The increasing penetration of variable renewable energy resources and new demands have significantly heightened the need for flexibility in power systems. Data centres present a unique opportunity to enhance power system flexibility due to their substantial yet controllable energy consumption and advanced technological capabilities. This paper provides an in-depth analysis of the potential role of data centres in improving power system flexibility. Initially, the flexibility requirements of modern power systems are defined, followed by an exploration of the flexibility assets and operational flexibility capabilities of data centres. Then, the flexibility capacities of data centres are examined, and the opportunities and benefits of leveraging this flexibility are explored, supported by case studies illustrating real-world examples. This paper underscores the vital role of data centres in the evolving energy landscape. In particular, the analysis reveals that data centres have a high potential to address the increasing flexibility requirements driven by the integration of renewable energy and the transition towards net-zero emission goals. Moreover, the findings emphasise key challenges, including ensuring Quality of Service (QoS) and adherence to Service Level Agreements (SLA), the need for further legislative development to facilitate data centres’ participation in energy markets and the provision of ancillary services, as regulatory frameworks differ across regions and variations exist in energy market structures. The findings provide actionable insights for policymakers, industry stakeholders and data centre operators, demonstrating how data centres enhance the stability, flexibility and efficiency of power systems

This study presents a detailed performance evaluation of a graphene-enhanced thermal interface material (TIM) conducted at RISE ICE Data Center. Tests were performed on Open Compute Project (OCP) Leopard servers using three different TIMs: conventional thermal paste, graphene-enhanced thermal pad GT90 from SHT Smart High Tech AB (SHT), and indium foil. Three sets of experiments were conducted: (1) air cooling with default chassis fan control in a bespoke server wind tunnel, (2) air cooling with controlled, fixed fan speeds and different heatsink mounting pressures operating in the wind tunnel and (3) immersion cooling tests with two coolant flow rates at fixed inlet temperatures. Results indicate that graphene-enhanced TIM and thermal paste exhibit similar performance in experiment (1), whereas indium foil TIM tests showed the undesired effect of increased CPU temperatures. In experiment (2), servers equipped with graphene-enhanced TIM showed lower CPU temperatures in comparison to the servers equipped with Indium foil TIM. In experiment (3), immersion cooling resulted in lower CPU temperatures overall, with the graphene-enhanced TIM again providing lower temperatures than indium foil at a similar mounting pressure. The findings suggest that the interfacial thermal conductivity and material compatibility of the GT90 TIM contribute to an improved performance in the tested immersion cooling system as well as the importance of mounting pressure. 

This study presents a detailed performance evaluation of a graphene-enhanced thermal interface material (TIM) conducted at RISE ICE Data Center. Tests were performed on Open Compute Project (OCP) Leopard servers using three different TIMs: conventional thermal paste, graphene-enhanced thermal pad GT90 from SHT Smart High Tech AB (SHT), and indium foil. Three sets of experiments were conducted: (1) air cooling with default chassis fan control in a bespoke server wind tunnel, (2) air cooling with controlled, fixed fan speeds and different heatsink mounting pressures operating in the wind tunnel and (3) immersion cooling tests with two coolant flow rates at fixed inlet temperatures. Results indicate that graphene-enhanced TIM and thermal paste exhibit similar performance in experiment (1), whereas indium foil TIM tests showed the undesired effect of increased CPU temperatures. In experiment (2), servers equipped with graphene-enhanced TIM showed lower CPU temperatures in comparison to the servers equipped with Indium foil TIM. In experiment (3), immersion cooling resulted in lower CPU temperatures overall, with the graphene-enhanced TIM again providing lower temperatures than indium foil at a similar mounting pressure. The findings suggest that the interfacial thermal conductivity and material compatibility of the GT90 TIM contribute to an improved performance in the tested immersion cooling system as well as the importance of mounting pressure.

The energy sector is essential in the transition to a more sustainable future, and renewable energies will play a key role in achieving this. It is also a sector in which the circular economy presents an opportunity for the utilisation of other resources and residual energy flows. This study examines the environmental and social performance of innovative energy technologies (which contribute to the circularity of resources) implemented in a demonstrator site in Luleå (Sweden). The demo-site collected excess heat from a data centre to cogenerate energy, combining the waste heat with fuel cells that use biogas derived from waste, meeting part of its electrical demand and supplying thermal energy to an existing district heating network. Following a cradle-to-gate approach, an environmental and a social life cycle assessment were developed to compare two scenarios: a baseline scenario reflecting current energy supply methods and the WEDISTRICT scenario, which considers the application of different renewable and circular technologies. The findings indicate that transitioning to renewable energy sources significantly reduces environmental impacts in seven of the eight assessed impact categories. Specifically, the study showed a 48% reduction in climate change impact per kWh generated. Additionally, the WEDISTRICT scenario, accounting for avoided burdens, prevented 0.21 kg CO2 eq per kWh auto-consumed. From the social perspective, the WEDISTRICT scenario demonstrated improvement in employment conditions within the worker and local community categories, product satisfaction within the society category, and fair competition within the value chain category. Projects like WEDISTRICT demonstrate the circularity options of the energy sector, the utilisation of resources and residual energy flows, and that these lead to environmental and social improvements throughout the entire life cycle, not just during the operation phase. 

The optimal utilisation of distribution grids requires the proactive management of volatilities caused by mixed energy resources installed into different grid levels, such as buildings, energy communities (ECs) and substations. In this context, proactive control based on predictions for energy demand and generation is applied. The mitigation of conflicts between the stakeholders' objectives is the main challenge for the control of centralized and distributed energy resources. In this paper, a bi-level approach is proposed for the control of stationary battery energy storage systems (SBES) supporting the local distribution system operator (DSO) at the transformer level, as well as distributed energy resources (DERs) operated by end customers, i.e., EC-members. Model predictive control (MPC)- based and hybrid approaches merging rule- and MPC-based control schemes are evaluated. Simulation studies based on a typical European low voltage (LV) feeder topology yield the performance assessment in terms of technical and economic criteria. The results show an advantage of hybrid approaches with respect to the DSO's cost savings from peak shaving. From the EC's perspective, both hybrid and MPC-based schemes can achieve effective cost savings from proactive energy management.

As the demand for data privacy and low latency grows, edge computation carried out at edge data center nodes is believed to become increasingly important for future telecom applications. Providers must consider various factors, including power consumption, thermal dynamics, and the ability to maintain high-quality service, in addition to deployment and service orchestration. This paper presents a detailed description of two different prototype edge data centers designed to investigate the power performance and thermal dynamics of edge nodes under various applied services. The prototypes were developed and tested at the RISE ICE Datacenter research facility. We present the results of power flow experiments in which input current from the grid was limited while the computational load was maintained using the energy stored in batteries. We further discuss implications for placing edge data center nodes in locations with temporal power constraints and opportunities for participation in support services at the grid level.

Emerging use-cases like smart manufacturing and smart cities pose challenges in terms of latency, which cannot be satisfied by traditional centralized infrastructure. Edge networks, which bring computational capacity closer to the users/clients, are a promising solution for supporting these critical low latency services. Different from traditional centralized networks, the edge is distributed by nature and is usually equipped with limited compute capacity. This creates a complex network to handle, subject to failures of different natures, that requires novel solutions to work in practice. To reduce complexity, edge application technology enablers, advanced infrastructure and application orchestration techniques need to be in place where AI and ML are key players.

This article investigates the problem of where to place the computation workload in an edge-cloud network topology considering the trade-off between the location-specific cost of computation and data communication. For this purpose, a Monte Carlo simulation model is defined that accounts for different workload types, their distribution across time and location, as well as correlation structure. Results confirm and quantify the intuition that optimization can be achieved by distributing a part of cloud computation to make efficient use of resources in an edge data center network, with operational energy savings of 4–6% and up to 50% reduction in its claim for cloud capacity.

Server power densities are foreseen to increase, and conventional air-cooling systems will struggle to cope with thermal demand. Single-phase immersion systems are a promising alternative to operate very intensive workload such as high-performance computing, cryptocurrencies mining or research activities. However, few companies deal with this kind of system and there is a lack of energy models that can reproduce an accurate analysis of the system behaviour. This study addresses the experimentation, data collection, and model validation of a single-phase immersion cooling system where 54 open compute project servers, each with a peak power of 400 Watts that are submerged and operated in a dielectric coolant. Results show the evolution of the thermal profile of the system under static and dynamic workloads, and it provides a correlation of server energy use under various system temperatures. The energy model is presented, validated against real data, and exploited to investigate the system response to different cooling conditions. In conclusion, the study demonstrates the validation of the energy model and supports the basis for further investigation. © 2023 The Authors

This study investigates the use of transfer learning and modular design for adapting a pretrained model to optimize energy efficiency and heat reuse in edge data centers while meeting local conditions, such as alternative heat management and hardware configurations. A Physics-Informed Data-Driven Recurrent Neural Network (PIDD RNN) is trained on a small scale-model experiment of a six-server data center to control cooling fans and maintain the exhaust chamber temperature within safe limits. The model features a hierarchical regularizing structure that reduces the degrees of freedom by connecting parameters for related modules in the system. With a RMSE value of 1.69, the PIDD RNN outperforms both a conventional RNN (RMSE: 3.18), and a State Space Model (RMSE: 2.66). We investigate how this design facilitates transfer learning when the model is fine-tuned over a few epochs to small dataset from a second set-up with a server located in a wind tunnel. The transferred model outperforms a model trained from scratch over hundreds of epochs.