A recent study found that the average American spends about 45% of their waking hours using information-accessing technology (computers, phones, etc). The computing equipment needed to maintain this level of activity is stored in facilities called data centers. The energy demands of such centers are sometimes up to 100x the energy needs of a typical commercial facility. As anyone who has sat with a laptop on their lap for an extended period of time knows, computers release a significant amount of heat.
As such, data centers use energy for two main functions:
- Running computing equipment
- Cooling the equipment
Here is what some of the biggest players in the industry are doing to make your internet browsing more sustainable by implementing data center energy solutions:
Google is using an approach, called machine learning, to analyze and interpret information collected from sensors throughout their facilities. Essentially, the computing equipment can learn from past energy usage trends, how to run the system more efficiently in the future. Engineers use a series of what could be described as “artificial neurons” to process information. The goal is to simplify and replicate the process of human cognition. This is a classic example of how a biomimicry problem-solving approach can help achieve sustainability goals.
Facebook has taken a geographic approach to solving the energy efficiency problem. The company recently opened a mega-sized data center in the arctic circle. The facility requires significantly less energy than typical data centers because it uses the cold arctic air to cool the computing equipment. The remainder of the facility’s energy needs come from renewable energy sources. Most impressively, Facebook is aiming to help other companies and organizations overcome these energy efficiency obstacles. The designs for their state-of-the-art data center are being made publicly available via a program called Open Compute.
Microsoft has implemented a strategy to power their Cheyenne, Wyoming data center with methane generated at a nearby waste water treatment plant. Typically facilities operating with biogas need to import methane. However, the Data Plant is directly tied to the sewage treatment plant, and is fueled entirely by the waste byproduct of the sewage treatment process. Microsoft continues to investigate ways to make this process more efficient, through the use of smaller fuel cells that could decrease the operating and maintenance costs of the facility.