Search

With the advent of the information age, the scale of data centers has developed unprecedentedly, in which the cooling system consumes a lot of energy. Therefore, it is crucial to adjust the operating parameters in real-time to maximize energy savings. In this article, the mathematical models of each component are established, and the functions of system power consumption and chip temperatures are obtained for the liquid-cooled data center based on cold plates. The method proposed in this article has good accuracy through experimental verification and can be used for the optimization of operating parameters. The server chips in data centers need to be maintained within a safe range, so we take the minimum system power consumption as the optimization goal and the chip temperature as the constraint condition to calculate the optimal cooling tower wind volume, primary side flow rate, and secondary side flow rate under different environmental temperatures and heat loads, and fit all three for real-time optimization and regulation. The results show that the intelligent control proposed in this article can save 42.7% energy and reduce PUE to 1.16 under variable heat load, and save 30.6% energy and improve PUE by 4.3% under variable environmental temperature. The intelligent control method described in this article provides guidance for real-time optimization in data centers. • The models of each equipment in the liquid-cooled data center based on cold plates were established. • The effect of three operating parameters on system power consumption and chip temperature are analyzed. • Operating parameters of the system for different ambient temperatures and thermal loads are optimized in real time. • The energy consumption of the cooling system is reduced and PUE of data center is improved. [ABSTRACT FROM AUTHOR]

As the demand for Internet services continues to rise, large-scale Internet companies have expanded their data center infrastructure across multiple global regions. The inherent spatial-temporal flexibility in scheduling workloads allows data center clusters to dynamically adjust their power consumption without overloading the transmission network. In this work, we introduce a novel approach for treating a geo-distributed data center cluster as a non-wire alternative resource and explore its potential role in the energy balancing market. Our methodology manages a geographically distributed data center cluster as a balance-responsible entity by employing optimal workload scheduling that considers its inherent spatial-temporal flexibility and market arbitrage opportunities. To achieve this, we develop a two-layer stochastic receding horizon control optimization algorithm to formulate a strategic bidding strategy for the data center cluster. To validate the effectiveness of our proposed approach, we conduct numerical experiments using real-world data center production traces and data from the Nordic balancing market, demonstrating its practical applicability and performance. • Explored spatial-temporal flexibility of data center cluster as non-wire alternative. • Applied a two-layer SRHC algorithm for online biddings and workload schedules. • Applied real-world data traces to illustrate the effectiveness and accuracy. • Real-world data showed the scheme's impact on cost reduction in balancing markets. [ABSTRACT FROM AUTHOR]

Data centers consume large amounts of electricity and water specifically for cooling purposes. Integrating free cooling technologies is an essential part of broad efforts to reduce data centers' energy, carbon, and water footprints. Adding a new technique to the existing portfolio of free cooling technologies for data centers, we present how all-day passive radiative cooling can reduce energy and water consumptions of data centers. All-day passive radiative cooling is a free cooling technique, in which sky-facing surfaces with high solar reflectivity and midinfrared emissivity spontaneously reject heat into the sky, without consuming electricity and evaporating water. Using custom-built physics-based models, we simulate three configurations that integrate all-day radiative free cooling into data centers. We show that data center cooling energy and water consumptions differ substantially depending on the integration configuration. Particularly, direct free cooling of data center return air gives the highest energy saving with an annual average of around 20.0% under the tropical climate of Singapore which hosts a large number of data centers, whereas free cooling of compressed refrigerant prior to the chiller condenser gives the highest water saving with an annual average of around 84.0%. These results offer multiple viable options for integrating radiative free cooling into data center cooling systems to simultaneously achieve water and energy savings. • Radiative free cooling is added to the existing portfolio of free cooling technologies for data centers. • Detailed physics-based models of a data center cooling system were built. • The effects of weather and operation conditions on the cooling system performance were investigated. • Direct free cooling of data hall return air yields the most energy saving with an annual average of around 20.0%. • Free cooling of compressed refrigerant of the chiller cycle gives the most water saving with an annual average of 84.0%. [ABSTRACT FROM AUTHOR]

Data centers (DCs) are gradually becoming one of the world's largest energy consumers. Owing to their spatial-temporal transferable characteristics, DCs show great potential in demand response (DR) programs. Considering the difference in task delay sensitivity (TDS), an elaborate DR model of DCs is proposed in this work. Furthermore, to ensure the TDS requirement and relieve the data transmission pressure, TDS-based unit migration cost matrices are designed in this work. Given that most DR dispatch for DCs is conducted relying on a central dispatching center, which imposes critical requests for the reliability of the dispatching system. In this work, we establish a novel 4-level joint optimal dispatch model for the DC operator, where dispatching centers installed at each level integrate district information from bottom to top and conduct optimal task dispatch from top to bottom. Moreover, an innovative dispatch autonomy mechanism is designed to realize district autonomy in the case of dispatching center breakdowns. Finally, simulation results prove that the proposed 4-level joint optimal dispatch model has a positive effect on enhancing renewable energy accommodation and improving the economic benefits of the DC operator. More specifically, by applying the proposed 4-level joint optimal dispatch model, the DCO's total cost is decreased by 32.25% in comparison with the result of no DR participation. It also proves that our designed dispatch autonomy mechanism can achieve district autonomy under the breakdown of dispatching centers at any level. • Propose a novel 4-layer joint optimal dispatch model for data center operator • Establish an elaborate demand response model of data centers based on task delay sensitivity • Design an innovative dispatch autonomy mechanism to realize district autonomy [ABSTRACT FROM AUTHOR]

The data usage surge drives greater data center demand, amplifying global CO 2 emissions. Mitigating climate change necessitates reducing data center CO 2 emissions. Reusing waste heat from data centers offers a potential energy efficiency boost and environmental impact reduction. This study utilizes liquid cooling technology to raise waste heat temperature for building space heating and introduces the concept of 'data furnaces,' where data centers directly supply waste heat to heat buildings on-site, reducing district heating consumption and lowering CO 2 emissions. Efficiently designing a heat recovery heat exchanger system that accounts for both heat rejection and cooling sides of a liquid cooling system is crucial for achieving complete heat recovery without using heat pump, a commonly overlooked aspect in existing literature. To address this issue, we propose two heat exchanger schemes: connecting the building space heating network to the secondary side (Scheme 1) and the primary side (Scheme 2) of the cooling distribution unit. Implementing these innovations leads to the elimination of dependence on a heat pump, substantially cutting energy and CO 2 emissions. Using TRNSYS software, we develop, model, and compare waste heat recovery schemes to curb district heating consumption and CO 2 emissions. To demonstrate broad implications of the proposed approaches for energy efficiency and sustainability in the data centers and building space heating networks, a showcase study examines constant 25 kW waste heat from a direct-to-chip liquid-cooled rack in an office building with 285.7 MWh annual space heating demand. A novel waste heat recovery rate relationship graph is created to assist system design, uncovering an unexpected result in Scheme 2: waste heat recovery decreases as outdoor temperature falls. In contrast, Scheme 1 maintains a stable waste heat recovery rate around 25 kW, regardless of outdoor temperature fluctuations. As a result, Scheme 1 reuses 155.2 MWh of waste heat annually compared to 138 MWh for Scheme 2. Schemes 1 and 2 yield annual electricity savings of 2290.5 kWh and 905.2 kWh, respectively, for the cooling system. Both schemes achieve profitability within a year through a 25-year life cycle analysis (LCC) and substantially reduce CO 2 emissions, with Scheme 1 saving 291,996 kgCO 2 and Scheme 2 saving 258,192 kgCO 2. The study addresses critical gaps in existing literature by emphasizes LCC. The proposed heat exchanger designs represent pioneering solutions for optimizing waste heat recovery, particularly in challenging climates. New findings offer substantial benefits to both liquid-cooled and air-cooled facilities, making significant contributions to achieve carbon neutrality in data center operations. • Non-heat pump data center waste heat recovery for space heating is studied. • Two schemes, recovering from both sides of cooling distribution unit, are compared. • The two schemes achieve payback in under a year due to the elimination of heat pump. • Waste heat recovery on the secondary side outperforms that on the primary side. • Novel relationship graphs facilitate data center waste heat recovery system design. [ABSTRACT FROM AUTHOR]

With the prevalence of cloud computing applications, cloud data centers (CDCs) are proliferating around the world. However, in the context of hybrid energy‑carbon markets environment, the high energy costs and environmental expenses caused by CDC operation are pressing CDC owners to restructure the future development of CDC in a more economical and low-carbon manner. The potential spatio-temporal transferability and reducibility of workloads provides CDCs with significant flexibility in their operations and thus may interact with the power grid as active demand users. Nonetheless, other than private data centers, public CDCs have no direct control over the workloads submitted by terminal cloud users. As such, this paper presents a bi-level model for CDC allocation planning, so as to incorporate cloud service-demand response (CS-DR) from the perspective of a hybrid electricity-carbon market. The upper level pertains to multi-domain resource collaborative planning model, which determines the optimum siting and sizing of CDCs as well as the incentive design for CS-DR program, with the objective of maximizing the total expected benefits of CDC. The lower level models correspond to the market clearing (electricity-carbon tariff model) of Independent system operator (ISO) and the decision-making of cloud users regarding CS-DR participation. The proposed model belongs to a bi-level mixed integer nonlinear programming problem with two non-convex lower levels, which can be intractable in mathematics. To solve such difficult problem, a hybrid solution method combining multiple linearization techniques and reformulation and decomposition (R&D) strategy based on column-and-constraint generation (C&CG) algorithm is developed. The proposed model is demonstrated on a modified IEEE 30-bus test case, and the simulation results verified the effectiveness of the proposed approach. • A bi-level model to allow cost-effective and decarbonizing planning of CDC under the electricity-carbon market is proposed. • Modelling of CDC from the both physical and information aspects. • An incentive-based approach to promote cloud users in the provision of flexibility is developed. • A hybrid solution method combining multiple linearization and decomposition and reformulation techniques is developed. • Effectiveness of the proposed approach is verified via comparative analysis. [ABSTRACT FROM AUTHOR]

Battery energy storage systems (BESSs) are widely utilized in various applications, e.g. electric vehicles, microgrids, and data centres. However, the structure of multiple cell/module/pack BESSs causes a battery imbalance problem that severely affects BESS reliability, capacity utilization, and battery lifespan. The available balance schemes introduce extra equalizers and suffer from slow balance speed due to the equalizer limits. To tackle this issue, a modular reconfigurable BESS (MR-BESS) topology is introduced in this paper, for which a fast battery balance method is proposed. This combination provides reconfiguration flexibility and fault tolerance capability without the need for any extra components, such as equalizers. Experimental results were performed on a scale-down prototype to verify the operation of the proposed reconfigurable BESS topology and the effectiveness of the fast battery balance method. • A novel modular reconfigurable BESS (MR-BESS) topology is proposed. • An equalizer-free active battery balance method for proposed topology is proposed. • A control algorithm for balance procedure that realizes fast balance speed. [ABSTRACT FROM AUTHOR]

Since Sweden joined the EU in 1995, importing food became easier and cheaper, leading to certain parts of the country, such as Norrbotten, becoming highly dependent on imports. This dependency, along with the inherent environmental impact of imports, could be significantly reduced by local farming. The environmental emissions originating from animal farming could be lowered even further by substituting the highly polluting soybean feed with, e.g., insect feed. This study examines the farming of mealworms, utilizing excess heat from a data center, part of a growing industry in Norrbotten county, as a means of alternative feedstock for animal production and a case study for industrial symbiosis. This industrial symbiosis project is in line with the EU's incentive to use other sources of protein and thus lower the EU's reliance on the import of foreign protein. Three different feeding approaches are tested, in a room heated with data center excess heat of 30 °C and at room temperature of about 20 °C. After the adult mealworms were harvested, a sample was taken to analyze their nutritional values. The results show that protein, lipid, and fiber content is 19,1 g, 12,6 g, and 2,7 g per 100 g, respectively. All amino acids except tryptophan were detected. This project concludes that it is possible to reach full-grown mealworms in about 8 weeks, which is about half the time stated in the literature. • Data center excess heat was utilized for mealworm farming. • 10 kg of mealworm was farmed with different temperatures, feed schedules and access to water. • The growth in length and weight was analyzed. • The nutrition composition was analyzed. • By using data center excess heat the production time of mealworms could be halved. [ABSTRACT FROM AUTHOR]

To better make use of the waste heat in distributed data centers, a new design of integrated waste heat recovery system that can meet different needs including heating for users and data center cooling is introduced. The system can switch among different operation modes including waste heat recovery mode, natural cooling mode, composite cooling mode and refrigeration mode based on the ambient temperature. The integrated waste heat recovery system is experimented in the enthalpy-difference laboratory, and the experimental results show well consistency compared with the modeling results. The operation strategies to switch among different modes are determined based on the experimental and modeling results. The paper introduced PUE (Power Usage Effectiveness) and ERE (Energy Reuse effectiveness) to comprehensively evaluate the cooling and energy reutilization effectiveness of the system. According to the evaluation result, the integrated waste heat recovery system has well performed energy effectiveness, environmental and economic effect. The ERE values of using the integrated waste heat recovery system in Chinese typical climate cities including Harbin, Beijing, Xi'an, Shanghai and Guangzhou are 0.87, 1.02, 1.04, 1.17 and 1.29 respectively. Using this system at 96 kW distributed data center instead of CRAC (computer room air conditioning) system can save 152.7 MWh electricity, 56.4 tons of standard coal, and 244.7 tons of CO 2 emission in Beijing annually. The payback periods of this new system in different climate regions in China are all within 3 years. • An integrated data center cooling and waste heat recovery system is proposed. • Operation performance of the system is verified by experiment and modeling. • Energy Reuse Factor in Beijing is 1.02, and payback period is within 3 years. [ABSTRACT FROM AUTHOR]

Multiple geographically dispersed data center buildings have been increasingly positioned in renewable-rich areas due to their high energy consumption. This paper proposes an optimal multi-energy portfolio for zero carbon data center buildings in the presence of proactive demand response programs. In the data center buildings, the complementarities of hydro-solar-wind hybrid renewable energy sources (RESs) are fully exploited via multi-energy conversion and storage devices, which are modeled as 100% renewable energy hubs for zero carbon multi-energy supplies. In order to exploit the hub-internal operational dispatchability and flexibility, a comprehensive energy consumption model of data center buildings is proposed to facilitate the demand response in terms of available serves, while the variable speed pumped storage (VSPS) is modeled to capture its dynamic operational characteristics. The energy‑carbon production, conversion, storage, and consumption are formulated as an energy‑carbon matrix, which are solved via a two-stage method to obtain the optimal portfolio. Case studies on southwest China data center buildings are implemented to demonstrate the effectiveness and superiority of the proposed methodology on cost-effective accommodation of RESs. Simulations results show that the system portfolio cost can be reduced by at most 17.8% with a higher operational flexibility. • A cost-effective data center building is proposed to accommodate zero carbon targets. • Hydro-solar-wind 100% renewables are formulated for multi-energy supplies. • VSPS and data center energy consumption are modeled to exploit inherent flexibility. • The energy-carbon production, conversion, storage, and consumption are optimized. [ABSTRACT FROM AUTHOR]

Recently, waste heat recovery has enabled data centers to serve as energy prosumers with flexible features that play essential roles in the coordination operation of integrated electricity–heat systems (IEHSs). The energy hub, as a critical component of IEHSs, provides a promising opportunity to improve energy efficiency. In this paper, the decentralized coordination optimization for IEHSs with data centers is explored to protect the confidential information of different entities. First, a novel data center based energy hub (DCEH) model is developed, where energy conversion, consumption and storage present high flexibility. Especially, in addition to heat recovery, several other controllable operational characteristics of a single data center are considered, involving servers, workloads and indoor temperature. Then, the coordination optimization model of IEHSs with DCEH (IEHSs-DCEH) is constructed by incorporating energy consumption cost and carbon emission, where the energy networks of IEHSs are considered. Moreover, a learning-aided relaxed alternating direction method of multipliers (LR-ADMM) algorithm is proposed to solve the dispatching model of IEHSs-DCEH considering communication packet loss. The proposed LR-ADMM algorithm embeds a momentum extrapolation based prediction technique, which can obtain the predicted value of missing boundary information without adding computational burden, even in continuous packet losses. Simulation results demonstrate that the developed coordination dispatching model can achieve higher economic and environmental benefits than the reference one that ignores data centers' flexibility. Additionally, the proposed LR-ADMM algorithm with proper prediction factors exhibits a faster convergence rate and robustness hedging against packet losses compared to the ADMM and relaxed ADMM approaches. • A highly flexible data center based energy hub model is developed. • Impacts of the flexibility in the data center based energy hub are investigated. • The coordination of integrated energy systems with data centers is optimized. • A learning-aided decentralized algorithm is proposed hedging against packet loss. [ABSTRACT FROM AUTHOR]

• An innovative dew point cooling system was designed, constructed, installed and real live tested in live data centre. • The system can provide continuously stable cooling air and maintain the required temperature and humidity in data centre. • The average COP of 29.7 was achieved during the testing period. • 90 % energy saving can be achieved for data centre cooling. • Suitability, stability and reliability of the system used in data centre environment were proved. This paper presents the development and application of a super performance dew point cooling technology for data centres. The novel super performance dew point cooler showed considerably improved energy saving and carbon reduction for data centre cooling. The innovations of this technology are built upon a series of technological breakthroughs including, a novel hybrid flat/corrugated heat and mass exchanging sheets, an innovative highly water absorptive and diffusive wet-material for the sheets which enable an intermittent water supply with well-tuned water pressure and flow rate, and the optimised fan configurations. Following a list of fundamental research including theoretical, numerical and lab experimental testing of a small scale prototype system, a specialist 100 kW rated data centre dew point cooling system was dedicated designed, constructed, installed and real life tested in an operational live data centre environment, i.e., Maritime Data Centre at Hull (UK) to investigate its dynamic performance, suitability and stability for application in operational data centre environment conditions. During the testing period, the system showed its reliability and capability to remove a tremendous amount of heat dissipated from the IT equipment and maintain an adequate space temperature in the operational live data centre. The dynamic data collection and analysis during the continuous testing and monitoring period showed the average COP of 29.7 with the maximum COP of 48.3. Compared to the existing traditional vapor compression air conditioning system in the data centre, the energy saving using the super performance dew point cooling system is around 90 %. The work presented in this paper include detailed innovation aspects of the technology and the system operation, as well as the established bridging knowledges, methodology and technical procedure for bringing this new technology into real life operation which involve in data centre survey, optimum design and modularization of the specialist cooling system for data centre application, proven system installation, operating method and cooling air management for data centre as well as the assurance of the continuous sufficient cooling supply to the data centre. [ABSTRACT FROM AUTHOR]

• The airflow management technologies in data centers are classified as long and short distance in the present review. • Hot-air recirculation, cold-air bypass, leakages, and airflow distributions are the most critical factors. • Strategies on improving airflow uniformity and preventing bypass and hot-air recirculation are reviewed. • Computational fluid dynamics plays essential role but requires prior calibration before further implementation. This study provides a review upon airflow management in data centers. Based on the available airflow path, cooling systems in data centers are categorized as long-distance cooling or short-distance cooling systems. Investigations on airflow management include tests in real data center or in simulated data center. Besides, the computational fluid dynamics (CFD) had been widely employed upon the thermal and airflow management of the data centers. For the long-distance cooling system, the airflow management normally adopts raised-floor configuration and hot/cold aisle arrangement. The major problems in airflow management include hot-air recirculation, cold-air bypass, leakages, over-provisioned and under-provisioned air supply, and airflow/temperature non-uniformity. The aforementioned effects often interact with the geometry layout of the data center. Related literatures regarding the effect of plenum depth, perforated tiles, enhanced facility such as induced bypass fans, infrastructure layout, aisle containment and leakage are discussed and compared. In addition, studies on the overhead air supply method are also examined and compared with the raised-floor ones. For the short-distance cooling system, the effect of server layout and heat exchanger layout concerning the airflow uniformity are investigated. It is found that the appropriate management of the original design into centralized server layout can ease the mal-distribution of airflow into the severs by 30%. This review aims to emphasize the criteria of implementing airflow management to data centers that serve as a reference guide for energy saving in data center as far as airflow arrangement is concerned. Moreover, some recommended future research efforts are also addressed. [ABSTRACT FROM AUTHOR]

Highlights • Data centers have a lot of excess capacity due to their redundant design. • This capacity can be used to provide ancillary services, such as primary regulation. • UPS systems are technically and economically viable in primary regulation. • Additional stress to battery systems is within battery specifications. • Participation to primary regulation can create significant revenue for data centers. Abstract Data centers have a significant potential to become a major source of flexibility in smart girds. They consume currently roughly 3% of all the electricity produced globally and are expected to only increase their consumption as the world becomes more connected and digitalized. As data centers are required to operate without any interruptions, they use power protection systems and energy storages. This paper investigates the technical and economic feasibility of dual-purposing these power protection systems, the uninterruptible power supplies, and their batteries in data centers to perform primary frequency regulation services. While the topic of data centers and demand response has been extensively covered in the current scientific literature, the focus has been on the demand response enabled by server workload shifting or hardware-enabled peak shaving. Based on an extensive literature review, there is a knowledge gap in the literature concerning primary frequency regulation and dynamic response enabled by modern power electronics systems in data centers. In this paper, this knowledge gap is bridged by suggesting a novel approach of taking advantage of the bidirectional operations capabilities of the uninterruptible power supply systems, thereby enabling them to provide dynamic power response from their battery systems. The feasibility of this approach is examined with the proposed method, which includes (1) an analysis of the required energy for primary regulation and the availability of this energy in a typical data center, (2) a simulation of activation events and their impact on the service life of the battery systems, (3) reaction speed and reliability considerations of the operations, and (4) an economic feasibility and balancing market analysis. The results show that as primary frequency regulation is an energy nonintensive service and data center battery systems are by design oversized for redundancy reasons, typical data centers have more than ample amounts of energy to participate in the primary regulation without jeopardizing their own processes. The results also show that by maintaining reasonable levels of usage, the battery systems can be operated within their specifications, and the demand response operations will not cause premature aging of the battery systems. The reaction speed of the power electronics is found to be very high and easily meet the current market requirements. While the achievable revenue from the primary regulation service is small compared for example with the electricity costs of the data centers, it is still significant as there is little to no impact on the daily business of the data centers. [ABSTRACT FROM AUTHOR]

• A job scheduling model for low-carbon data center operation is established. • Practical challenges such as job dependency and heterogeneity are considered. • A reinforcement learning algorithm is proposed to obtain online scheduling policy. • The effectiveness of proposed scheme is verified with real-world scenarios. As the digitalization of the economy and society accelerates, the enormous and fast-growing energy consumption of data centers is becoming a global concern. With the unique power consumption flexibility introduced by computing job scheduling, data centers could play an important role in enhancing the capability to integrate renewable generation as a demand-side resource. In this paper, we propose an online job scheduling scheme for low-carbon data center operation from an information and energy nexus perspective. We formulate the job scheduling problem as a Markov decision process in which job dependencies, job heterogeneity, and quality of service are considered comprehensively. To address the challenges of large-scale heterogeneous computing jobs, we propose a deep reinforcement learning-based approach to solve the energy-aware scheduling problem and achieve an optimal online policy. The case study results based on real-world data illustrate that the proposed scheme can effectively reduce the carbon footprint and energy cost of a data center while maintaining the quality of service for cloud products. [ABSTRACT FROM AUTHOR]

Meshed microgrids have been used in a plethora of specialised applications that demand increased system resilience, from data centres to the international space station. When resilience maximisation is the desideratum, topology design is the fundamental factor determining the overall system performance. Very few published papers on this problem are found in literature and they take into consideration partial available topologies, or they fail to quantify resilience or assess microgrid stability. To address these gaps on microgrid topology planning (MTP), this paper proposes a holistic optimal topology design framework, comprised of six stages: (a) graph generator to extract all possible connected, non-isomorphic networks for a given number of nodes, (b) optimal asset positioning upon each generated graph using mixed-integer linear programming, (c) small-signal stability assessment and (d) post-outage flow matrix calculation for each microgrid to quantify repercussions from fault/damage of all interconnecting lines, (e) quantification of resilience metrics originally applied to distribution systems to assess vulnerability and criticality of each line, along with classic graph-theoretical indicators and finally, (f) aggregation of said metrics to rank the produced microgrid topologies. The proposed methodology is evaluated through detailed discrete simulations to assess its efficacy and the dynamic stability of the optimal microgrid topology. [Display omitted] • A novel 5-step resilient topology design for isolated DC microgrids is presented. • Assets are optimally sized and positioned on buses with the focus on resilience. • Small-signal stability check is included to identify inherent topology instability. • Microgrid resilience is quantified by metrics deriving from distribution grids. • The superior performance of the optimal topology is proven under worst-case outages. [ABSTRACT FROM AUTHOR]

This paper proposed an optimal planning method for regional integrated electricity–heat systems with data centers (DC-RIEHS) considering wind power uncertainty to reduce economic costs and improve system flexibility. First, a novel data center model is developed, where various flexible operating characteristics of the data center are modeled in detail, including thermal inertia of indoor air, rescheduling of delay-tolerant workloads, dynamic voltage frequency scaling technique of servers and heat recovery. The heat recovery system is specially built based on the electric cooling system of the studied data center and heat pump, which connects electric power systems and district heating systems. On this basis, a three-stage distributionally robust chance-constrained (DRCC) planning model of DC-RIEHS is proposed. It is converted into a tractable single-stage mixed-integer conic model by applying affine decision rules to hedge against wind power uncertainty. The proposed planning method determines the optimal capacities of energy equipment and the scheduling decisions of representative days considering wind power uncertainty. Simulation results demonstrate that the flexible resources in a data center can significantly reduce system costs and actively participate in coping with wind power uncertainty. Compared with the chance-constrained optimization method under Gaussian distribution, the proposed DRCC method exhibits better out-of-sample performance in terms of violation probability and system operation results under the wind uncertainty with spatial–temporal correlation. • A data center model is presented by considering various flexible characteristics. • The critical equipment is planned for integrated energy systems with data centers. • Distributionally robust optimization is used to deal with wind power uncertainty. • Impacts of the flexible resources in the data center are investigated. [ABSTRACT FROM AUTHOR]

• A single-phase immersion cooling system with LHTES devices is proposed. • An innovative LHTES device with palmate leaf-shaped fins is designed by bionic. • Heat charge and discharge durations reduce by 21.0% and 38.2%, respectively. • Performance of biomimetic LHTES devices is less influenced by inclination angles. • Flow rate of 20 L/min and large heat-transfer temperature difference are suggested. Improving the power usage effectiveness of data centers has become a significant challenge for carbon peaking and carbon neutrality. To address this, here we propose a single-phase immersion cooling system with latent heat thermal energy storage (LHTES) devices to recover waste heat. Furthermore, an innovative LHTES device with palmate leaf-shaped fins is designed by bionic techniques. The phase change behavior and thermal transport patterns of biomimetic and traditional LHTES devices are investigated, and the effects of the arrangement and working conditions are revealed. Additionally, the proposed solution's application prospect and economy are presented. The results indicate that the performance of palmate leaf-shaped finned LHTES devices is significantly enhanced compared to traditional ones, whose charge and discharge durations reduce by 21.0 % and 38.2 %, respectively. Moreover, the natural convection development enhances the melting performance, but its effect on solidification processes is insignificant. Considering the trade-off between power consumption and LHTES efficiency, a horizontal biomimetic LHTES device with a flow rate of 20 L/min and a large heat-transfer temperature difference is suggested in practical engineering. Notably, the maximum reduction in annual power consumption for using biomimetic LHTES devices is 2397.4 kW·h, and the minimum payback period is only 2.0 years. [ABSTRACT FROM AUTHOR]

• Algorithm can be sensitive to algorithm setting, affecting optimality and robustness. • The discrepancy between the digitized and initial objectives cannot be ignored. • The system dynamics that can affect potential performance improvement is revealed. • Algorithms can obtain energy savings and violation reductions in some scenarios. • Actor-critic, off-policy, and model-based algorithms exhibit better performance. Deep Reinforcement Learning has been increasingly researched for Dynamic Thermal Management in Data Centers. However, existing works typically evaluate the performance of algorithms on a specific task, utilizing models or data trajectories without discussing in detail their implementation feasibility and their ability to deal with diverse work scenarios. The lack of these works limits the real-world deployment of Deep Reinforcement Learning. To this end, this paper comprehensively evaluates the strengths and limitations of state-of-the-art algorithms by conducting analytical and numerical studies. The analysis is conducted in four dimensions: algorithms, tasks, system dynamics, and knowledge transfer. As an inherent property, the sensitivity to algorithms settings is first evaluated in a simulated data center model. The ability to deal with various tasks and the sensitivity to reward functions are subsequently studied. The trade-off between constraints and power savings is identified by conducting ablation experiments. Next, the performance under different work scenarios is investigated, including various equipment, workload schedules, locations, and power densities. Finally, the transferability of algorithms across tasks and scenarios is also evaluated. The results show that actor-critic, off-policy, and model-based algorithms outperform others in optimality, robustness, and transferability. They can reduce violations and achieve around 8.84% power savings in some scenarios compared to the default controller. However, deploying these algorithms in real-world systems is challenging since they are sensitive to specific hyperparameters, reward functions, and work scenarios. Constraint violations and sample efficiency are some aspects that are still unsatisfactory. This paper presents our well-structured investigations, new findings, and challenges when deploying deep reinforcement learning in Data Centers. [ABSTRACT FROM AUTHOR]

• A concept for servers with nominal power usage effectiveness less than one is presented. • A novel system combining loop heat pipe and thermoelectric generators for server cooling was proposed. • The coupling system can harvest electrical energy during cooling electronic chips. • Net electrical power more than 1 W was generated when cooling a 250 W server CPU. With the rapid increase in the power consumption of global data centers, it is becoming more and more important to reduce the power usage effectiveness (PUE) value, which is a key metric for evaluating the energy efficiencies of data centers. The previously proposed advanced thermal management methods can make PUE values very close to 1. To further reduce the PUE of the air-cooled data center, a novel looped heat pipe (LHP) and thermoelectric generator (TEG) coupled system based on a series–parallel scheme was first proposed. Attributed to the passive heat dissipation apparatus with extremely low thermal resistance and high-efficiency thermoelectric power generation devices as adopted in this design, the power output, especially the net power output, could be greatly improved compared to the power outputs achieved in previous research, and an electrical power of more than several watts was obtained while cooling a server CPU, which enable servers possibly with PUE values less than one. The results showed that, under the condition that the junction temperature was below 85 °C, by using nine 40 mm TEGs, the minimum PUE of 0.997 was achieved at a Reynolds number of 930 and a heat load of 155 W. [ABSTRACT FROM AUTHOR]

• An analysis model for dynamic available energy of UPS in IDC is proposed. • Integrated dispatching strategy of the UPS and conventional units. • A reallocation method based on dispatching error and SOC of BESS improve the optimization of SOC and the revenue of the IDC increased by about ¥5950. • Proposed strategy increases UPS utilization by about 85% and reduces the MAE and RMSE of the SOC fluctuation to 12%-19%. The internet data center (IDC) can improve the stability of power system and increase the utilization of uninterruptible power supply (UPS) with battery energy storage system (BESS) and hydrogen fuel cell (HFC) by participating in dispatch operations. This paper proposes a hierarchical dispatch strategy assisted by model predictive control (MPC) for UPS in IDC including available energy analysis, the upper-level power system dispatch strategy and the lower-level IDC dispatch strategy. The data security can be ensured through available energy analysis based on load forecasting information of IDC. Then, the upper-level dispatch model aims at increasing the utilization of UPS and the revenue of IDC via minimizing the total generation cost. Meanwhile, the lower-level dispatch method with MPC reallocates the upper-level dispatch instructions between BESS and HFC by balancing power tracking error and the state of charge of BESS. Finally, the effectiveness of the strategy is verified by several experiments in MATLAB platform. Numerical results show that the proposed strategy increases the utilization rate of the UPS by about 85% and the revenue by about ¥5950, and the SOC fluctuation is reduced to 12%-19%. [ABSTRACT FROM AUTHOR]

With the emergence of energy communities, where a number of prosumers invest in shared generation and storage, the issue of fair allocation of benefits is increasingly important. The Shapley value has attracted increasing interest for redistribution in energy settings — however, computing it exactly is intractable beyond a few dozen prosumers. In this paper, we first conduct a systematic review of the literature on the use of Shapley value in energy-related applications, as well as efforts to compute or approximate it. Next, we formalise the main methods for approximating the Shapley value in community energy settings, and propose a new one, which we call the stratified expected value approximation. To compare the performance of these methods, we design a novel method for exact Shapley value computation, which can be applied to communities of up to several hundred agents by clustering the prosumers into a smaller number of demand profiles. We perform a large-scale experimental comparison of the proposed methods, for communities of up to 200 prosumers, using large-scale, publicly available data from two large-scale energy trials in the UK (UKERC Energy Data Centre, 2017, UK Power Networks Innovation, 2021). Our analysis shows that, as the number of agents in the community increases, the relative difference to the exact Shapley value converges to under 1% for all the approximation methods considered. In particular, for most experimental scenarios, we show that there is no statistical difference between the newly proposed stratified expected value method and the existing state-of-the-art method that uses adaptive sampling (O'Brien et al., 2015), although the cost of computation for large communities is an order of magnitude lower. • Systematic study of Shapley value approximation methods for community energy applications. • New deterministic method for approximating the Shapley value for large communities. • New methodology to allow systematic comparison of different Shapley approximations. • Experimental comparison using data from two large-scale energy trials from the UK. • Redistribution very close to real Shapley values, but at much lower computation cost. [ABSTRACT FROM AUTHOR]

Electricity cost has become a critical concern of data center operations with the rapid increasing of information processing demand. Data center microgrid (DCMG) is a promising way to reduce electric energy consumption from traditional fossil fuel generators and the billing cost, by effectively utilizing local renewable energy, e.g., wind power. However, uncertainties of wind power generation and real-time workload of data center would have significant impacts on the operational efficiency of DCMG, especially when it is in the island mode. For this reason, a novel affinely adjustable policy based robust multi-objective optimization model under flexible uncertainty set is proposed in this paper, which simultaneously optimizes wind power curtailment, the operation cost, and the over-plus level of computation resource, while considering uncertainties of both the wind power and real-time workload. Through numerical simulation studies, the validity of robust multi-objective optimization model for the island operation of DCMG is verified. Besides, the effectiveness of the proposed methods, i.e., the novel affinely adjustable policy and the flexible uncertainty set, in handling uncertainties are evaluated. Compared to the conventional robust multi-objective optimization model, the proposed approach reduces the operating costs of about 10% in average while maintaining similar reliability in numerical simulations. Moreover, the complex quantitative relationship among these multiple objectives is further investigated. Simulation results indicate the minimization of wind power curtailment and over-plus level of computation resource increases about 25% of the operation cost. These quantitative relationships can well support the decision making of DCMG operation management. • A novel robust multi-objective scheduling model is proposed to study the relationship between the multi-objectives of the islanded data center microgrid. • The flexible uncertainty set and the affine adjustable policy are developed to effectively handle the uncertainty factors, i.e., less conservative Pareto optimal solutions could be obtained. • Numerical simulation studies are executed to verify the validity of robust multi-objective optimization model and the effectiveness of the developed novel affinely adjustable policy as well as the flexible uncertainty set. [ABSTRACT FROM AUTHOR]

• An Internet data center planning model is proposed. • The role of Internet data centers in the energy transition process is focused on. • The carbon response model of Internet data centers is established. • The carbon footprint of Internet data centers is traced. • A linearized model of Internet data centers is presented. With the advent of the information era and the advance of the Internet, the electricity demand of Internet data centers (IDCs) sees rapid growth worldwide. The demand response of the IDCs is a potential resource to realize load management in smart grids. In this paper, we investigate the impact of the demand response of IDCs on the electricity network transition roadmap under carbon pricing policies. Different from other types of loads, IDCs can offer both temporal and spatial load regulation potentials. With their unique demand response characteristics, IDCs can help to smooth the output of renewable energy in the energy transition process. To this end, a two-stage energy transition framework is proposed. The proposed framework work reveals the interaction between the transmission expansion planning and the carbon response of IDCs. In the first stage, a transmission network expansion model, considering the early retirement of the coal-fired power plants (CFPPs), is proposed. The utilities are compelled to prioritize carbon-constrained infrastructure augmentation in their capital programs. In the second stage, the carbon response of the IDCs is modeled. The adjusted carbon emission flow (CEF) model is proposed to track the carbon footprint of IDCs. The operation and server expansion strategies of IDCs are optimized. The IDCs' self-elasticity to the carbon price and the cross-elasticity to the electricity price is formulated to better understand the underlying driver of carbon emissions in electricity markets. With the proposed model, the operation and expansion strategies of IDC can respond to the carbon price signals actively. The proposed method is verified on the IEEE 30-bus system. The simulation results show that the proposed method can save $0.518 billion annually. Furthermore, the carbon response of the IDCs can help the network realize energy transition smoothly and reduce the total carbon emission. [ABSTRACT FROM AUTHOR]

• A data-driven flexibility assessment scheme of internet data center is proposed by investigating the temporal shifting capability of periodic batch workloads. • The power consumption model of data center is modeled by a data-driven manner based real-world workload processing and energy consumption data. • The periodic arrival patterns and job dependencies are incorporated into the assessment model to avoid infeasible job scheduling. • The effectiveness and accuracy of the proposed scheme is illustrated with real-world production workload trace from one of the major cloud service providers. Considering its unique operational and power consumption characteristics, internet data center (IDC) has been intensively investigated as a promising candidate to provide flexibility for electric power system. In this paper, a data-driven flexibility assessment scheme for IDC is proposed by investigating the temporal shifting capability of periodic batch workloads, which are the major flexibility source in the workload scheduling and execution process. We develop a four-step assessment procedure by identifying the periodic jobs, extracting key operational patterns, mapping the power consumption with workload execution, and quantifying the flexibility associated with power system operation, all of which are established in a data-driven manner. In addition, we adopt real-world production workload trace to verify and demonstrate the effectiveness of the proposed flexibility assessment scheme. [ABSTRACT FROM AUTHOR]

• A data center energy saving method through UPS and server coordinated optimization. • Shift VM allocation to server allocation by considering the one-to-many physical configuration. • Propose a two-step VM allocation method by using BFD and genetic algorithm in sequence. • Up to 2.47% total power reduction can be achieved under the assumed case. Recent years witness a rapid increasing of data center energy consumption, among which the power losses of uninterruptible power supply (UPS) and server dominate. This paper proposes a mission-driven two-step virtual machine (VM) commitment by considering the characteristics of UPS power losses in virtual machine allocation. To formulate the optimization problem, a power model of data center is built by combining server power with power losses of rack level UPSs. Further, a two-step virtual machine allocation method is proposed, in which the best fit decreasing algorithm first groups the VMs and then the genetic algorithm searches for the optimal allocation of servers equipped with the grouped VMs. A data center with 100 servers and 100 VMs is simulated using the public data collected from PlanetLab platform. Results have verified that the total power consumption is reduced by up to 2.47%. In parallel, possible combinations need to be searched decrease by 3.35 × 10182 times in the VM allocation problem in the case with 100 servers and 100 VMs. [ABSTRACT FROM AUTHOR]

• Deep Reinforcement Learning cooling control provides energy savings potential. • Soft Actor Critic algorithm is used to reduce training time and improve data efficiency. • The approach provides additional insights into the cooling control approach by a learning agent. • DRL cooling control in large-scale data centers exhibit less energy savings than smaller data centers due to efficient infrastructure. To reduce cooling energy consumption, data centers are recommended to raise temperature setpoints of server intake. However, in tropical climates, Data Center operators are still found to be operating at lower temperatures. In this paper, we demonstrate that using a floating setpoint with a lowered temperature value for tropical climates reduces the overall energy consumption of Data Centers as opposed to raising the temperature in a static manner. We achieve this by applying a deep reinforcement learning algorithm to a hybrid data center model that was built from data collected off a highly efficient data center. This generates an optimal control strategy which minimizes the costs of energy consumption while operating within the required set of operational constraints. Following which, we evaluate the behavior of the control strategy to account for the exact sources of energy savings. The deep reinforcement learning algorithm learns by continually interacting with the built Data Center model without any prior knowledge of the Data Center. The algorithm is trained under the full-load and the part-load configuration of the Data Center. Testing results show that further energy savings of up to 3% and 5.5% (under full load and part load respectively) can be achieved with targeted cooling provisioning while operating within constraints in an already cooling-efficient Data Center. We find that while building level optimization studies of Data Centers generally improve energy efficiency, the source of energy savings is not well accounted for. Consequently, our studies show that the reduction of server fan usage and not the reduction of cooling energy consumption is the main contributor of energy savings in a deep reinforcement learning-driven data center operating in the tropics. [ABSTRACT FROM AUTHOR]

• New performance enhancement mechanism for looped heat pipe operation is proposed. • Different looped heat pipe prototypes were developed for comparison. • An air-cooled looped heat pipe with 550 W capability at horizontal position was achieved. • The nominal power usage effectiveness of the LHP module is very low. • The novel looped heat pipe module was implemented into server onboard cooling test. The rapid development of data centers has brought huge energy consumption around the world, and an effective, reliable and energy-saving cooling method is urgently demanded for data centers cooling. This paper proposed air directly cooling looped heat pipe (LHP) modules with excellent thermal performance that can be used for server cooling in data centers based on the unconventional enhancement mechanism for looped heat pipes. The proposed LHP with no compensation chamber has an auxiliary wick which connects the primary wick in the evaporator and extends into the condenser to promote the liquid replenishing. Meanwhile, another approach for reducing the sensible heat leak in the replenishing liquid was introduced with aid of an additional fin and fan assembly installed very close to the evaporator entrance. By means of the above approaches, significant improvement in maximum sustainable heat load of the LHP in horizontal direction was achieved. The improved LHP module has a maximum heat dissipation capacity of 550 W with the minimum total thermal resistance of 0.16 °C/W in horizontal direction, and with a power usage effectiveness as low as 1.02. [ABSTRACT FROM AUTHOR]

• A green and sustainable method for packing servers is investigated. • Cooling principles on cooling efficiency and packing capacity are constructed. • The problem is formulated as bin-packing models via segmenting and gridding. • Mathematical analyses and solution strategies help to pack servers efficiently. An Underwater Data Center (UDC) is an underwater vessel full of computing servers and designed with a cooling system using cold water in the ocean. A UDC vessel is composed of cabinets for computing servers, and the cabinets are finally packed into racks that facilitate the installments of the computing servers. We formulate the problem of packing cabinets into the vessel by considering the water volume required for server cooling and the placement of the cabinets of servers. Due to different patterns of racks, a set of mixed-integer linear programs were developed, and they were solved by methods with several stages. As demonstrated by experiments, the proposed models and strategies effectively solve the simulated instances generated by considering the cooling principles. The cabinets with more considerable differences in their width and height are better than single-sized cabinets because the smaller cabinets make the packing process more flexible. Moreover, multi-sized cabinet packing is significantly better than single-sized cabinet packing. Additionally, it is not always appropriate to use racks with large capacities. However, racks with multiple capacities may improve the space and energy utilization degrees when the size and the designed heat power consumption of the data center are fixed. [ABSTRACT FROM AUTHOR]

This paper addresses the technical and economic issues associated with waste heat recovery in data centers through the use of absorption cooling machines. The theoretical possibility of utilizing the heat dissipated by a server, or a number of servers, to power an absorption system, which in turn produces cooling for other servers in the data center, is investigated. For this purpose, a steady-state thermodynamic model is developed to perform energy balance and exergy analyses for a novel configuration of an on-chip two-phase cooling system and an absorption refrigeration system. This combination is created by replacing the condenser in the on-chip cooling circuit with the generator of an absorption refrigeration cycle. The performance of the developed model in simulating both LiBr–water and water–ammonia absorption cooling systems is examined through verification of the model results against the reference data available in the literature. The verification indicates the superiority of LiBr–water absorption system for data center/server operating conditions. Therefore, a LiBr–water absorption refrigeration system is modeled in the novel combined heat recovery system. For these systems it is shown that the traditional definition for the coefficient of performance (COP) is not appropriate to evaluate the performance and, in its place, introduce a new figure of merit. Through a sensitivity analysis, the effects of server waste heat quality, server coolant type, solution peak concentration, solution heat exchanger effectiveness, evaporator temperature, and operating pressures on the performance of the novel system are investigated. Finally, using the thermodynamic model and cost information provided by the absorption refrigeration industry, an economic analysis is carried out to calculate the payback period when this technology is used for data center waste heat recovery. [ABSTRACT FROM AUTHOR]

• System design of energy-saving superconducting cables from renewable energy to a 100 MW data center. • Technical design of compact multi-layer superconducting cables with the details of materials' characteristics. • Total loss of superconducting cables decreasing down to 1.7% of the loss of conventional AC/DC cables. • Economic advantages of superconducting cables over conventional cables due to the extremely low operating costs. • Promising solution using energy-saving superconducting cables for the hyper-scale data centers due to the technical and economic benefits. With the significant increase of global data center infrastructure, how to delivery electric energy to data centers in an efficient and sustainable way needs to be carefully considered. This article presents a novel study on the energy-saving superconducting cables from the renewable energy source to a 100-MW-class data center, with the comparison using the conventional AC and DC power transmissions (9 cases in total). The design of superconducting cables and AC and DC power cables are shown with the details of materials, electrical and thermal characteristics. For all these transmission schemes, both the capital investment costs and operating costs are derived, and the economic analysis is carried out with the factors of different voltage levels, transmission distances from clean energy to the data center, and operating years. Results show the transmission schemes by superconducting cable have both the technical and economic advantages over the conventional AC and DC transmission schemes (e.g., for the 10 kV 10 km case, the total energy loss of superconducting transmission is merely 1.7% of conventional AC, and after 10 years the total cost of superconducting transmission is also cheaper), which can be a promising solution of energy transmission for the high-power data centers in the future. [ABSTRACT FROM AUTHOR]

• A PEMFC-based CCHP system integrated with methanol reforming and dehumidification. • Realize the cooling and dehumidification of the data center simultaneously. • The dynamic performance under the cooling-led strategy is studied. • The temperature and humidity of the data center are controlled at 25 °C and 55–63%. For data centers with strict requirements on temperature and humidity, this study proposes a combined cooling, heating and power (CCHP) system integrated with methanol-reforming and dehumidification based on proton exchange membrane fuel cells (PEMFCs). The system consists of a methanol-reformer, preferential oxidation (PROX) reactor, PEMFC stack, adsorption chiller, desiccant air conditioning (DAC), and phase change material (PCM) heat storage tank. Taking a data center in Harbin as the research object, the dynamic response characteristics of the system are studied, and the system has good dynamic response characteristics. The CO concentration in the reformed gas is controlled by controlling the O 2 /CO ratio in the PROX. The CO concentration in the transformation gas can be controlled below 4 ppm. The cooling and dehumidification requirements of the data center are met by controlling the cooling power of the system and the air intake conditions of the data center. The results show that the cooling power of the system rises for 400 s at the beginning of the operation, and stabilizes after about 2 h, and the temperature and relative humidity of the data center can be controlled at 25 °C and 55–63%. The novel system develops an efficient way to recover low-grade waste heat of PEMFC stacks for cooling and dehumidification in data centers. [ABSTRACT FROM AUTHOR]

• The synergistic relationship between the power consumption and temperature difference in heat transfer process is revealed. • A synergy operation factor is defined to guide the practical operation optimization. • An experiment is carried out to verify the accuracy of the synergy optimization analysis. • Nearly 30.9% of the energy was saved due to the synergy optimization analysis in a real data center. Most industrial heat exchange processes are carried out between two fluids. Usually the thermal fluids are driven by pumps or fans. The power consumption of pumps and fans is quite high in the transport systems, especially in data centers. Nearly 60% of the computer room air conditioning (CRAC) system's energy consumption is consumed by pumps and fans in winter. A lot of literatures discussed the heat transfer enhancement in heat transfer process but few focused on the energy consumption. In this paper, a new synergy optimization method is proposed to save energy in heat transfer process. The heat transfer constraint equations of the heat transfer process are established. Taking the energy efficiency as the objective function, the synergistic relationship between the power consumption and temperature difference is derived analytically by using variation principle. A synergy operation factor is defined to guide the practical operation optimization of heat transfer process. The closer the synergy operation factor to 1, the higher the system energy efficiency is. An experiment of a separated heat pipe system is carried out to verify the accuracy of the synergy optimization analysis. The synergy optimization analysis is implemented in a real data center, and the cooling system can save energy by 30.9% under test conditions. [ABSTRACT FROM AUTHOR]

• Cooling solution adopts heat exchanger/evaporative water chiller on air/water sides. • Thermal control is enhanced using overhead downward flow with cold aisle containment. • Effects of supply air temperature and velocity on data center performance are studied. • A good power usage efficiency of 1.38 is achieved for planned large-scale data center. This study proposes the container data center with the featured cold aisle containment (CAC) as effective thermal control strategy. In design, the overhead downward flow system is implemented with a heat exchanger arranged right above the data center on the air side and an evaporative water chiller on the water side to form the cooling approach. The cold airflows and hot exhausts of racks are separately transported by the contained cold and hot aisles to alleviate the problem of cold and hot air mixing. The measurements of air temperature and velocity of racks are used to validate the prediction accuracy of the computational fluid dynamics (CFD) model. The performance metrics in terms of the rack cooling index (RCI), return temperature index (RTI), supply heat index (SHI) are used to examine the design effectiveness of the proposed test data center. The simulations are then extended to assess the air distribution and thermal management at varied supply air temperatures and velocities for a large-scale data center to be built in the green energy technology demonstration site of the Shalun smart green energy science city. Overall, the calculated average PUE of 1.38 for the large-scale data center is notably less than the average PUE of 1.59 from the results of 2020 data center industry survey, indicating the potential savings of cooling energy and cost. This paper demonstrates a generalized approach as an easily adaptable, cost-effective solution for data centers to be deployed in tropical and subtropical areas. [ABSTRACT FROM AUTHOR]

This elaboration presents the configuration of the cooling system of the POLCOM Data Center which utilizes commercially available components, proposes steering strategy and analyzes their functionality. The designed architecture of cooling and dedicated control system is presented to demonstrate the novelty and customization with respect to the demanding outdoor temperatures existing in the climate of the Małopolska Province. The cooling system control and operation is illustrated by a comprehensive analysis of compressor and freecooling modes. The discussion on embedded functionality of chillers is undertaken. According to features and limitations of the complex solution, the cooling system reached the annual average coefficient of performance of 8.63 in 2015 (increase of 110% compared to the year 2014), operating 65% of the time during the year in compressor mode and 35% in freecooling mode. This coefficient in compressor mode amounted to 4.39 while in freecooling mode totaled 16.50. It was proved that in the real case under consideration they generated losses in electricity consumption amounting to 557MWh per year. The real-time experimental data collected from the commercial Data Center installation are used to present a unique operation of a such complex system. • Custom controller incensed efficiency and performance of cooling system. • The freecooling mode is more energy-efficient compared to the compressor mode. • In freecooling mode, a single chiller works more effectively than two simultaneously. • The operating modes strongly correlates with external temperatures distribution. • Embedded controllers limits freecooling operating time and results in energy losses. [ABSTRACT FROM AUTHOR]

[Display omitted] • Full physics-based IT equipment CFD model is built. • Machine learning is integrated with detailed IT equipment CFD model to improve the CFD model accuracy at off-design conditions. • Novel approach for developing compact IT equipment models which can predict its power consumption, airflow demand, and exhaust air temperature is proposed. • The derived IT equipment compact model is validated with the experimental results. In most data centers, performance reliability is often ensured by setting the amount of airflow provided by the cooling units to substantially exceed that which is needed by the IT equipment. This overly conservative strategy requires additional energy expenditure, which inevitably results in a huge amount of energy being wasted by the cooling system. To eliminate adopting such wasteful policies, conducting proper management of airflow, temperature, and energy is critical. To that end, this work proposes a novel approach to developing a compact IT equipment model at off-design conditions. This model is designed to support thermal and energy management functions in data centers. The benefit of this model is that it can accurately predict not only the IT equipment power consumption, but also the amount of flowrate required for the equipment and the air temperature leaving the equipment. While the compact model's power consumption was derived as a function of CPU utilization, its flowrate demand and exhaust temperature were obtained from a dynamic detailed CFD model. Results from the compact model were validated with experiments where the maximum mismatch was found to be 5.7% in the outlet temperature field and 11.4% in flowrate. Compared to a state-of-the-art IT equipment compact model, the developed model was found to reduce the prediction error of the equipment's flowrate and outlet air temperature by up to 5.2% and 9.3 % that of the state-of-the-art IT equipment compact model, respectively. [ABSTRACT FROM AUTHOR]

• A novel spray cooled rack is developed for data center cooling in the tropics. • This new cooling system eliminates the need for high energy consuming chiller. • Microprocessor temperature of spray cooled scheme is lower than air cooled scheme. • Characterized system thermal performance over a wide range of operating parameters. • Spray cooled rack system contributes to significant energy savings. A lab-scale spray cooled rack system with a server and simulated heaters, operating with near-ambient temperature coolant of approximately 30 °C, was developed for data center cooling application. This system has significant energy saving potential as it eliminates the need for a chiller unit, which is usually necessary in conventional air-cooling schemes. The system was designed to dissipate a maximum heat load of 5.6 kW. Despite its higher operating temperature, our investigation shows that the spray cooled system leads to better cooling performance of the microprocessors as compared to conventional air-cooling scheme. A durability study for more than 40 h of cumulative intermittent operation showed no degradation in system performance when operated at high chamber temperature conditions (>35 ℃). The thermal performance of the spray cooled system and the effects of parameters such as spray nozzle flow rate, condenser fan power, condenser flow rate and heat loads were systematically characterized. It is found that the thermal performance of the simulated heaters is mainly affected by the spray nozzle flow rate, while the condenser fan power and condenser cooling water flow rate strongly affect the condensation rate and are, therefore, critical parameters for spray chamber pressure control. In summary, this work demonstrates the possibility of employing the newly developed spray cooled system for the thermal management of high-performance servers of data centers in tropical climates and identified key operating parameters that are useful for the design, operation and implementation of this system. [ABSTRACT FROM AUTHOR]

• Review the state-of-the-art research works of integrated energy systems of data centers and smart grids. • Propose future integration scenarios for data centers and smart grids. • Analyze the challenges of implementing integrated energy systems of data centers and smart grids. Cloud computing platforms are critical cyber infrastructures in modern society. As the backbone of cloud systems, data centers act as large energy consumers in today's power grids. The integration of on-site renewable energy sources and energy storage systems further transforms data centers to be energy prosumers (producers-and-consumers). As a result, optimizing data centers' energy production and consumption, and exploiting their potential of actively engaging in the external grid's planning, operation, and control has been drawing increasing attention in the last few years. This paper conducts a comprehensive review of the state-of-the-art research efforts on integrated energy systems of data centers and smart grids. A taxonomy of such integration scenarios is provided. Consequently, this paper identifies several future application scenarios of integrating data centers and smart grids, which serves as a roadmap towards future research. This article is expected to provide a useful reference for researchers and engineers in the areas of energy systems and cloud computing. [ABSTRACT FROM AUTHOR]

[Display omitted] • Development of workload and cooling management framework for multi-rack data centers. • Pareto optimization problem accounts for energy, exergy, and computing efficiencies. • 2-D workload distribution and cooling parameters for a wide range of load factors. • Maximizing cooling energy efficiency paradoxically degrades computing performance. • Extended strategy to manage load factor by data center modularization. The rapidly rising computing workloads in data centers (DCs) have necessitated new approaches to ensure effective performance and resilience that minimize the associated cooling energy. The literature on thermally-aware workload management provides strategies to reduce this energy cost, while typically ignoring the reduction in cooling capacity due to thermodynamic irreversibility and computing performance per unit energy consumption. Hence, we provide an approach that considers coefficient of performance COP c , exergy efficiency η ex , and a new metric, computing performance ratio CPR. In contrast to existing methods that consider one-dimensional workload distributions, the temperature predictions from a physics-based zonal model are used to optimize cooling for two-dimensional workload distributions in a multi-rack DC. The investigation reveals physics associated with two-dimensional workload management for multi-rack DCs, provides a framework for trade-offs between COP c , η ex , and CPR , explains the influence of IT load factor LF on different objectives, and describes how parameters obtained from single- and multi-objective problems can vary. Our findings show that COP c , and η ex can be improved by up to 20% and 8% by regulating the chilled water temperature and airflow setpoints while increasing the LF degrades the CPR by 7.5%. These results enable an extended approach for heterogeneous LF management in large-scale DCs. [ABSTRACT FROM AUTHOR]

• Novel approach using porous partitions is introduced for improving overall data center energy efficiency. • The effect of partitions on data center performance metrics is evaluated. • Practical measures for reducing airflow non-uniformity is investigated to reduce energy losses. • Parametric study for investigating the partitions geometry impact on the Supply heat index is conducted. This investigation focuses on improving the airflow distribution in data centers. In many data centers vortices form in the plenum during operations. These vortices cause spatial and temporal non-uniformities and may give rise to hot regions in the data center which in turn impacts performance and reliability of the IT equipment. The current study identifies a novel approach using porous partitions in the plenum and demonstrates a significant generalized approach that is easily adoptable in existing and future data centers. For improving the overall data center energy efficiency and the cooling system effectiveness by eliminating a critical source of inefficiency. The results of quantitative and qualitative analyses of the underfloor plenum pressure field, perforated tiles airflow rate, and air temperature at the rack intake side with and without partitions are presented. Different data center configurations are studied using physics-based experimentally validated Computational Fluid Dynamics (CFD) model. The CFD model results showed that the partitions eliminated the presence of vortices in the underfloor plenum and thus enabled a more uniform pressure distribution and tile airflow delivery. Regarding rack inlet temperature, the results showed that the partitions significantly improved the air temperature at the rack inlet. Finally, a geometrical parametric study is performed. An ideal design is and demonstrated to improve the Supply Heat Index (SHI) by about 10%, while the amount of IT equipment that exceeded the ASHRAE recommended supply air temperature (SAT) was reduced by about 40%, and the floor leakage was cut in half. [ABSTRACT FROM AUTHOR]

• Integrated planning of Internet data centers and battery energy storage systems. • Coupled cyber-physical modelling of Internet data centers and a smart grid. • Multi-objective modelling of the planning task. Modern power grids have been becoming complex cyber-physical systems integrated with distributed energy sources and information and communication facilities. With prevalence of cloud computing, geo-distributed, networked data centers have become an integrated part of modern grids. The coupling impact between data centers and smart grids thus becomes an important consideration. This paper proposes an integrated planning scheme that optimally determines the locations and capacities of interconnected Internet data centers and battery energy storage systems in a smart grid. The model is formulated as a multi-objective optimization problem, in which both computational performance metrics of Internet data centers and operational criteria of the grid are coordinately considered as three inter-related but conflict objectives; the coupling impact between the cyber and energy resources are modelled. An advanced evolutionary algorithm – Multi-Objective Natural Aggregation Algorithm is used to solve the model. Extensive case studies are conducted to demonstrate the reasonability and effectiveness of the proposed integrated planning method. [ABSTRACT FROM AUTHOR]

• A new indicator, coefficient of power usage effectiveness (COPUE) is proposed. • Energy performances of data centers in various climatic conditions is compared. • A simplified method is developed to present benchmark cooling technology. The power usage effectiveness (PUE) is commonly used as the key performance indicator to evaluate the energy performance of data centers. However, using only PUE cannot enable a fair comparison when data centers are operating in different regions, due to the unneglectable impacts of climatic conditions on the power consumption of cooling systems. To solve this problem, a new indicator, coefficient of PUE (COPUE), is proposed, which is defined as the ratio of the measured PUE of real data centers to the local benchmark PUE. The benchmark PUE is reckoned based on the current most commonly used cooling technology, which consists of water-cooled chillers and water cooling towers. A simplified method for calculating benchmark PUE is also developed. The degree hour of water cooling is introduced to consider the impacts of local climatic conditions. It presents the annual accumulated hours, in which chilled water is needed to satisfy the cooling demand of data centers. Through several case studies, COPUE has been proved to be an effective indicator for comparing the energy performance of data centers. When the same cooling technology is adopted, it can reflect how good the design and operation are; while, when different cooling technologies are adopted, it can be used to demonstrate which one is superior. [ABSTRACT FROM AUTHOR]

• Co-optimization of server provisioning and power procurement. • Chance constraint is introduced to ensure quality of service guarantee. • Expected-value constraint is introduced to improve renewable energy utilization. • Dispatchable workloads can be deployed to increase renewable energy utilization. • Metrics are proposed to measure efficiency and sustainability of data centers. The reduction of greenhouse emissions is becoming a major goal of energy-intensive industries, such as data centers, and there have been significant efforts to achieve sustainable operations by meeting electricity consumption using renewable energy generations. Specifically, it has been a common practice for data centers to use renewable energy via on-site solar power generation to directly offset electricity consumption by renewable energy to contribute to environmental sustainability. However, the introduction of intermittent and non-dispatchable renewable energy generations for powering data centers that generally host time-varying workloads presents a significant challenge, and thus, this study mainly focuses on how to improve renewable energy utilization for data center operations considering the integration of co-located solar power generation and battery energy storage. The main objective is to develop a mathematical optimization model for energy-efficient and sustainable data center operations to minimize energy cost while ensuring the desired level of renewable energy utilization and the required quality of service guarantee. In particular, this study proposes a two-stage stochastic program integrated with an expected-value constraint and a chance constraint, and an integer programming and sampling-based approach are adopted to solve the problem to investigate optimal data center operations. The comprehensive numerical experiments are conducted to evaluate the proposed model compared with benchmark models for various parameter settings, and the results show that the proposed model can be successfully implemented to enable data centers to achieve the desired renewable energy utilization while improving energy efficiency. [ABSTRACT FROM AUTHOR]

• Analyze the server power trends from generation to generation. • Classify and compares existing power consumption models of servers. • Summarize the application scenarios of the power consumption models. • Discuss the model parameters for model construction. • Identify the future research directions for improving data centers' energy efficiency. This study provides an overview of power consumption models of servers in data centers. The server is the basic unit of both power and heat flow paths; therefore, its power consumption model can be used for both energy management and thermal management. Investigations of server power trends were carried out according to the data from the Standard Performance Evaluation Corporation (SPEC). It is found that a heavier workload can be handled without consuming more energy, and the difference between the peak power and idle power of the servers is not consistent from generation to generation. Furthermore, the existing power consumption models are categorized as additive models, baseline power + active power (BA) models, and other models based on calculation formula and other factors. Specifically, there are four forms of BA models: linear regression models, power function models, non-linear models and polynomial models. Besides, these models have been compared in terms of accuracy. It can be found that the polynomial model and the linear regression model perform better in terms of accuracy. Additionally, the model applications are summarized. Considering server architecture upgrades and technological innovation, the establishment of the new model and its application scenarios are discussed. Moreover, in-depth and accurate power consumption models must be extensively researched and applied to effectively improve data centers, including information technology (IT) equipment and cooling equipment, in terms of overall energy performance. [ABSTRACT FROM AUTHOR]

• A model to predict the temperature within an IT server enclosure is developed. • The flow field is predicted based on mechanical resistances. • The influence of various system factors on the cooling power consumption is evaluated. • Passive servers should be placed in racks located closer to the cooling unit. • Use of two in-row cooling units is more energy efficient than using a single unit. Traditional data center cooling methods cannot yet control cooling airflows and temperatures on demand, creating an intrinsic inefficiency. A recent solution places row-based cooling unit adjacent to servers and places the entire assembly within an enclosure, which improves airflow distribution and provides rapid real-time control. This is, in particular attractive for micro-data centers where traditional room-based cooling is less energy efficient. Spatiotemporal predictions of temperatures are required to control and optimize data center performance as the system configuration, and other parameters are varied. Current methods, such as proper orthogonal decomposition, machine learning, and heuristic models are inapplicable in practice because they require a prohibitively large number of a priori simulations or experiments to generate training datasets. We provide an alternative a computationally inexpensive training-free model for enclosed micro-data centers that are integrated with in-row cooling units that requires no a priori training. The model determines the air flowrate within each zone based on a mechanical resistance circuit analysis. These flowrates are then introduced into a zonal energy balance to predict the temperature of each zone. The methodology is validated with experimental measurements and coupled with a power consumption calculation. Its applicability is demonstrated by evaluating the influence of various system factors, such as IT server configurations, cooling unit air, and water flowrates and the numbers of cooling units, on the temperature distributions, and total cooling power consumption. The method can improve micro-data centers control and help to optimize the design of any data center with a row-based cooling system. [ABSTRACT FROM AUTHOR]

• Discuss data centers from perspective of energy prosumers in district energy system. • Review renewable energy usage in data centers and the related advanced controls. • Review data center's waste heat recovery and reuse for district heating. • Review existing green data centers and related economic/environmental analysis. • Identify future research directions for improving data center's overall performance. As large energy prosumers in district energy systems, on the one hand, data centers consume a large amount of electricity to ensure the Information Technologies (IT) facilities, ancillary power supply and cooling systems work properly; on the other hand, data centers produce a large quantity of waste heat due to the high heat dissipation rates of the IT facilities. To date, a systematic review of data centers from the perspective of energy prosumers, which considers both integration of the upstream green energy supply and downstream waste heat reuse, is still lacking. As a result, the potentials for improving data centers' performances are limited due to a lack of global optimization of the upstream renewable energy integration and downstream waste heat utilization. This study is intended to fill in this gap and provides such a review. In this regard, the advancements in different cooling techniques, integration of renewable energy and advanced controls, waste heat utilization and connections for district heating, real projects, performance metrics and economic, energy and environmental analyses are reviewed. Based on the enormous amount of research on data centers in district energy systems, it has been found that: (1) global controls, which can manage the upstream renewable production, data centers' operation and waste heat generation and downstream waste heat utilization are still lacking; (2) regional climate studies represent an effective way to find the optimal integration of renewable energy and waste heat recovery technologies for improving the data centers' energy efficiency; (3) the development of global energy metrics will help to appropriately quantify the data center performances. [ABSTRACT FROM AUTHOR]

• A framework for data center to bid into electricity market and follow regulation signal in real-time. • A risk-constrained bidding strategy is developed to determine the optimal energy and frequency regulation bids. • Dummy load is introduced to increase the amount of regulation service provision of data center. • Bi-linear server power consumption model and rule-based control enable data center to follow regulation signal accurately. • The proposed regulation service provision framework reduces electricity bill by 12%. The rapid adoption of cloud storage and computing services led to unprecedented growth of data centers in the world. As bulk energy consumers, large-scale data centers in the U.S. rack up billions in electricity costs annually. Fortunately, the operational flexibility of data centers can be leveraged to provide valuable frequency regulation services in smart grids to mitigate the indeterminacy of the renewable generation resources. Specifically, this paper aims to leverage computational flexibility provided by servers, such as dynamic voltage frequency scaling and dummy loads. This paper develops a comprehensive framework for data center's frequency regulation service provision in both hour-ahead market and real-time operations. A risk constrained hour-ahead bidding strategy along with a real-time data center power consumption control algorithm are developed to minimize electricity bills and the total response time of the requests. The introduction of dummy load, realistic bi-linear server power consumption model, and probabilistic forecast of electricity and frequency regulation service prices enable the data center to accurately follow frequency regulation signals, while reducing the financial risks associated with electricity market participation. The simulation results show that the proposed frequency regulation provision framework results not only in significant cost reduction for data centers, but also limits degradation in quality of service. Meanwhile, the stability and reliability of a power grid will be improved by the frequency regulation service provision. [ABSTRACT FROM AUTHOR]