Putting the heat on data center cooling costs

Data centers are becoming one of the single largest industrial energy users, consuming 61.4 billion kilowatt-hours annually in the U.S. alone, and producing more than 43 million tons of C0₂ each year.

Between 40 percent and 60 percent of this annual energy spend is consumed by cooling systems for data center IT equipment. Consequently, there is an enormous market opportunity to deliver power and cooling solutions that can reduce costs dramatically and improve IT availability (see Data centers: $7B in annual energy costs).

As data centers get larger, they are getting thirstier as well. The enormous volume of water required to cool high-density cloud computing server farms is making water management a growing priority for data center operators. For example, a 15-megawatt data center can use up to 360,000 gallons of water daily.

On top of the water required for cooling, water is consumed at the power generation level. According to the U.S. Geological Survey, 39 percent of the water used in the United States is for power production.

In view of this sobering data, data center designers and facility managers must reevaluate how to go about reducing power and cooling requirements. This is going to involve a much more stringent evaluation of next-generation computer room air-conditioning (CRAC), air-handling units (AHU) and chiller systems that have demonstrable efficiency advantages over legacy systems.

Clearly, the table stakes for breakthrough cooling solutions in the $6 billion retrofit and new construction data center market are huge. While this data center “plumbing” may not be as glamorous as the more visible computing side, it will constitute a major piece of the new energy efficiency infrastructure of the 21st century (see IBM rides third wave of cleantech). The next “Cisco of cooling” will emerge from this wholesale change in data center economics.

A good metric that the industry is relying on more and more is a power utilization efficiency (PUE) rating. Currently, data centers deploying cooling infrastructures based on legacy systems are averaging a PUE of 2.25, according to the Uptime Institute. With the introduction of next-generation CRACs, AHUs and chiller systems, PUEs can be reduced to 1.25 without sacrificing temperature or humidity control. To put that in a financial context, lowering a PUE from 2.25 to 1.25 can slash spending from $0.44 per ton-hour for cooling to less than $0.05 per ton-hour.

Facility managers should be very careful to avoid being seduced by various technologies vying for dominance in the data center cooling market. This is especially true with airside economizing (ASE) solutions that have been the rage in the industry of late. In point of fact, ASE uses more energy, requires additional consumable expenditures for filters, and does not provide consistent temperatures and humidity levels.

Refrigerant side economizing (RSE), on the other hand, can reduce refrigeration compressor energy by up to 100 percent. By deploying RSE solutions directly in the return air based on the outside wet-bulb temperature, rather than dry bulb temperature relied upon by ASE, economizing hours can be increased by as much as 50 percent, as compared to airside economizing and typical waterside economizing.

Next-generation cooling solutions also play a vital role in reducing water usage on-site at data center facilities and at regional power plants. A typical 100-ton water-cooled data center consumes 1.2 million gallons of water per year at the on-site facility. This can double or triple depending on cycle concentration rates in cooling towers.

Consequently, a 100-ton data center consuming 1.7 million kWh per year for cooling on average requires 3.2 million gallons of water for the cooling load at the power plant. With a next-generation data center-cooling infrastructure, water consumption can be drastically reduced to 900,000 gallons at the power plant per year due to the total energy savings.

Like death and taxes, the absolute certainty surrounding data centers is that they will continue to expand to support more users, more compute-intensive equipment and applications, and increasing demand for higher availability (see Booming data-center business creates troubles for India). Simply put, that demand calls for even greater energy consumption.

This leaves data center managers with two vexing choices: either invest in upgrading the power and cooling infrastructure of existing data centers via a major retrofit, or build out entirely new facilities, both of which can be significantly capital-intensive alternatives (see Study: India leads the pack in green server rooms). In the case of new data center construction, costs can run as high as $2,000 to 3,000 per square foot.

There is a general trend toward consolidating many smaller data centers into one or a few large data centers. The theory behind these centralized data centers is that they can support higher equipment densities and more effectively use available floor space than a number of smaller centers can. That’s why computing hardware is increasingly moving to space-efficient form factors such as 1U rack servers and blade servers, which allow higher computing density in a given floor space.

However, this trend toward increasing server density is resulting in higher power densities in many data centers. The impact of these dramatically higher power densities is that, in many cases, power and cooling capacity are the primary constraints to expansion of computational capacity within a data center (see Study: Computing industry not embracing green IT).

Another impact of higher energy densities is that server hardware is no longer the primary cost component of a data center. The purchase price of a new (1U) server has been exceeded by the capital cost of power and cooling infrastructure to support that server. And this will soon be exceeded by the lifetime energy costs for that server alone. This represents a significant shift in data center economics that seriously challenges conventional cooling strategies.

The U.S. Environmental Protection Agency forecasts that data centers will double their cumulative power and water consumption levels by 2011 and every five years thereafter. The industry faces a legitimate and daunting question: Is there enough water supply to support the energy needed for power hungry data centers?

According to the U.S. Department of Energy's Energy Information Administration (EIA), data centers use more water than the paper, food, auto, wood, petroleum, and plastics industries. Only the chemical industry's 153 billion kWhs and metal industry's 144 billion kWh require more water because of power usage.

Historically, water use has been seriously overlooked when designing and permitting a data center. Given the total sustainability footprint that data centers leave behind, the industry is now at a tipping point with respect to implementing new solutions that reduce energy and can reduce water consumption significantly.

Rick Cockrell is a senior energy efficiency designer with more than 20 years in the energy efficiency/refrigeration space. He is currently the chief technical officer of Core4 Systems, a data center cooling vendor that aims to cut water use and cooling costs with next-generation computer room air-conditioning  and air-handling unit systems. Contact him at rcockrell@core4sys.com.

Want to author a guest column yourself? We welcome contributions, and would like to hear from you. Guidance and directions here.

Cleantech developments making news in the past 24 hours

• Metabolix, ADM to open new bioplastics factory
• Rockwool nears construction on Indian insulation factory

View the last week of cleantech developments here | Subscribe to full feed