Weathering the Oncoming Carbon Allocation Storm

With impending government regulation of carbon emissions, the development of an energy cost and carbon allocation strategy will be essential in minimizing the business and financial impact on an organization. Within an office building, the data center will most likely represent the biggest source of carbon emissions and wasted energy. A logical first step, in order to avoid excessive energy expense, is to learn how to measure the data center's carbon footprint. 

A data center is energy intensive. In fact, data centers can be 40 times as energy intensive as the surrounding office building and have been identified as one of the fastest growing consumers of energy.

Figure. Energy consumption profile comparison of data center and office building

The U.S. Environmental Protection Agency (EPA) is mandating that government agencies develop strategies for energy efficiency in government-operated data centers with a target of 20 percent improvement by 2011. Private sector data centers may soon be asked to meet mandated carbon emissions (CO₂) limits.

Historically, data center design and operations have been focused on reliability and capacity without a focus on optimizing the infrastructure for efficiency. In fact, when IT and facilities staff collaborate, energy and carbon savings opportunities on the order of 20 to 90 percent, are possible in the realm of existing data centers,

Energy use is a substantial cost of IT operations, in some cases exceeding the cost of IT hardware. This cost pressure, combined with the realization that data centers can be much more efficient, has caused forward thinking business executives to make energy management a priority.

A Call for Action

Private organizations and governments are proposing that all corporations be assigned a carbon allocation. Under such an allocation plan, exceeding the allocation would result in fines and other penalties. The concept of carbon credits, and the ability to trade carbon credits, has been introduced in many parts of the globe.

Business professionals must first determine their energy usage and then convert that usage into carbon emissions. To determine whether a data center's carbon footprint is being reduced, the first step is to properly measure actual consumption. Data center location, IT load, and electrical efficiency are the three key variables that impact data center power consumption the most and are the key drivers for the development of data center energy consumption measurement tools.

First Key Factor: Location

Outdoor temperatures and humidity levels are a significant influencing factor on energy consumption. A geographical location which experiences extreme temperatures and humidity levels will consume more energy as the data center physical infrastructure systems (e.g., power and cooling systems) work harder to maintain consistent, moderate temperature and humidity levels.

The local source of power generation will also have a major impact on a data center's carbon footprint. In France, for example, most of utility generated power comes from nuclear energy. A data center in France would have a much lower carbon footprint, from a daily operations perspective, than one located in the Midwestern US. In the case of the U.S. data center, the energy source "mix" may be 60 percent coal, 20 percent oil, 10 percent natural gas and 5 percent hydro and 5 percent wind farms. The data center in central France would draw 95 percent of its electricity from a nuclear power plant. A nuclear reactor does not emit CO₂. A fossil fuel burning power plant emits significant quantities of CO₂.

Figure. Electricity generation by fuel type: USA vs. France

Second Key Factor: IT Load

IT load reflects how much power the IT equipment in the data center consumes. The IT load consists of all of the IT hardware components that make up the IT business architecture:  servers, routers, computers, storage devices, telecommunications equipment, as well as the security systems, fire and monitoring systems that protect them. The higher the load, the more power will be required and the higher the carbon footprint.

Third Key Factor: Electrical Efficiency

The traditional practice in data centers of oversizing the physical infrastructure to support the IT load has a negative impact on overall data center efficiency and therefore impacts carbon footprint. Data centers are oversized in order to build in a fat margin for error in terms of estimating data center capacity. Oversizing results in underutilization of equipment (such as servers that are plugged in 24 hours a day but that are very rarely used). Fortunately, new generations of modular scalable IT and physical infrastructure equipment allow for a "pay as you grow" approach that helps to optimize equipment utilization. Advancements in capacity planning software also allow for much more accurate prediction of data center capacities and power consumption.

Measurement tools

Automated measurement tools such as the APC by Schneider Electric Data Center Carbon Calculator, Efficiency Calculator, Energy Allocation Calculator, and Power Sizing Calculator TradeOff Tools™ (http:// tools.apc.com ) can be accessed at no cost and can help data center professionals understand how they use electricity, and to recognize how changes in efficiency can impact carbon footprint.

Figure. Tools calculate carbon footprint and yearly per-server energy cost

Together these tools provide an estimate of efficiency and carbon footprint. Entering the data into each of the tools requires some knowledge about the particular data center's environment (e.g., number of servers, type of cooling systems, power distribution). In some cases, IT staff and facilities staff will work together to gather appropriate input data. Third parties, such as APC by Schneider Electric, can provide assistance by offering data center efficiency assessments that can also provide valuable input to the tools.

Simple, no-cost models of energy use can be used to allocate energy and carbon based on standard, average units of IT capacity. Such models are not precise, but have sufficient accuracy to be useful in a data center energy management system.

For more information on carbon allocation and data center carbon emissions see APC white papers #66, "Estimating a Data Center's Electrical Carbon Footprint" and #161, "Allocating Data Center Energy Costs and Carbon to IT Users" available for review at www.apc.com.