I recently found out, for example, that the computing equipment in my home office last year consumed 803 kilowatt-hours of power and directly resulted in the emission of 889 lbs. of carbon dioxide and other greenhouse gases into the atmosphere. (See "The numbers behind the numbers" at the bottom of the article for the formulas used.) My personal contribution to global warming also included 1.4 lbs. of sulphur dioxide and about half a pound of nitrogen oxide, all by-products of the power-generation process serving my office, according to the Independent System Operator of New England.
All that energy came with a financial cost as well, of course. Could I save money and help save the planet by reducing my energy use without compromising my business? To answer that question, I set out to eliminate every wasted watt of energy -- and wasted dollar -- I could find in my office setup.
The result was shocking. Based on an audit of my own office's energy use, I estimated that when I started, my equipment added $112 (£60) to my annual electricity bill, or 8.5% of the total for my household. Had I been more careful in the selection, configuration and use of the equipment, I could have saved as much as 80% of that -- and put $90 (£45) back into my pocket.
If that sounds like small change in the grand scheme of things, multiply it by the 36 million home offices in the U.S. that use computer and communications equipment, according to market research company IDC. If all offices in the U.S., both commercial and home, used Energy Star-certified equipment, individuals and businesses would save $1.8 billion in energy costs, and the resulting reduction in greenhouse emissions would be equivalent to removing 2.7 million cars from the roads, according to the U.S. Environmental Protection Agency.
My biggest savings came from some simple changes. I replaced or eliminated inefficient equipment and changed how I configured and used it. After a few big wins, however, the law of diminishing returns began to take hold. The hardest part was deciding what trade-offs I was willing to make to save ever smaller increments of power.
But I'm getting ahead of myself. I couldn't do anything to make my office more energy efficient without knowing where I already stood. "Measure to find out, experiment to optimise, and keep measuring," advises Amory Lovins, chairman and chief scientist at the energy efficiency think tank Rocky Mountain Institute. I decided to follow that advice by performing an energy audit.
Measuring the problem
Fortunately, an energy audit is something you can do yourself with the help of an inexpensive metering device such as P3 International Corp.'s Kill A Watt meter. That device has a retail price of $39.99 (£20), but I found one for less than $20 (£10) on Amazon.com. Even better, to my surprise, I was able to borrow one from my local library. The Kill A Watt plugs into a power outlet and has its own outlet on the front for attaching the device you want to monitor. Among other things, the device displays power draw in watts and tracks cumulative power consumption over time in kilowatt-hours.
By placing all of my devices on two daisy-chained power strips and plugging one of them into the Kill A Watt, I was able to track power consumption for my equipment as a whole. "Your IT equipment should use an average of just 0.2 watts per square foot if you spec and operate it optimally," says Lovins. Mine wasn't even close.
This wasn't really a surprise: "energy-efficient" is not a term I would have expected to apply to my home office. The 120-square-foot space contained a mish-mash of about a dozen devices spread across twelve feet of desk space:
-- A Lenovo and a Dell laptop, each with a docking station
-- A 19-in. CRT monitor
-- A laser printer and a multifunction ink-jet printer (both long in the tooth
-- A network storage device
-- A set of powered speakers
-- A cell phone and charger
-- A two-line phone and a cordless headset
-- A cable modem and a wireless router
With the exception of one laptop, none of it was Energy Star-certified.
My workday typically includes talking on the phone, taking notes, checking e-mail, doing research online, printing out 20 to 30 pages of notes and research and writing up stories in Microsoft Word. At the end of the workday, I simply turned out the lights and walked out of the office. The monitor turned off after 15 minutes of inactivity, and I blithely assumed that everything else went into low-power mode. It didn't.
The Kill A Watt is the key to an energy audit.
With the Kill A Watt's help, I discovered that my typical daytime load ranged from 55 watts at idle to more than 160 watts, or about 1.3 watts per square foot, when I was sitting at my desk and typing at the keyboard. Had I been using a desktop instead of a laptop, that number would have been at least 50 watts higher.
Because the wattage measurements varied so widely, I decided to look at total power consumption over a 24-hour period. It turned out that my office sucked away 2.2 kilowatt-hours (kWh) of power during that time. Since all of that power is converted to heat, I was also warming up my office to the tune of 562 British Thermal Units (BTUs) per hour. That's roughly one-sixth of the heat output from a 1,000-watt hair dryer, heat that my fan had to work to remove.
(For those who find these measurements confusing: Watts measure the load, or energy demand, of a device; think of it like the weight of an object you're trying to hold in your outstretched arm. Kilowatt-hours measure how much energy the device actually uses over time, similar to the amount of work you'd have to do to hold up that heavy object for a while. And a BTU is the energy it takes to increase the temperature of 1 pound of water by 1 degree F.)
As I made energy-saving changes to my office, I continued to measure the aggregate energy consumption to determine the bottom-line results. As I said, I managed to cut my energy usage by more than half. Here's how.
The low-hanging fruit
I began by replacing my CRT, which consumed about 90 watts (more than half of my peak load) when operating and 3 watts in standby mode. I chose a ViewSonic Corp. 19-in. LCD panel (available for about $200), which pulled 29 watts when in use and less than 1 watt in standby mode, shaving about $18 (£9) a year off my bill all by itself. It was also easier on my eyes -- a big benefit when you look at a screen eight hours a day.
That was my biggest single savings, but you can do even better here. While my monitor is Energy Star-certified, there's a big difference -- as much as 10 watts in some cases -- between the energy efficiency leaders and those that just make the cut. The EPA doesn't give you the complete power comparison breakdown for Energy Star monitors, but Pacific Gas and Electric Co. has calculated those numbers because it uses them to determine rebates to vendors.
You won't find those rankings on the PG&E Web site (although it's considering publishing them), but the utility does make the information available in spreadsheet form, which we've duplicated below.
My ViewSonic LCD came with a bonus: embedded speakers. Since I play my music at low volume anyway, I decided to turn off my powered speakers, which consume at least 5 watts when in use and 3 watts when idle. But it turned out that shutting off the power button only cut their power use from 3W to 2W -- they continue to demand energy as long as they're plugged in.
This prompted me to go back and test other components, and I quickly realised that simply turning a piece of equipment off doesn't stop it from continuing to pull power. Like dripping faucets, my network storage device, speakers and Dell docking station continued to leak 5W of power even when turned off.
To eliminate that load completely, I would need to put them on a separate power strip that I could switch off. I eventually did that for most of them. Some new power strips, such as BITS Ltd.'s Smart Strip Power Strip, and uninterruptible power supplies, such as American Power Conversion Corp.'s soon-to-be-released BE750G, automatically cut power to devices connected to selected outlets when your computer goes into standby mode, which would make them a good solution to this problem.
But dealing with the network storage device -- a Western Digital NetCenter -- wasn't so simple. Western Digital Corp. has replaced the NetCenter with the My Book World Edition, but that model also has no sleep mode. The company says it is working on adding power management features in future models. Unlike the other peripherals, however, I couldn't just put the NetCenter on a power strip and turn it off without thinking through the consequences.
I use the device as a shared file server and to back up both my office computers and a home computer elsewhere in the house. The NetCenter also has a print server feature that I use to share the LaserJet to all the computers.
My home machine primarily uses a direct-attached ink-jet printer and accesses the LaserJet only infrequently, so I could live with turning on the NetCenter the few times a month when the home PC needed to print to the LaserJet. In the interest of eliminating those last couple of watts as well, I decided to forgo file sharing and only turn on the NetCenter once a day for about 20 minutes to complete backups, and I put it on a power strip so I could eliminate its power drain entirely at night.
I also decided to replace the LaserJet, which has plenty of miles on it and has always had an annoying tendency to produce curled pages. When the smaller, faster, more energy-efficient LaserJet 1018 went on sale for $64.99 after a $50 rebate -- less than the cost of a toner cartridge for the 1200 -- I was sold. The new unit consumes just 2 watts in standby mode, 4 watts less than the LaserJet 1200. When compared using the Typical Electricity Consumption operating power tests (download PDF) required by the EPA's newest Energy Star guidelines for imaging equipment, the 1018 draws 65 fewer watts during operation -- 23% less power -- than the 1200.
Finally, I decided to turn off my secondary laptop, the Dell, which used to spend most of its time sitting in standby mode. Now I turn it completely off when I'm not using it, just leaving it plugged in to keep the battery from discharging. And to reduce its power consumption for those times when I do use it, I have configured power management to turn off the monitor, spin down the disk and go into standby mode after 15 minutes of inactivity and to hibernate after 30 minutes.
I then turned to the power management settings for my primary machine, a Lenovo ThinkPad T43. While its monitor timeout was already set to 15 minutes, the disk timeout was set to "Never." I changed the settings to match the Dell, the setttings recommended by a Lenovo Group Ltd. spokesperson
The Rocky Mountain Institute's Lovins suggests even more aggressive settings. "Since your hard drive is designed to spin up and down an enormous number of times, don't hesitate to cut its turnoff setting from 15 minutes to, say, two or three. The same for the monitor," he says. But since most people never even turn on power management settings, I decided to stick with the manufacturer's recommendations this time around and see where that got me.
After all that, typical power consumption when I was at my desk dropped to about 70 watts, or 0.58 watts per square foot, and the off-hours load dropped to just 42 watts. Total power consumption came in at 1.33 kWh over a 24-hour period, or 485 kWh a year -- a 40% improvement in energy efficiency. That would reduce my annual electricity bill to $68, a savings of $44.
At this point, the trade-offs got a bit stickier. I've been using my office for eight to 10 hours a day but still leaving everything powered on around the clock. I decided to start turning everything off by flipping the big red switch on the power strip at the end of each workday.
By doing this, I am reducing not just operating hours per workday, but also the number of days that my office equipment is in service, which will drop from 365 days a year to 250 after vacations and weekends are eliminated. The upshot is that my annual power consumption should come down to just 0.59 kWh a day, which projects out to just 147.5 kWh a year. Bottom line: I could cut my IT equipment energy bill to just $21 -- a $91 savings.
I went further. I pushed my disk and monitor timeouts to 3 minutes. I kept standby mode at 15 minutes but set hibernate for 20 minutes -- the lowest setting.
Then I turned to my cable modem and wireless router, each of which consumes 5 watts. I rent the cable modem for $3 per month, so I checked with my Internet service provider Comcast Corp. about getting a new one. The latest model would save just 1 watt over my current unit. I could exchange it at no charge, but Comcast wouldn't deliver it. So to reduce 1 watt's worth of my carbon footprint, I'd have to increase it by a 30-mile drive to the nearest Comcast office. Not worth it.
Another option was to trade my cable modem for a consolidated wireless gateway device that would also replace my wireless router and consume half the power of the two devices together. For that, though, Comcast would charge me a mandatory $150 installation charge plus $5 per month. Ridiculous.
So I looked into buying and installing my own wireless gateway. Comcast wouldn't support it, of course, but the real problem was that I'd need to spend a substantial sum to save just 5 watts. The Linksys WCG200 wireless gateway, for instance, sells for $130, and installing it would save only about $6 per year in electricity costs and $36 a year in cable modem rental charges -- about a three-year payback.
Unlike replacing the printer and monitor, buying a new gateway wouldn't offer any immediate advantages other than saving 5 watts, so I decided to wait for a normal replacement cycle to take that step. In the interest of pushing the power savings envelope, however, Cisco-Linksys LLC offered to loan me a WCG200 for this story, and I put it on one of my power strips and set it in operation for my next round of measurements.
Those changes brought my total energy consumption down to 0.54 kWh per day. But was that sustainable? Putting the gateway on a power strip and turning it off at the end of the workday meant that no one could use the Internet after hours from the family computer without going into my office to turn on the wireless gateway and wait for it to start up and reconnect. That was doable, but not very convenient. Losing the shared printing was also inconvenient, but the home system doesn't use the LaserJet much. My family could accept that.
Remembering to punch the power button on the NetCenter every time I wanted to back up -- and then to turn the unit off again -- was a challenge. I put a daily backup reminder in my calendar at lunchtime to help ensure that backups were completed.
But actually buying my own wireless gateway didn't fit my budget. I returned it, restored the cable modem and wireless router, and put them on a timer that cycles them on at 7:00 a.m. and off at 11:00 p.m. That allows the family to use the Internet connection in the evening while reducing demand by 10 watts overnight, saving another 0.08 kWh per day.
Today, my office equipment consumes about 0.63 kWh a day and draws about 70 watts of power during work hours and less when idle. Aside from the direct energy savings, that translates into 316 fewer BTUs/hour of heat coming off my equipment. My office is noticeably cooler, particularly later in the day. Over the next year, I should consume 645 kWh less power, shave $90 off my electric bill, and reduce my share of utility CO2 emissions by 715 pounds. That's an energy efficiency compromise I can live with.
The numbers behind the numbers
For a given device:
kilowatts (kW) = watts/1,000
kW x total hours = kWh
Total kWh x rate = total electrical cost
For my office:
kWh x $0.14 per kWh [my rate] = total cost
365 days x 24 hrs = 8,760 hrs/yr
250 days x 8 hrs = 2,000 working hrs/yr
Total hours x kWh x $0.14 = annual cost of electricity
Formula to calculate BTUs:
watts x 3.413 = total BTU per hour
Formulas used to calculate greenhouse gas impact:
Source: Public Service Company of New Hampshire/Independent System Operator of New England
SO2: kWh/yr x .00175 lbs/kWh = SO2 lbs/year
NOx: kWh/yr x .00054 lbs/kWh = NOx lbs/year
CO2: kWh/yr x 1.107 lbs/kWh = CO2 lbs/year
(National average from DOE: 1.55 lbs/kWh)