The Data Centre, Demand Response and the Microgrid

The Microgrid discussion in the data centre sector is gathering pace.

Whether it is Microgrids, Distributed Energy Resources (DERs), co-gen, grid-interactive, islanded or integrated (decoupled or coupled) - there is growing consensus that data centres will require onsite or locally generated power delivered through microgrids.

To say microgrids will become a feature of future large data centre design would be to overlook the idea that technically most data centres of scale can be configured to operate as local microgrids.

For data centres to participate in Demand Response (DR) schemes adds a layer of complexity. Designing and building a power plant or augmenting an existing power design using assets already in place is not straightforward for any industrial application such as a large data centre that wants to be a power provider to the grid or a local community.

While some suppliers are offering turnkey microgrid solutions, it is not simply a case of choosing a microgrid and plugging it into your data centre - or plugging your data centre into it.

For the data centre operator, the starting point requires becoming more familiar with the nomenclature of the ‘behind and in front of the meter’ power systems.

To begin, first choose your favourite microgrid definitions and microgrid modes.

Microgrid modes

Typical microgrid modes are described as grid coupled (connected/integrated), islanded, (disconnected/decoupled), integrated, remote, rural, community, industrial etc. There are also many definitions. To quote one, the NREL says: “A microgrid is a group of interconnected loads and distributed energy resources that acts as a single controllable entity with respect to the grid. It can connect and disconnect from the grid to operate in grid-connected or island mode. Microgrids can include distributed energy resources such as generators, storage devices, and controllable loads. Microgrids generally must also include a control strategy to maintain, on an instantaneous basis, real and reactive power balance when the system is islanded and, over a longer time, to determine how to dispatch the resources. The control system must also identify when and how to connect/disconnect from the grid.”

For example, you may wish your data centre to be in an Islanded Microgrid scenario. That is, in normal operation, the microgrid is not connected to the main power grid and relies on local power generation. This could be engine-based, for sustainability purposes using RERs with energy storage back-up, or be a mix.

Any separation from the grid means island grids must control voltage and frequency, often while managing energy demand to loads.

Challenges identified with Island Microgrids include maintaining stability of voltage and frequency output. Local generation must also be capable of switching from “grid-following mode” to “grid-forming mode.”

Clearly demand response requires some form of flexible grid connection. An Integrated Microgrid shares many characteristics of an Islanded Microgrid such as local load, distributed energy resources (DERs) for generation and power storage, distribution management and control, with the ability to operate independently. However, the Integrated Microgrid has an additional interconnection with a large regional or national power grid.

Microgrid controllers

All microgrids require a controller. The IEEE 2030.7 Standard for Specification of Microgrid Controllers provides a basis for planning and specifying a microgrid. The IEEE describes its standard, saying: “A key element of microgrid operation is the microgrid energy management system (MEMS).”

MEMS include the control functions that define the microgrid as a system that can manage itself, operate autonomously or grid-connected, and seamlessly connect to and disconnect from the main distribution grid for the exchange of power and the supply of ancillary services. The scope of this standard is to address the functions above the component control level associated with the proper operation of the MEMS that are common to all microgrids, regardless of topology, configuration, or jurisdiction.

The IEEE standard is said to reduce microgrid complexity to two steady state (SS) operating modes and four types of transitions (T). SS1 – Steady State Grid Connected; SS2 – Stable Island.

T1 – Transition from Grid Connected to Steady State Island (Planned); T2 – Grid Connected to Steady State Island (Unplanned); T3 –Steady State Island reconnect to Grid; T4 – Black Start into Steady State Island.

Beyond the technical

Building a microgrid dedicated to a particular data centre facility or campus that can provide demand response will also entail a whole new set of planning considerations, navigating a new supply chain ecosystem and other new challenges.

It will entail becoming even more familiar with power market structures. For example, the UK Government Review of Market Arrangements (REMA) which is due to report in October will change how electricity is traded. And the EUDCA recently responded to the EU’s proposed changes to the block’s power market. It cited concerns about making PPAs easier to access for its members.

The case for investing in Microgrids is strong despite the significant challenges that exist.

For data centres looking to become power providers by establishing Demand Response schemes using microgrid-generated power, the financial rewards and sustainability benefits can be significant for those aware of the barriers to overcome.