Subtask A: Basic principle, Performance indicators of data center heat exhaust processes

Basic classification of data center heat exhaust systems. Develop unified thermal resistance model for data center heat exhaust processes, including both cooling and waste heat recovery systems. Identify common performance indicators, to evaluate electricity consumption, water consumption and so on. For waste heat recovery systems, give methods to evaluate the extra electricity consumption for heating purpose. Using the unified evaluation methods to analyze the performance of data center cooling and waste heat recovery systems under different climates all over the world. Analyze practical cases of air-cooled systems and air-liquid fusion systems.

The scope of this subtask is as follows:

1. Classification and typical configurations of data center cooling systems: Data center cooling systems will be classified from the perspectives of end cooling methods (air-cooled, liquid-cooled), cooling system configurations (centralized chilled water systems, air-cooled AHU systems, etc.), and the independence of air-cooled/liquid-cooled cold sources (air-cooled systems, liquid-cooled systems, air-liquid hybrid systems). A review of typical data center cooling system configurations will be conducted, and representative system configurations adopted by participating countries will be identified.
2. Based on the heat exchange process, a thermal resistance model for the heat dissipation process in data centers is established. This model is then applied uniformly to the analysis of cooling processes and waste heat recovery systems for mutual comparison.
3. Performance evaluation indicators for cooling systems: Based on the aforementioned classification framework, cooling performance evaluation indicators (including PUE, WUE, exergy-based optimization, entransy-based optimization, etc.) will be defined for various cooling system configurations. For air-cooled systems (including centralized chilled water systems and air-cooled AHU systems), the performance evaluation indicators will be specified. For liquid-cooled systems, the coupling relationship between computing power energy efficiency and cooling parameters, as well as cooling system performance, will also be investigated.
4. Evaluation indicators for data center waste heat recovery systems (summer heat rejection systems): For scenarios involving data center waste heat utilization, the accounting methodology for heat pump electricity consumption under winter heat supply mode utilizing heat pumps for waste heat recovery will be investigated. The annual PUE calculation method incorporating heat pump energy consumption of the waste heat recovery system will be explored, with clarification on whether heat pump electricity consumption should be counted within the total data center power consumption or alternative reasonable treatment approaches. The feasibility of sharing equipment between summer electric chillers and winter heat pumps during heat pump selection for waste heat recovery will be analyzed, i.e., whether the electric chiller design should comprehensively consider winter heat supply requirements.
5. Comparative analysis of liquid-cooled and air-cooled waste heat recovery systems: The differences in system configurations between liquid-cooled and air-cooled systems under waste heat recovery scenarios will be compared, together with their impacts on evaluation indicators, considering the direct utilization of liquid-cooled waste heat without heat pump temperature elevation.
6. Evaluation of typical heat rejection systems across global climate zones: Based on the proposed evaluation methods, performance evaluation and analysis of typical data center heat rejection systems across different global climate zones will be conducted. Typical system configurations and key parameter designs shall be provided during the evaluation.
7. Collection of real cases: Collect real case studies of data center air-cooled systems and air-liquid fusion systems. Primarily consider the terminal cooling source system. Analyze the performance levels of these case studies.

Subtask B: Evaporative Cooling application and Water consumption of data center cooling system

Basic classification of evaporative cooling systems applied in data centers: IEC AHU, cooling towers, IEC cooling towers and so on. Introduce the basic system structures using different evaporative cooling systems, analyzing the performance of different systems under different climates of different countries. Water consumption performance analysis of different cooling systems, corresponding to the water resources, give the potential of water saving. Real cases collection of data center cooling systems using different IEC/DEC technologies during participating countries.

In data center cooling systems, heat is typically rejected to the ambient environment through evaporative cooling processes. The most commonly used evaporative cooling technologies include indirect evaporative cooling air handling units (IEC-AHU), direct evaporative cooling towers, and indirect evaporative cooling towers. The performance of these technologies varies across different climatic regions and system configurations. Key performance indicators generally include Power Usage Effectiveness (PUE) and Water Usage Effectiveness (WUE).

It is important to emphasize that water consumption has received increasing attention in data center cooling systems. Therefore, evaluating the water usage characteristics of different evaporative cooling technologies, in conjunction with regional water resource availability, is essential for determining their applicability and deployment boundaries.

The scope of the subtask is as follows:

1. Analyze the performance of the following representative system configurations across multiple climatic regions in different countries, with key metrics including Power Usage Effectiveness (PUE) and Water Usage Effectiveness (WUE):
   1. Indirect evaporative cooling air handling unit (IEC-AHU) based air cooling systems;
   2. Centralized chilled water systems coupled with cooling towers for air cooling applications;
   3. Centralized chilled water systems coupled with cooling towers for liquid cooling applications;
   4. Centralized chilled water systems integrated with indirect evaporative cooling towers for air cooling applications;
   5. Centralized chilled water systems integrated with indirect evaporative cooling towers for liquid cooling applications.
2. Water consumption characteristics, and potential water-saving strategies.
3. Real cases collection of data center cooling systems using different IEC/DEC technologies during participating countries.

Subtask C: Data Center liquid cooling systems

To meet the chip-level high-efficiency cooling requirements of high power density data centers, key technical research will be conducted on typical liquid cooling system configurations, chip thermal management, internal heat transfer performance, system parameter design, and coolant property evaluation. Through comparative analysis and optimized design of different liquid cooling approaches, theoretical guidance and technical support will be provided for the engineering application and standardized design of data center liquid cooling technologies. And collect cases of actual testing at the chip end.

The scope of this subtask is as follows:

1. Typical liquid cooling system configurations: Classified by cooling form and liquid working state, typical liquid cooling system configurations including cold plate liquid cooling (single-phase, two-phase, jet impingement), immersion liquid cooling (single-phase, two-phase), and on-chip microchannel liquid cooling with directly integrated microchannels on chip surfaces will be reviewed.
2. Chip temperature uniformity: The temperature uniformity among different chips within the same liquid cooling circuit, as well as the temperature distribution uniformity within a single chip itself, will be investigated. The distribution characteristics and formation mechanisms of local hot spots will be analyzed, and solutions or mitigation measures for local hot spots will be proposed.
3. Liquid cooling heat transfer performance and heat rejection temperature requirements: The internal heat transfer performance of different liquid cooling approaches (including heat transfer coefficient, thermal resistance, critical heat flux, etc.) will be comparatively analyzed. Two-phase cooling physics, including boiling stability and vapor management.
4. The coolant temperature requirements for satisfying chip heat dissipation demands will be specified.
5. Liquid cooling system temperature parameter design and system design: The temperature parameter design (such as coolant supply/return temperature, temperature difference, etc.) and system design of liquid cooling heat rejection systems will be conducted. Typical design approaches for air-liquid hybrid systems will be developed. Dynamic and transient thermal behavior under fluctuating computational loads.
6. Liquid cooling coolant property evaluation: Thermophysical characterization of advanced dielectric coolants and refrigerants (including PFAS-related concerns); The flow characteristics (pressure drop, flow distribution uniformity, etc.) of liquid cooling coolants within a single cabinet and among different cabinets, as well as the vaporization and corrosion characteristics of coolants, will be investigated. Comprehensive performance evaluation and applicability analysis of currently commonly used liquid cooling coolants (such as deionized water, ethylene glycol aqueous solutions, fluorinated fluids, etc.) will be conducted. Long-term reliability, degradation, fouling, and coolant aging.
7. Collection of real cases: Collect case studies of actual tests on chip terminals using various liquid cooling methods. Analyze the heat dissipation performance levels in these cases.
8. Based on this, compare the characteristics of different liquid cooling methods.

Subtask D: Waste heat recovery of data centers: system design and applications

Potential configurations of data center waste heat recovery systems include multi-stage heat pump systems combined with thermal energy storage, as well as heat pump-assisted or direct heat exchange using plate heat exchangers. The selection of system configuration primarily depends on the temperature level (grade) of the available waste heat and the temporal and spatial mismatches between heat supply and demand. To achieve efficient utilization of waste heat, systematic investigations are required on system configurations and operational parameters of waste heat recovery systems, with the aim of developing a set of practical and transferable design methodologies.

The scope of the subtask is as follows:

1. Develop system configurations and identify key operating parameters for data center waste heat-based district heating systems.
2. Investigate the integration of data center waste heat with thermally driven cooling technologies to meet air-conditioning demands requiring low-temperature cooling sources, and evaluate system configurations and feasibility.
3. Analyze the matching characteristics between multi-stage heat pumps and thermal energy storage in waste heat heating systems to improve system performance and operational flexibility.
4. Real cases collection of data center waste heat recovery systems.

Subtask E: Computing and Power Synergy for data centers

Reasonable scheduling of data center computing workloads and renewable energy resources to achieve higher energy utilization efficiency is an important approach for improving data center energy efficiency. In addition, increasing attention has been paid to the relationship between cooling system temperature levels and chip energy consumption.

The scope of the subtask is as follows:

1. Give the feasibility analysis of data center demand response to power grid
2. Give guiding methodologies and fundamental principles for the coordination between data center computing workloads and power systems
3. Establish the relationship between liquid-cooling coolant temperature levels, chip power consumption, and cooling system energy consumption

Subtask F: Non-technical challenges, governance, policy recommendations

Data center cooling and waste heat recovery are also strongly influenced by non-technical factors. In areas such as data center waste heat utilization, non-technical challenges related to economic feasibility, stakeholder responsibilities, risk allocation, and ownership models still need to be addressed.

Different countries have accumulated diverse experiences in this regard and face challenges that are similar in nature yet differ in specific contexts.

The scope of the subtask is as follows: Give practical guidance for practitioners, policymakers, municipalities, district heating companies, data center operators, and technology suppliers.