Operational research (OR) is transforming how businesses approach complex challenges and Alpro is a prime example of its impact in action.

A European leader in plant-based food production, Alpro has significantly improved its energy efficiency and production planning by applying operational research techniques. With large-scale factories and energy-intensive processes, the company needed a smarter way to manage its resources in a volatile market environment.

The Challenge

Alpro faced a multifaceted operational challenge: how to efficiently align energy consumption with production schedules, given the constraints of internal generation capabilities and external energy prices. Their factories draw power from a combination of boilers, generators, and solar panels, but also rely on external electricity markets where prices fluctuate unpredictably. At the same time, energy demand within the factories changes based on production requirements, while maintenance windows add further constraints. The result was a complex optimisation problem—one that required a robust solution capable of integrating energy generation, market dynamics, factory loads, and operational constraints into a single decision-making framework.

The Solution: A Two-Model Approach

Alpro partnered with GAMS Software to develop a decision-support system to solve its energy problem. The team worked with Alpro's planners to turn operational needs and data into a mathematical framework that could find the best energy solutions. This system, which included two custom optimisation models, was integrated into a user-friendly, web-based interface that also allowed for conditional bidding on the day-ahead electricity market.

  • Model 1: Day-Ahead Market Bidding This model manages the crucial task of bidding in the electricity day-ahead market. It automatically gathers data from various sources, including spot price forecasts, solar energy projections, factory demand, and maintenance schedules. The model then generates conditional bids, determining if Alpro should buy or sell energy and under what conditions. For example, the model might suggest bidding to buy 0.8 MW if the cleared spot price is below €30/MWh, not buy or sell for prices between €30 and €69.99/MWh, and offer to sell 0.8 MW if the price is above €70/MWh. This automation saves Alpro valuable time and ensures energy planning is based on up-to-date information, giving them a competitive edge.
  • Model 2: Asset Load Optimisation Once the day-ahead bids are accepted, the second model takes over. It uses that information to optimise the load of each of Alpro’s energy-generating assets. This is done on a 15-minute interval basis, taking into account the factory’s energy consumption prediction and the expected production from its solar plant. The output is a detailed schedule of how to operate each asset over time, which is automatically passed to Alpro's control systems. This ensures constant, optimal energy management with minimal manual effort, and crucially, allows Alpro to maximise the use of its renewable solar energy generation.

The Impact

Since deploying the models, Alpro has seen impressive results. The energy management process is now three times faster, leading to a 25% reduction in operational costs. Having a more efficient energy system also improves Alpro's competitiveness in the electricity market, while supporting its sustainability goals. The Alpro energy team stated the new solution helps employees control energy assets in a fraction of the time, allowing them to maximise renewable energy from their solar panels.

The project's success, as highlighted by Justine Broihan at the EURO 2025 conference, hinged on three crucial lessons: vital collaboration between operational teams and technical experts; essential iterative development with continuous feedback; and critical change management, which required strong cultural buy-in for automation. The project proved that OR is an invaluable tool for improving a company's performance, efficiency, and sustainability.