns the model and IP? Can we run it in-house?
You do. Deliverables are provided under a permissive license, built on open-source dependencies. We set you up to run everything internally (repo, environments, CI), and train your team so you’re not dependent on us.
How do you validate results and deal with uncertainty?
Back-tests against historical periods, sensitivity and stress testing on key parameters, and peer-reviewed assumptions. We present drivers and ranges (not just point estimates) so decisions reflect uncertainty realistically.
What do you need from us? How do you handle data security?
Objectives/KPIs, any constraints or policies, and access to relevant datasets (or contacts). We sign NDAs, use private repos, minimize PII, and can work inside your cloud/VPN if preferred. All data handling is documented and reproducible.