Summary

The global transition toward renewable energy is not only reshaping the way electricity is generated but is also redefining the nature and behaviour of electrical loads. Traditionally, load patterns were relatively stable and predictable, largely governed by residential, commercial, and industrial demand profiles. However, with the increasing deployment of large-scale renewable energy systems, a new class of bulk electrical loads is emerging, specifically data centres driven by cloud computing and artificial intelligence applications and electrolysers for green hydrogen production. These new loads are not only large in magnitude but are also dynamic, nonlinear, and sensitive to grid conditions, presenting unique challenges to system planners.

Electrolysers, which are central to the green hydrogen economy, are being increasingly adopted across various sectors including transportation, steel manufacturing, chemical processing, and energy storage through power-to-gas systems. Their integration into the power grid introduces a new dimension of complexity, as their electrical characteristics vary widely depending on the technology used, such as alkaline, proton exchange membrane (PEM), or solid oxide electrolyser cells (SOEC). These technologies differ in terms of efficiency, control response, ramping behaviour, and ancillary power requirements. Electrolysers typically draw high levels of real power, especially when scaled to industrial capacities, and their operation can involve variable power factors, non-trivial reactive power consumption, and harmonic emissions due to the widespread use of power electronic converters. When operated flexibly to follow renewable generation, their ramp-up and ramp-down behaviour must be carefully coordinated with grid stability requirements. The choice between operating electrolysers as baseload consumers or as flexible, interruptible loads has significant implications for power system design.

Parallel to this, data centres are emerging as one of the most rapidly growing sources of electricity demand, a trend accelerated by the proliferation of digital services, edge computing, and artificial intelligence applications. Unlike traditional industrial loads, data centres exhibit a near-constant power draw throughout the day, characterized by highly nonlinear load behaviour arising from switch-mode power supplies in server farms and networking equipment. This nonlinearity makes them prone to harmonic distortion, requiring advanced power conditioning systems and harmonic filtering solutions. Moreover, data centres are extremely sensitive to voltage sags, frequency deviations, and transient disturbances, necessitating stringent power quality standards and often dedicated electrical infrastructure. The presence of significant ancillary loads, such as HVAC systems, lighting, and cooling systems, adds to their complexity, with the relative proportion of these loads being highly location-dependent. The development of hyperscale data centres, creates concentrated load pockets that can have substantial local and regional grid impacts.

This paper presents a examination of the technical characteristics of both electrolysers and data centres from a power system integration perspective. It explores their dynamic behaviour and operational interactions with variable renewable energy sources. The paper highlights the need for enhanced planning strategies, including grid codes, interconnection standards, and control frameworks that are robust enough to manage these complex and evolving load profiles. As power systems continue to decarbonize and digitize, understanding and addressing the unique challenges posed by these emerging bulk loads will be critical for maintaining grid reliability and ensuring a stable energy transition.

Additional informations

Publication type Session Materials
Reference C4_10564_2026
Publication year
Publisher CIGRE
Country India
Study committees
  • Power system technical performance (C4)
File size 967 KB
Price for non member 30 €
Price for member 30 €

Authors

YERUBANDI* VMS Prakash - Central Transmission Utility of India Limited, India; PAL Ashok - Central Transmission Utility of India Limited, India; GUPTA Manju - Central Transmission Utility of India Limited, India; BAGADIA Vikas - Central Transmission Utility of India Limited, India; DAS Partha Sarathi - Central Transmission Utility of India Limited, India; KUMAR Ajay - Central Transmission Utility of India Limited, India

Keywords

Artificial Intelligence, Data Centres, Dynamic behaviour, Electrolyser, Load, Power quality

Load Characteristics of Emerging Bulk Loads – Electrolysers and Data Centres in the Renewable Energy Era