Data Center Microgrid Research Articles

Stochastic optimization for capacity configuration of data center microgrid thermal energy management equipment considering flexible resources

The rapid development of Internet and cloud computing technologies has led to the expansion of the scale of data centers, resulting in a continuous increase in the demand for data processing and storage. This not only results in higher energy consumption but also makes thermal management a key factor of energy management in data centers. A capacity configuration method is proposed for the core thermal management equipment of data center microgrids, electric boilers, cooling systems, and heat pumps. This method not only considers the uncertainty factors of the source and load but also effectively utilizes the characteristics of various flexible resources in data centers, such as the adjustable ability of different types of batch processing loads, air thermal inertia, and waste heat recovery, significantly improving the energy utilization efficiency and enhancing the system’s economy and flexibility. In addition, a wind power scenario generation method based on conditional least squares generative adversarial networks was proposed to improve the generation quality of wind power scenarios, and a random optimization model was constructed. Considering a certain province’s data center microgrid as a case study, the effectiveness of the model was verified, and a sensitivity analysis of the relevant parameters provided important references for microgrid planning. This study provides a new perspective and tool for the thermal management of data center microgrids and comprehensive utilization of renewable energy, which is of great significance for promoting the improvement of data center energy efficiency and sustainable development.

Joint optimization of operational cost and carbon emission in multiple data center micro-grids

As the internet data centers are mushrooming, the energy consumption and carbon emission of data centers are increasing rapidly. To cut down the electricity cost and carbon emission of the data centers, we proposed an optimization method to reduce the electricity cost and carbon emission in geo-distributed multiple data centers. In the proposed method, the carbon tax is considered in the overall operation cost to optimize the carbon emission. The spatial and temporal flexibility of computational workload is fully utilized by considering the difference in renewable energy power output, local electricity and carbon emission of multiple geo-distributed data centers to achieve a better performance. Furthermore, the nonlinear characteristics of the power loss of uninterruptible power supply (UPS) are considered in the optimization. To verify the proposed optimization method, simulation of six cases is carried out with realistic data, and results have proved the proposed method can reduce the operational costs by 4.93%–12.7% and decrease carbon emissions by up to 10%.

Hybrid energy storage configuration method for wind power microgrid based on EMD decomposition and two-stage robust approach

Data centers are usually characterized by high energy loads, which raises increasing sustainability concerns in both academic and daily usage. To mitigate the uncertainty and high volatility of distributed wind energy generation, this paper proposes a hybrid energy storage allocation strategy by means of the Empirical Mode Decomposition (EMD) technique and the two-stage robust method. First, this paper conducts the evolution analyses for the over- and under-evaluated uncertainty of wind power fluctuation under different time scales. Second, we employ the EMD technique to configure a high-frequency flywheel energy storage device, realizing the wind power transformation from large fluctuations to small fluctuations and the convergence of the wind power fluctuation curves in minute- and hour levels. Finally, based on the hour-level wind energy stable power curves, we carry out two-stage robust planning for the equipment capacity of low-frequency cold storage tanks and lithium bromide chillers. The case study on a data center microgrid in northeastern China confirms the effectiveness of our proposed strategy.

Cooperative online schedule of interconnected data center microgrids with shared energy storage

The rapid increasing of data centers calls for an efficient method to reduce high operation costs and carbon emissions. This paper proposes a cooperative online schedule framework for local interconnected data centers considering shared energy storage. A time-average optimization problem is built to reduce the overall operation cost with essential operational constraints. Multiple randomness and the temporal coupling constraints of energy system levels make it challenging to solve the stochastic problem. Adapting the Lyapunov optimization theory with relaxation and perturbation technique, an online algorithm is proposed to find an acceptable solution to the problem to achieve both good service coordination and efficient energy coordination. The algorithm does not require prediction of random parameters, and has a fast convergence rate. Especially, the theoretical conditions for convergence and the distance between the acceptable and real optimal solution are derived in a closed form. Simulation studies show that the algorithm can effectively reduce operation cost of data centers, carbon emissions, and dependence on the main grid.

A Multicriteria Optimal Operation Framework for a Data Center Microgrid Considering Renewable Energy and Waste Heat Recovery: Use of Balanced Decision Making

The rapid development of data centers (DCS) leads to a huge challenge in their energy consumption and environmental impact. It is promising to establish DC microgrids (DCMGs) for solving these issues, considering utilizing renewable energy (RE) generation and waste heat recovery systems. However, the efficient energy management of a DCMG is a topic to be pursued. In this article, we propose a multicriteria optimal operation framework for the DCMG by coordinatively scheduling the energy supply and demand. In this framework, multiple criteria are comprehensively considered via a multicriteria optimization (MCO) approach. Then, we adopt an augmented constraint algorithm to address the corresponding MCO problem. Afterward, a balanced decision-making (BDM) method is further proposed to determine the optimal scheduling solution. Finally, a case study and results analysis verify the effectiveness of the proposed multicriteria optimal operation framework for the DCMG.

Robust multi-objective optimization for islanded data center microgrid operations

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.

Energy Optimal Dispatch of the Data Center Microgrid Based on Stochastic Model Predictive Control

Renewable energy outputs such as wind turbines and photovoltaics, as well as data center workloads are both random and uncertain. In order to enhance the stability and economy of the data center in actual operation effectively, a multi-time scale optimal dispatch method for the data center microgrid based on stochastic model predictive control is proposed in this paper. In the day-ahead scheduling stage, the characteristic of the data center that batch workloads are allowed to be served delayed is considered. And the scenario analysis method is applied to describe the uncertainty of loads and renewable energy outputs, based on which an economic optimization scheduling model is established to minimize the system operating cost. The intra-day scheduling utilizes the rolling optimization and feedback correction of model predictive control to correct the deviation of loads and renewable energy outputs and adjust the day-ahead dispatch plan in real time, which ensures the effectiveness of day-ahead plan and the stability of system operating. Through the simulation results of a typical data center microgrid, the effectiveness of the proposed method is verified.

Flexible Multi-Energy Scheduling Scheme for Data Center to Facilitate Wind Power Integration

Due to the intermittent and uncontrollable nature of wind resources and inflexible operation of conventional generation units, they present challenges for the power system to integrate more wind power. With its unique flexibility on the demand side, the data center can be considered as an effective solution to relieve wind curtailment. Moreover, with the help of waste heat recovery module, the data center can reduce the utilization of conventional thermal units especially in the residential heating sector which increases the flexibility of system operation and facilitates more renewable integration. In this paper, a flexible workload management and resource scheduling model are proposed to achieve a multi-energy co-optimization for data center and enhance the integration of wind power. A two-stage stochastic programming model is formulated to address the uncertainties involved in this process. The proposed model is examined by a simulative data center microgrid and the numerical results demonstrate its effectiveness and robustness.

Towards green data center microgrids by leveraging data center loads in providing frequency regulation

In an electricity grid, imbalance between generation and load need be corrected within seconds so that frequency deviations will not threaten the stability and security. This is especially important for a low inertia microgrid operated in islanded mode, which is equipped with a limited number of synchronous generators in regulating frequency. To this end, for data center microgrid with limited on-site generators and increased green energy, when isolated from the utility grid, frequency deviation due to generation-load imbalance could be potentially corrected by conventional generating units as well as data center loads. Focusing on high PV penetrated data center microgrid operated in islanded mode, this paper explores effective control strategies for data center loads to participate in primary frequency response. By analyzing unique operational characteristics of traditional and PV generation units, uninterruptible power supply (UPS) units, and power consumption characteristics of IT components and cooling systems, the proposed load control strategy design effectively utilizes primary FR capabilities while not compromising data center quality of service (QoS) requirements. Numerical simulations via MATLAB/Simulink illustrate effectiveness of the proposed load control strategy in enhancing renewable energy penetration without compromising the system stability and security, providing a viable solution for future green data centers.

Integrated Stochastic Energy Management for Data Center Microgrid Considering Waste Heat Recovery

The rapidly developing data center industry results in a large amount of energy consumption. Considering its unique demand characteristics, it becomes desirable to improve its energy efficiency by reusing its waste heat. In this paper, an integrated energy management scheme is proposed to optimize the operation cost for a data center microgrid. The waste heat recovered from data center operation is optimally scheduled with other resources in the integrated energy management model to minimize the operation cost of data center microgrid. Taking the uncertain nature of renewable supply, power and heat demand into consideration, a two-stage stochastic programming model is formulated and analyzed under different test cases. The numerical study results demonstrate the effectiveness of the proposed integrated energy management model.

Planning of Distributed Internet Data Center Microgrids

This paper proposed an optimal planning model for locational distributed Internet data centers' (IDCs) computing and power supply infrastructures-IDC microgrids. The objective is to determine the optimal IDC sites, microgrid types (i.e., ac, dc, or Hybrid), IDC server types (ac or dc) and capacities, as well as onsite distributed energy resource types and capacities of each site to be built, under a single decision criterion - economics. The overall planning cost including investment cost, operation cost, and reliability cost associated with the IDC microgrid installation and operation is minimized. Both the long term investment and short term operation details of distributed data center microgrids are studied and modeled in the proposed integrated planning problem. Numerical simulations demonstrate the effectiveness of the proposed model over a 4-year planning horizon considering the evolution of data center servers. Factors which may impact the planning decisions are analyzed through a set of detailed sensitivity analysis.

Emission-Aware Stochastic Resource Planning Scheme for Data Center Microgrid Considering Batch Workload Scheduling and Risk Management

Internet data centers are booming in the last few years, leading to gigantic amount of energy consumption and greenhouse gas emission. In this paper, a data center microgrid including conventional units, energy storage, and renewable generation is modeled to optimize its electricity bill and carbon footprint. Considering the randomness of electricity price, renewable output and workload distribution, the volatility risk is incorporated to measure the associated operation risk. Then, a day-ahead emission-aware stochastic resource planning scheme is formulated to decide the strategy on power procurement, energy storage operation, batch workload allocation, and unit commitment of conventional units. Simulation results show the effectiveness of the proposed approach.

Distributed Real-Time Energy Management in Data Center Microgrids

Data center operators are typically faced with three significant problems when running their data centers, i.e., rising electricity bills, growing carbon footprints and unexpected power outages. To mitigate these issues, running data centers in microgrids is a good choice since microgrids can enhance the energy efficiency, sustainability and reliability of electrical services. Thus, in this paper, we investigate the problem of energy management for multiple data center microgrids. Specifically, we intend to minimize the long-term operational cost of data center microgrids by taking into account the uncertainties in electricity prices, renewable outputs and data center workloads. We first formulate a stochastic programming problem with the considerations of many factors, e.g., providing heterogeneous service delay guarantees for batch workloads, interactive workload allocation, batch workload shedding, electricity buying/selling, battery charging/discharging efficiency, and the ramping constraints of backup generators. Then, we design a realtime and distributed algorithm for the formulated problem based on Lyapunov optimization technique and a variant of alternating direction method of multipliers (ADMM). Moreover, the performance guarantees provided by the proposed algorithm are analyzed. Extensive simulation results indicate the effectiveness of the proposed algorithm in operational cost reduction for data center microgrids.

Toward Optimal Operation of Internet Data Center Microgrid

This paper studies the optimal operation of internet data centers (IDCs) in microgrid environments. The proposed model is to minimize the IDC microgrid operation cost, which consists of the operation cost of onsite generations, cost of purchasing power from the main grid, cost due to the repeated charging and discharging of onsite energy storage system (ESS), and the cost due to IDC workload shedding. Both preventive and corrective operation solutions, including power purchase from the main grid, onsite generation dispatch, as well as ESS and IDC workload scheduling are provided for provisioning IDC workloads while considering credible islanding scenarios of the IDC microgrid. Demand response capability of the IDC microgrid is achieved through the proposed optimal solution. The use of ESS for facilitating electricity load shifting is further analyzed via a sensitivity study for minimizing the operation cost of IDC microgrid. Numerical simulations demonstrate the effectiveness of the proposed optimal operation model for IDC microgrid.

A Comprehensive Model of Data Center: From CPU to Cooling Tower

Aiming at addressing environmental challenges, large data centers, such as Facebook, Google, and Yahoo, are increasing share of green power in their daily energy consumption. Such trends drive research into new directions, e.g., sustainable data centers. The research often relies on expressive models that provide sufficient details, however, practical to re-use and expand. There is a lack of available data center models that capture internal operating states of the facility from the CPU to the cooling tower. It is a challenge to develop a model that allows to describe complete data center of any scale including its connection to the grid. This paper proposes such a model building on the existing work. The challenge was to put the pieces of data center together and model behavior of each element so that interdependencies between components and parameters and operating states are captured correctly and in sufficient details. The proposed model was used in the project “Data center microgrid integration” and proven to be adequate and important to support such study.