Modelling and control of a free cooling system for Data Centers

Beghi, Alessandro; Cecchinato, Luca; Mana, Giuseppe Dalla; Lionello, Michele; Rampazzo, Mirco; Sisti, Enrico

doi:10.1016/j.egypro.2017.11.156

Cited by 29 publications

(18 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In [8], adaptive vent times (AVT) is considered for local cooling, and a cooling power and temperature model is built, based on which a cooling control scheme is proposed to reduce the cooling power with the constraint of rack thermal requirement. Free cooling method is considered in [9] and a cooling model containing all the main plant components is built based on First-Principle Data-Driven (FPDD) techniques. The cooling energy consumption optimization is achieved using the Particle Swarm Optimization method.…”

Section: Related Workmentioning

confidence: 99%

“…Without consideration of the cooling system and the hot spot issue of servers, these methods improve the energy efficiency of the IT system but usually at the cost of increasing the cooling energy consumption. In other works, such as [8][9][10], adjusting the cooling facility such as the supply air temperature and flow rate to achieve the minimum cooling energy consumption is the primary concern. In these works, a slightly simplified cooling model is usually formulated to reduce the calculation overhead, based on which the cooling adjustment is optimized and the performance is verified.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Cooperatively Improving Data Center Energy Efficiency Based on Multi-Agent Deep Reinforcement Learning

Chi

Song

et al. 2021

Energies

View full text Add to dashboard Cite

The problem of high power consumption in data centers is becoming more and more prominent. In order to improve the energy efficiency of data centers, cooperatively optimizing the energy of IT systems and cooling systems has become an effective way. In this paper, a model-free deep reinforcement learning (DRL)-based joint optimization method MAD3C is developed to overcome the high-dimensional state and action space problems of the data center energy optimization. A hybrid AC-DDPG cooperative multi-agent framework is devised for the improvement of the cooperation between the IT and cooling systems for further energy efficiency improvement. In the framework, a scheduling baseline comparison method is presented to enhance the stability of the framework. Meanwhile, an adaptive score is designed for the architecture in consideration of multi-dimensional resources and resource utilization improvement. Experiments show that our proposed approach can effectively reduce energy for data centers through the cooperative optimization while guaranteeing training stability and improving resource utilization.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Cooperatively Improving Data Center Energy Efficiency Based on Multi-Agent Deep Reinforcement Learning

Chi

Song

et al. 2021

Energies

View full text Add to dashboard Cite

show abstract

“…DC cooling systems are classified into two types. One is an air-based cooling system [6,7]. The other one is a liquid cooling system [8].…”

Section: Introductionmentioning

confidence: 99%

Thermal Energy Storage Performance of Tetrabutylammonium Acrylate Hydrate as Phase Change Materials

Kiyokawa

Tokutomi

Ishida³

et al. 2021

Applied Sciences

View full text Add to dashboard Cite

Kinetic characteristics of thermal energy storage (TES) using tetrabutylammonium acrylate (TBAAc) hydrate were experimentally evaluated for practical use as PCMs. Mechanical agitation or ultrasonic vibration was added to detach the hydrate adhesion on the heat exchanger, which could be a thermal resistance. The effect of the external forces also was evaluated by changing their rotation rate and frequency. When the agitation rate was 600 rpm, the system achieved TES density of 140 MJ/m3 in 2.9 hours. This value is comparable to the ideal performance of ice TES when its solid phase fraction is 45%. UA/V (U: thermal transfer coefficient, A: surface area of the heat exchange coil, V: volume of the TES medium) is known as an index of the ease of heat transfer in a heat exchanger. UA/V obtained in this study was comparable to that of other common heat exchangers, which means the equivalent performance would be available by setting the similar UA/V. In this study, we succeeded in obtaining practical data for heat storage by TBAAc hydrate. The data obtained in this study will be a great help for the practical application of hydrate heat storage in the future.

show abstract

“…Modeling and numerical simulation, along with selective experimentation, have been used to obtain the flow and temperature fields, as well as heat transfer rates in data centers (Iyengar, 2007;Amemiya et al, 2007;Abdelmaksoud et al, 2010b;Nada and Elfeky;Beghi et al, 2017). These results can be used to determine hot spots in the electronic hardware and areas that need heat transfer enhancement, so that the thermal management system may be designed to meet the stringent temperature constraints of electronic systems that typically require device or electronic chip temperatures to be lower than about 80 o C. Variations in flow configuration and operating conditions like flow rates and inlet temperatures, that are controlled by the use of chillers, may then be employed to obtain satisfactory thermal management.…”

Section: Introductionmentioning

confidence: 99%

Thermal Management of Data Centers Under Steady and Transient Conditions

Jaluria¹,

Sunder²,

Benner³

2020

Frontiers in Heat and Mass Transfer

View full text Add to dashboard Cite

Data centers are of crucial importance today in the storage and retrieval of large amounts of data. Most organizations and firms, ranging from banks and online retailers to government departments and internet companies, use data centers to store information that can be recovered efficiently and rapidly. As the deployment of data centers has increased, along with their capacity for data storage, the demands on thermal management have also increased. It is necessary to remove the energy dissipated by the electronic circuitry since the temperature of the components must not rise beyond acceptable levels that could damage them or affect their performance. The energy required for the cooling of data centers is generally quite substantial, with typical data centers requiring hundreds for megawatts for heat removal. Therefore, it is important to develop efficient thermal management systems to reduce power consumption and minimize the environmental impact. Extensive work has been done in recent years on the cooling of data centers. Though most studies have focused on steady state processes, time-dependent behavior is also of considerable interest because the thermal load is generally not constant but varies with time. Fluctuations in the load may arise or the load may peak at certain times. Efficient time-dependent distribution of load between various data centers, if several data centers are available to a given organization, or between different servers of a given data center may be used to improve the efficiency of the heat removal process and thus reduce the overall power consumption. Since environmental conditions vary from location to location, appropriate data centers may be selected, from those available, in order to use the most favorable location to reduce the overall cooling load. These aspects demand a study of both the steady and time-dependent behavior of data centers. The results obtained may be used to optimize the thermal management system and reduce the overall energy consumption. This paper reviews some of the relevant work done on data center cooling, with particular attention given to load distribution and to the data center location that determines the local environmental conditions. These results could form the basis for optimizing the cooling strategy for a system of data centers.

show abstract

Modelling and control of a free cooling system for Data Centers

Cited by 29 publications

References 4 publications

Cooperatively Improving Data Center Energy Efficiency Based on Multi-Agent Deep Reinforcement Learning

Cooperatively Improving Data Center Energy Efficiency Based on Multi-Agent Deep Reinforcement Learning

Thermal Energy Storage Performance of Tetrabutylammonium Acrylate Hydrate as Phase Change Materials

Thermal Management of Data Centers Under Steady and Transient Conditions

Contact Info

Product

Resources

About