Digital Twin Inserts "Sustainability" Wings into Data Center

In order to operate enterprises in a more sustainable way, enterprises are increasingly demanding digital infrastructure, not only for the sake of cost and efficiency, but also from the perspective of the environment.

Arno van Gennip, vice president of global IBX operation engineering at Equinix, said, "From design to construction to facility management, digital twinning is becoming the key to improving data center efficiency and reducing customers' carbon emissions."

Digital twinning helps to centralize the data from different key areas in the * * * enjoyment environment, which enables IT, engineering, finance, procurement and construction teams to explore and simulate the trade-offs among various factors such as performance, finance and environment earlier in the process. The direct impact of various efficiency improvements in equipment and space utilization is to reduce energy consumption and carbon emissions. Digital twinning also helps to improve the efficiency of construction and operation, reduce waste, reduce staffing requirements and related environmental impacts.

many enterprises and data center operators (such as Nvidia) may create digital twin workflows from various simulation modeling tools that combine the functions of engineering, CAD and data center information management (DCIM). More and more DCIM manufacturers (such as Schneider Electric) directly introduce the digital twinning function into their tools. Vendors such as dassault systemes and Future Facilities provide more integrated digital twins for data centers. Manufacturers such as Nvidia have also begun to introduce new tools such as Nvidia Air to optimize the physical and logical layout of data centers.

in operation

Equinix cooperates with Future Facilities to build digital twins for enterprise data centers. Digital twinning can help engineers ensure that the cooling system and the connected ecosystem provide the required capacity and the best efficiency. Engineers can compare the expected behavior of the data center with the actual behavior, as well as the energy use.

"This gives us a deeper understanding of the possibilities of maintaining and optimizing energy efficiency," said van Gennip.

Equinix engineers and partners worked together to build a 3D model of the physical data center. This twin model of data center is based on various factors, such as the capacity and density of computing equipment in the data center and the path of cooling system. The centralized digital twin platform can help engineers to predict the possible impact of predicted changes on power distribution, space utilization and cooling paths by using real-time data (such as power and temperature). These real-time data are integrated into existing models for accurate analysis and prediction, so that data center twins can improve efficiency by predicting energy demand.

dassault cooperates with many leading hyperscale data center enterprises to design and build the next generation data center.

"The biggest challenge they face is how to shorten the project preparation time to keep up with the increasing demand, and how to make the data center more sustainable by reducing energy and water consumption and waste during construction and operation," said Marty Rozmanith, director of sales strategy for architecture, engineering and construction (AEC) industry of Dassault.

making management easier

Kasper Dessing, director of global building management optimization at Digital Realty, a data center real estate investment trust, believes that in the past, data center management was divided into several islands, and each island focused on a certain aspect of management facilities.

therefore, managers in different fields may not be able to see the bigger picture. This is particularly important when considering the maintenance of facilities, both now and in the future. Data center will produce a large amount of data, but human beings can't capture, collect and manage these data well. As digital services become more and more complex, this situation will only get worse.

Dessing said: "Through digital twinning, we can present various elements and dynamics in the facility in a virtual way, and simulate actual behaviors in real time in various operating scenarios."

Digital Realty found that the operation of the general data center is not good enough because of the huge amount of data and the interdependence between different components. Because of this, Digital Realty integrates the digital twins of their facilities with proprietary artificial intelligence and machine learning platforms, analyzes thousands of data streams, so as to be able to track all components in the facilities and make real-time adjustments, and also predict future behaviors, so as to carry out predictive maintenance, save time and reduce costs.

this visibility into the relationship between facilities and different components helps to improve the design of new facilities and make them more efficient. Not only that, Digital Realty also uses digital twins and their artificial intelligence platform to optimize energy consumption.

Dessing said: "Sustainability is our top priority, and optimizing the energy consumption of each facility will help us reduce the cost and reduce the impact on the environment."

Not everyone has the technical expertise of simulating while making decisions, so Digital Realty integrates a recommendation engine into their digital twin platform.

"In this way, more people can use the technology instead of relying on experts all the time."

putting the pieces together

a large amount of data will be generated in the process of designing, building and operating a data center, which is saved in different formats and stored in different systems. Rozmanith said that it is very challenging to manage and organize data through proper access control and change management. Digital twinning can bring data of multiple disciplines, different levels of development (LOD) and multiple dimensions, which enables different stakeholders to collaborate around a single fact source in real time. Those more complex digital twinning technologies combine various technologies and use an integrated digital twinning to simulate the processes of heat, structure, electricity, control and monitoring, manufacturing and assembly.

Rozmanith explained, "With a common platform, everyone works around a single source of facts, which not only saves time, but also improves the quality and overall data center delivery capacity. This platform is a driving force for change for us. "

Teresa Tung, chief technical expert of Accenture Cloud, said, "As we integrate more data and simulations to connect engineering design, construction scheduling and operation processes, interoperability between different digital twins has become a big challenge."

Tung's team is working with data center vendors to apply data and domain expertise to the analysis process to determine the number and configuration of simulations needed to drive hypothetical forecasting. They use domain knowledge maps (the same technology used in Internet search) to capture these requirements and map the relationships between different elements.

Carsten Baumann, director of Schneider's strategic planning and solution architect, said that providers are increasingly adding digital twinning functions to DCIM tools to simulate the possible impact of infrastructure upgrade before actual deployment. He believes that open standards can simplify the integration between data center equipment and management tools, thus making it easier to make digital twins a part of daily data center workflow.

Let's take a detailed look at 19 ways that digital twinning can improve the sustainability of design, construction, operation and planning:

Design

Place new servers

"Perhaps in the data center industry, the biggest impact of using digital twinning technology is the air flow management and IT equipment placement," Baumann said.

the demand for deploying computing, storage and network resources is growing rapidly, which brings great challenges in infrastructure. Just because there is physical space in a specific rack or location does not mean that there is enough power supply, access and cooling capacity.

when a seemingly simple installation and deployment may require a major upgrade of the power supply or a better alternative, digital twinning can help.

increasing the density

increasing the equipment density in the data center can reduce the impact of new facilities on the climate.

Loren Absher, director of enterprise agility of information services group (ISG), said that digital twinning can help to optimize the design of data center, improve all related elements such as power supply, wiring, cooling requirements, airflow and even the integrity of raised floor, so as to prevent catastrophic failures, and also help to change the physical workflow needed to increase density.

improving thermal performance

cooling is the second largest energy consumption factor in the data center, second only to the equipment itself. The cooling system of modern data center includes coolers, pipes and HVAC equipment.

digital twins can use thermal simulation to understand the behavior of the cooling system and improve its performance.

Rozmanith said that some people often combine the 1D simulation of the equipment chain representing the change of the number of water chillers and the pipe size with the 3D computational fluid dynamics (CFD) analysis of air flow to find the best balance between cold air and equipment cooling, so as to optimize energy consumption.

Assessing seasonal impacts

Dan Kirsch, managing director and co-founder of TechStrong Research, said that digital twinning can also help data center designers to better plan seasonal climate change, so that designers can plan ahead according to the impact of external seasonal climate change to reduce overall operating costs and energy consumption.

"Digital twinning allows us to truly customize and optimize the design according to customers' specific needs and site conditions without field experiments," Kirsch said.

create modular components

dassault cooperates with large data center operators to create modular components that can be reused in different data center designs.

Rozmanith said that digital twinning can help enterprises to define and configure the properties of these modules, thus shortening the design, procurement and installation time by configuring them according to orders, thus helping to reduce the environmental impact of new data centers.

test and verify equipment

Bruno Berti, senior vice president of Americas products at NTT Global Data Center, said that they are using digital twinning to test and verify equipment before deploying it into the data center.

these new workflows enable them to build and test electrical and generator modules, so that engineers can have any potential process failures before the products are put into production, reduce the impact of wastes on the environment, improve risk assessment, accelerate the development of new products, and improve the reliability and flexibility of data centers. In addition, digital twinning also helps to arrange predictive maintenance and reduce maintenance costs.

Optimizing battery performance

Greg Ratcliff, chief innovation officer of Vertiv, a data center equipment manufacturer, said that digital twinning can be used to model and design systems to improve battery health and life expectancy, thus reducing the environmental impact of manufacturing new batteries. In this case, digital twinning can help the team to use battery health measurement and facility details to simulate different design options and predict the health status and service life of each battery.

Ratcliff said: "If a single battery in the battery pack fails, the whole battery pack will fail, so it is very important to monitor the operation of each battery."

evaluating environmentally friendly alternatives

data center operators can use digital twinning technology to evaluate the performance, environmental benefits and potential defects of new methods.

for example, Kao Data uses digital twin tools to virtually test and deploy a refrigerant-free indirect evaporative cooling (IEC) system, which uses water evaporation instead of mechanical systems to cool air in hot weather. This method helps Kao Data improve the efficiency of power utilization and reduce the impact on the environment.

Building

Streamlined construction

Digital twinning can simulate complex tasks, assembly, equipment use and personal safety, and also improve the cooperation among suppliers, integrators and contractors in the design and construction ecosystem to eliminate friction in the process.

Rozmanith said that better simulation and collaboration can shorten the construction time, reduce problems, avoid rework, and reduce the number of information requests and safety accidents, which helps Dassault's customers shorten the time to market by 1-15% on average, and reduce the possible environmental impact caused by long construction time.

reduce construction waste

data center designers are using digital twinning to better plan construction, so that staff can work more efficiently, reduce waste and shorten the time between different construction stages.

Kirsch said: "By creating a virtual model of the data center and a complete bill of materials, designers can optimize every detail of the data center assembled by construction workers."

This planning method can reduce the waiting time for one team to complete tasks in other teams. Generally speaking, it is not easy to reduce the waste in the process of data center construction. Kirsch said that many components in this process cannot be reused or recycled, and finally they can only enter the landfill.

Operation

Providing maintenance advice

Digital twinning helps to identify the root cause of the problem and provide maintenance advice for quick repair to reduce energy consumption.

For example, Equinix's factory in Amsterdam adopted a digital twin model, according to which they had to clean the cooling tower and adjust the fan, both of which used to consume more energy than the model expected. van Ge