English

Five Ways to Improve Data Center Energy Efficiency and Cut Energy Costs

Updated on Jun 1, 2022 by
4.0k

Data Center Energy Consumption is getting all the limelight due to growth of this industry. Operating data centers can become financially burdensome if the data center power infrastructure is ill-designed or obsolete. According to the International Energy Agency, 1% of all global energy is consumed by data centers and they are expected to devour 1/5th of the world’s power supply by 2025.

However, by implementing a series of wise management choices, even a small-sized data center can save thousands of dollars. Below are some of the smartest ways data centers can increase energy efficiency while dramatically reducing the costs of operation.

Five Ways to Improve Data Center Energy Efficiency and Cut Energy Costs

Investing in Modern Cooling Systems

The fact that in most data centers, the cooling equipment accounts for 30 to 60 percent of the utility bill is alarming. The numbers do not have to be so dire since most of the expended power is the result of inefficient practices and failure to meet the recommended settings for equipment operation set by ASHRAE. A hypothetical data center with 1000 servers can easily save tens of thousands of dollars by simply reducing 25 percent of its cooling costs.

By introducing simple and implementable cooling solutions, companies can decrease the immense utility loads. First, data centers must deploy hot aisle/cold aisle configurations. One can create a more uniform air temperature by alternating the cold aisle with a hot air exhaust. To prevent the air from the two aisles from mixing, you can use blanking panels inside equipment enclosures.

Minimize “bypass airflow” that affects about 60 percent of the cool air supply. Ensure that the cool air is recycled back to cooling units instead of circulating throughout the data center. The computer room air-conditioning units must be oriented perpendicular to the hot aisles and towards the enclosures to maximize cooling efficiency.

Data centers can further accelerate the cooling output by investing in air-handlers and chillers. The latter is equipped with high-end technologies such as air or water-side economizers, variable frequency drives (VFDs), humidity settings, and temperature adjustments.

Software-Based Smart Design

Experts believe that enhanced data center energy efficiency boils down to the simple question of which solutions serve the greatest workload with the lowest amount of hardware needed. Software-based smart designs and storage solutions can ensure greater availability.

Data centers must identify underutilized pieces of equipment and bring the power down. It would be more cost-effective in the long run to rid oneself of an underused application and implement more cost-effective ways to serve the need.

It is also possible for modern IT equipment to operate safely in warmer environments instead of overcooling the equipment, wasting money and energy. By using temperature sensors and data center infrastructure management (DCIM) software, one can safely raise temperatures while ensuring both the efficiency and safety of the equipment.

This is achieved by adjusting your temperature sensors. The best way is to place no more than six sensors in a rack. By installing DCIM software, you can adjust appropriate rack inlet thresholds that alert you once there is the danger of over-temperature.

A modern DCIM tool is equipped with built-in cooling charts that help companies achieve compliance with the changes recommended. You must also refer to the cooling charts to ensure you’re not trespassing any of ASHRAE’s (American Society of Heating, Refrigerating, and Air-Conditioning Engineers ) thermal guidelines. Also, keep tabs on the thermal heat along with performing periodic checks on your temperature trends.

One last thing data centers can do to increase efficiency in this department is to remove “bloatware”. Bloatware refers to software that is ineffectual and drains power by using excessive CPU cycles. Instead, equip the place with more efficient software that can reduce the CPU cycle, and generate greater processing output for an equal amount of power input.

Leveraging Virtualization

Virtualization allows data centers to consolidate servers and storage on a shared platform all the while maintaining the desired segregation between applications, operating systems, and data. It is possible to run most applications on virtual machines that share hardware with other applications. This enhances hardware utilization and brings down the power consumption rate by dramatically reducing the number of storage devices and servers. This is important because according to one study, storage and servers together account for over 50% of the energy consumed in a data center.

Consolidating the hardware and opting for virtualization are great ways for data centers to reduce the need for more space, greater power, and cooling resources. By reducing the number of physical assets, data centers can automatically bring down all related costs.

Here again, DCIM software would be inevitable to maximize the output. Not only would you be able to track the location of all of your physical assets, but also keep tabs on power consumption, and ghost servers. It will also allow you to identify the stranded capacity for the safe deployment of more devices in fewer racks. The software also enables data centers to identify key metrics to gauge performance, for example, CPU utilization, system processes, memory usage, virtual machines per cluster, number of hosts, hosts per application and OS, etc.

While virtualization may be the solution to many of your energy-related problems, it will not solve all. A certain amount of underutilized equipment would be inevitable to meet periodic peak loads. However, It can still go a long way in migrating workload more efficiently.

Innovative Tariffs

Implementing innovative tariffs is another way to derive data center efficiency. Already, many organizations have introduced green data center initiatives to kick-start energy-efficient practices amongst both internal users as well as external customers. The key is to ensure compliance from all parties involved.

By charging customers taxes that are tariffed on the basis of the energy consumed, data centers can drive a culture of awareness and accountability around energy efficiency, ensuring better consumption behavior.

Utilizing Airflow Management

Implementing airflow management is a critical energy-saving practice. Data centers with raised floors can easily suffer from uncontrolled air leakage that results in disproportionate quantities of cool air reaching the IT equipment. Energy efficient data centers must seal the raised floors properly and remove unnecessary blockages placed under the floors.

Underfloor blockages are also responsible for inhibiting cool air from reaching the IT equipment. This in turn results in fan energy surcharge as well as higher HVAC supply air temperatures. Such blockages can be controlled using visualization methods to study airflow patterns. Any obstructions under the floor must be removed or relocated. Air leaks, primarily at the pipe and cable penetration points, must be fixed immediately. Moreover, the installation of high-caliber temperature sensors greatly helps with the optimization process.

Overhead cable trays are the best alternative if underfloor blockage and clutter control is impossible. Schneider Electric has researched that this move can lower cooling fan power consumption by 24%.

Conclusion

While the above practices are by no means exhaustive in ensuring higher data center energy efficiency, they still form a host of implementable solutions that require little to no external costs with massive returns on investment. The key is to create a culture of awareness that can minimize unnecessary power consumption by energy-intensive equipment.

You might be interested in

Knowledge
Knowledge
Knowledge
Knowledge
Knowledge
Knowledge
See profile for Sheldon.
Sheldon
Decoding OLT, ONU, ONT, and ODN in PON Network
Mar 14, 2023
386.2k
Knowledge
See profile for Irving.
Irving
What's the Difference? Hub vs Switch vs Router
Dec 17, 2021
367.6k
Knowledge
See profile for Sheldon.
Sheldon
What Is SFP Port of Gigabit Switch?
Jan 6, 2023
335.6k
Knowledge
Knowledge
See profile for Migelle.
Migelle
PoE vs PoE+ vs PoE++ Switch: How to Choose?
Mar 16, 2023
420.6k