Grid congestion does not mean waiting years for a heavier connection. With peak shaving, an EMS and steerable flexibility a plant stays within its existing connection, and Crux Digits builds that control on your own meters and SCADA.
Don't wait years for a heavier connection. With peak shaving, an energy management system and steerable flexibility you keep your plant inside its existing grid connection, and cut energy cost now. Crux Digits builds the operational AI that steers batteries, loads and processes automatically and in real time. Fixed prices, you own the code, EU AI Act- and GDPR-aware.
Last updated: 14 September 2026
Industrial sites don't have to wait years for grid reinforcement. By flattening demand peaks with peak shaving and battery storage, running an energy management system (EMS) that automatically steers loads, storage, generation and flexible processes, and shifting consumption to low-cost or low-congestion hours, a plant can stay within its existing grid connection and cut cost now. AI turns meter, sensor, price and process data into automatic, real-time decisions, not a dashboard someone watches.
Crux Digits is an AI consultancy and custom software development company in the Netherlands, founded in 2022 and based in Nieuwegein in the Utrecht region, working for SMEs (MKB) and mid-market organisations. Bilingual (EN/NL), with transparent fixed pricing (a €2,500 AI audit, a €20,000 Production-ready MVP and a production build from €50,000), delivered by a named expert (founder Tom Joseph), with 13 real case studies, EU AI Act- and GDPR-ready. A focused alternative to the large enterprise consultancies.
Grid congestion (netcongestie) simply means the electricity grid is full in many places: more transport capacity is being requested than the grid can deliver at that moment. But that is the problem, not your action plan.
The reflex is to apply for a heavier connection. In practice that is no short-term fix: TenneT and the regional grid operators (DSOs) have paused or delayed connections across large parts of the Netherlands, and the queues for extra transport capacity run into years. Grid reinforcement (new cables, heavier substations) is a decade-long infrastructure project, not an order you place this month.
So the win is not in requesting more capacity, but in using the capacity you already have more intelligently. That is exactly where a plant manager does have levers to pull.
You don't have to wait on the grid operator to act. Four levers keep your site inside its existing connection and cut cost:
Flatten your demand peaks with a battery: at peak moments the battery discharges, so you stay within your contracted connection value and avoid expensive peak tariffs and overrun penalties.
A system that continuously measures and automatically steers: loads, storage, generation (solar, CHP) and flexible processes. The EMS is the brain that actually operates the other three levers.
Shift flexible loads (cooling, charging, batch processes) to low-cost or low-congestion hours based on day-ahead and imbalance prices. Run heavy work when power is inexpensive, not when the grid peaks.
Time-based transport rights and non-firm connections reward you for shifting off-peak. Through congestion management and redispatch (e.g. GOPACS) the grid operator even pays you to power down at busy moments.
Go deeper on each lever: peak shaving & battery storage and the energy management system for industry.
The four levers already exist. What's missing is the intelligence that operates them together, automatically, at the right moment. A battery without smart control shaves the wrong peaks; an EMS that only measures is an expensive dashboard.
Operational AI turns meter, sensor, energy-price and process data into decisions that steer batteries, loads and processes in real time: charge the battery ahead of the forecast peak, move a batch into the least expensive imbalance window, offer flexibility when it pays. Not a report someone reads in the morning, but a system that makes the best move every second: inside your rules and safety limits.
That is what we mean by "operational AI for energy-constrained industry": software that turns the physical limit of your grid connection into a daily cost advantage, instead of a dashboard that merely makes the problem visible.
This approach connects to our broader work in AI for the energy sector and predictive maintenance.
We work in three fixed steps with transparent pricing, and you own all the code and models we deliver:
We measure your load profile, peaks and connection, and deliver a prioritised list: which lever pays off most, with projected savings and payback.
A working control loop on your own metering and process data, in weeks, proving the peaks actually flatten before you invest big.
A full system, integrated with your EMS, battery and processes, in production and maintained.
Every engagement is checked against EU AI Act risk tiers and GDPR from day one. Compliance is built in, not bolted on.
We speak in mechanisms, not promises: the gain comes from avoiding peak tariffs and overrun penalties, from shifting to less expensive hours, and from being paid for offered flexibility. How much that returns depends on your load profile, your connection and your processes, which is why every engagement starts with an audit that makes the expected saving per lever concrete.
Our prices are fixed: €2,500 audit / €20,000 Production-ready MVP / from €50,000 production (excl. VAT). See our pricing.
Take a production site with 2 MW of contracted capacity that peaks at 2.4 MW twice a day: at the morning start-up and when the cooling compressors coincide with a batch. Assume the grid operator turns that overrun into a higher contracted value, and that reinforcement to 2.5 MW sits three years down the waiting list. These are assumptions to show the mechanism, not a client result. The audit replaces them with your own quarter-hour values.
A 500 kW, 1 MWh battery absorbs those two peaks of about 45 minutes each. An EMS that starts the cooling compressors half an hour earlier and holds the batch until start-up is over keeps the peak structurally under 2 MW. Shift a further 250 kWh a day from the most expensive to the least expensive hour and, at a day-ahead spread of €60 per MWh, that is about €15 a day, or roughly €3,750 a year over 250 production days.
The biggest item in this example is not price optimisation but the connection itself: staying within 2 MW means you can expand today instead of in three years. What that is worth depends on what you do with the room. The €2,500 audit runs the numbers on your own meter data and gives the expected saving and payback per lever.
Dive into the levers and see how our energy, maintenance and software expertise come together:
Start with a free 30-minute consultation: tell us your load profile and grid problem, and we'll map the fastest route to lower peaks and cost.
Book a free consultation →