Solutions provider takeaway: This section of the chapter excerpt from "The Green and Virtual Data Center" by Greg Schulz teaches solutions providers how they can assess and improve energy usage in the data center. Learn the different approaches for determining energy consumption, the various metrics for measuring energy and which information is important to clients.• Determining your energy usage in the data center
• How to create an energy-efficient data center
• About the author
When was the last time you looked at your business energy bill or your home electric bill? If you have not done so recently, look at or ask someone about what your facilities energy bill looks like. Also, look at your home energy bill and see how many kilowatt-hours you used, what the base energy rates are, what surcharges and other fees are assessed, and other information. Once you have reviewed your energy bill, can you determine where, when, and how electrical power is being used?
There are different approaches to determining energy usage. One is to take a macro view, looking at how much energy a data center consumes in total and working down to individual devices. The opposite approach is to work backwards from a micro view of components in a device and total the results across the various devices. Measured electricity usage can be obtained from utility bills, meters, or facilities management software tools, depending on what is available. Sources for estimating power consumption on a device or component level are vendor-supplied information, including equipment faceplates, and site planning guides. Other sources for determining energy usage are power meters and analyzers as well as power distribution devices that can also measure and report on power usage.
Electrical power is typically charged by utilities based on the number of kilowatt-hours used or the number of 1,000 Watt (W) of electricity used. For example, a server that draws 500 W consumes 0.5 kWh, or a storage device that consumes 3,500 W when being actively used consumes 3.5 kWh. Note that while energy usage varies over time and is cumulative, energy from a utility billing standpoint is based on what is used as of an hour. That is, if a server draws 500 W, its hourly energy bill will be 500 W or 0.5 kWh, as opposed to 500 x 60 seconds x 60 minutes. Likewise, electrical power generation is quoted in terms of kilowatt-hours or mega (million) watt-hours (MWh) as of a given point in time. For example, an 11- MW power plant is capable of producing 11,000 kWh at a given point in time, and if usage is constant, the energy is billed as 11,000 kWh.
Typically, energy usage is based on metered readings, either someone from the utility company physically reading the meter, remote reading of the meter, or, perhaps, estimated usage based on historical usage patterns, or some combination. Electric power is charged at a base rate (which may vary by location, supply, and demand fuel sources, among other factors) per kilowatt- hour plus fuel surcharges, peak demand usage surcharges, special fees, applicable commercial volume peak usage minus any applicable energy saver discounts. For example, for voluntarily reducing power consumption or switching to standby power generation during peak usage periods, utilities may offer incentives, rebates, or discounts.
IT technology manufacturers provide information about electrical energy consumption and/or heat (Btu per hour) generated under a given scenario. Some vendors provide more information, including worst-case and best-case consumption information, while others provide only basic maximum breaker size information. Examples of metrics published by vendors and that should be visible on equipment include kilowatts, kilovolts, amperage, volts AC or Btu. Chapter 5 discusses various PCFE-related metrics and measurements, including how to convert from watts to Btu and from amperes to watts.
To calculate simple energy usage, use the values in Table 2.1, selecting the energy costs for your location and the number of kilowatt-hours required to power the device for one hour. For example, if a server or storage device consumes 100 kWh of power and the average energy cost is 8¢/kWh, energy cost is $70,100 annually. As another example, a base rate for 1 kWh might be 12¢/kWh but 20¢/kWh for usage over 1,000 kWh per month. Note that this simple model does not take into consideration regional differences in cost, demand, or availability, nor does it include surcharges, peak demand differentials, or other factors. The model also does not differentiate between energy usage for IT equipment operation and power required for cooling. The annual kWh is calculated as the number of kWh x 24 x 365.
A more thorough analysis can be done in conjunction with a vendor environment assessment service, with a consultant, or with your energy provider. As a start, if you are not in the habit of reading your monthly home energy bill, look to see how much energy you use and the associated costs, including surcharges and fees.
Table 2.1 Example Annual Costs for Various Levels of Energy Consumption
Greg Schulz is the founder of The StorageIO Group, an IT consulting firm. He has worked as a programmer, systems administrator, disaster recovery consultant and capacity planner for various IT organizations and also worked for several vendors before joining an analyst firm and later forming StorageIO. In addition, Schulz is a prolific writer, blogger and speaker who regularly appears at conferences and other events around the world. He is the author of "The Green and Virtual Data Center" and "Resilient Storage Networks".