Colloquia

Samenvatting

The increasing electrification of sectors, such as mobility and heating, is placing growing pressure on regional electricity grids. Distribution system operators (DSOs), including Coteq in the eastern Netherlands, are facing challenges in accommodating new or expanded customer connections due to limited grid capacity. This thesis investigates how energy hubs, defined as organizational collectives in which companies jointly contract a grid connection and align their energy consumption accordingly, can be used to mitigate these constraints. The objective is to identify promising customer groups whose combined energy demand profiles are complementary, thereby reducing overall peak load without requiring physical grid reinforcement.

A general methodology is developed to analyse electricity consumption profiles of large users and cluster companies with complementary load behaviour. First, recurring patterns are extracted using time-series analysis to construct a representative weekly profile. Then, a detailed search is performed to identify customer combinations with minimal peak overlap and high load flatness. The potential of demand-side flexibility is also explored, using a modular approach that classifies consumption into uncontrollable, time-shiftable, and buffer components. By applying profile steering techniques, load profiles are adjusted within technical constraints to improve coordination among users.

The model is applied to real data of the large energy users of Coteq, which are 658 customers. For 470 customers, a representative weekly profile could be established based on recurring consumption patterns. Using clustering techniques, customers were grouped based on the natural spread in their peak timings. Results show that, even without active control, clusters of complementary users can achieve peak reductions of 10–30%, particularly among small and medium-sized customers. Incorporating flexibility, based on SBI sector classifications, led to further reductions of 30–40% within selected clusters. A total peak reduction of 100.8 A was achieved across all clusters with flexibility applied, compared to 70.7 A without.

Although the overall reduction remains modest compared to the total consumption of 2890.44 A, the results demonstrate that clustering based on behavioural patterns and sector-specific flexibility is a scalable, data-driven strategy. The method enables DSOs to gain insight into potential load coordination, guiding future connection planning and improving grid utilization without compromising service quality or requiring immediate infrastructure upgrades.