Energy Internet and eVehicles Overview
Governments around the world are wrestling with the challenge of how to reduce carbon dioxide emissions. The current preferred approaches are to impose carbon taxes and implement various forms of cap and trade. However another approach to help reduce carbon emission is to “reward” those directly who reduce their carbon footprint and complement their existing lifestyle. One possible reward system is to provide homeowners with free fiber to the home or free wireless products and other electronic services if they deploy micro renewable energy sources for their ICT equipment and use eVehicles for energy transportation. Not only does the consumer benefit, but this business model also provides new revenue opportunities for small businesses, network operators, and eCommerce application providers.
Linking renewable energy with the Internet using eVehicles and dynamic charging where vehicle's batteries are charged as it travels along the road, may provide for a whole new "energy Internet" infrastructure for linking small distributed renewable energy sources to users. For more details please see:
Free High Speed Internet to the Home: http://goo.gl/wGjVG
High level architecture of Building Zero Carbon Networks: http://goo.gl/juWdH
Wednesday, March 31, 2010
See also Educause paper on this subject:
A major step forward in turning university campuses into "Living Labs of the Greener Future" - UCSD's Energy Dashboardhttp://bit.ly/boeJJ2
UC San Diego Energy Dashboard to Help Campus Curb Appetite for Power
San Diego, March 29, 2010 -- After an extensive period of testing, researchers have launched an Internet portal to showcase the real-time measurement and visualization of energy use on the University of California, San Diego campus.
The UC San Diego Energy Dashboard (http://energy.ucsd.edu/) allows users to see up-to-the-second information on a structure-by-structure basis for 60 of the largest buildings on the La Jolla campus. The data is provided by UC San Diego Physical Plant Services from over 200 energy meters providing energy usage at the building level. The portal also features information coming from roughly 40 individual power meters that measure energy consumption in the office, e.g., a computer and monitor drawing power from a single socket. A denser deployment of meters, which would measure and display individuals’ energy use, is currently under planning and development.
The Energy Dashboard grew out of a simple premise. “If you cannot measure energy use, you will not be able to make much headway in reducing your energy footprint,” said Yuvraj Agarwal, a Research Scientist in the Jacobs School of Engineering’s Computer Science and Engineering (CSE) department.
“Energy models of buildings are decades old, and nobody was looking to see if those were still valid,” added Agarwal, principal architect of the dashboard. “People tend to think that by shutting off the lights in an office, they’ve done their part for the environment. In fact, our measurements indicate that personal computers can account for almost 25 percent of energy consumption of a building, and most of the time, these PCs are turned on but are not actually in use. If you also include servers and data centers, the contribution of so-called IT equipment can be a staggering 50 percent of total baseline energy use, because a lot of the energy is used during nights and weekends when utilization for these PCs and servers tends to be very low.”
The tools available on the Energy Dashboard include real-time power measurement of the entire UCSD campus; energy consumption for each building; and power usage of individual devices such as PCs and servers that are plugged into electrical sockets in some CSE offices. The campus meters are all viewable by the public, but access to the individual meters is currently restricted to the owner of that meter (for privacy reasons).
The Web portal provides statistics updated at least once every minute on total power consumption, power generation, imports from San Diego Gas & Electric, and a comparison between power usage and production. (UC San Diego produces about 82 percent of its annual energy load using 1.2 megawatts of electricity from photovoltaic panels and a 30-megawatt natural gas-fired co-generation plant.) To locate energy-use data on each building, visitors to the Energy Dashboard can select the UC San Diego School of Medicine, Scripps Institution of Oceanography, or any of the university’s six undergraduate colleges (e.g., both the CSE Building and Atkinson Hall are located on the Warren College campus).
“According to some estimates, buildings account for roughly 70 percent of electrical power use in the United States and approximately 40 percent of greenhouse gas emissions,” said Gupta, who is also the associate director of Calit2 on the UCSD campus. “UC San Diego is rapidly becoming an important testbed for technologies to improve energy efficiency, and the Energy Dashboard is an important step toward achieving that goal.”
The researchers were able to identify where peaks in energy consumption came from and the primary components of baseline energy use – including IT’s large energy drain even when computers were not in use (e.g., at night or on weekends when the computers are often left on, just in case the user ever wants to connect in remotely or they are running a background application that requires the machine to be powered on).
“Buildings with a large IT footprint can therefore reduce consumption significantly by decreasing their base energy load,” concluded Agarwal. “Our ability to look at energy use in fine detail gave us greater insight about how to reduce power consumption significantly in these campus buildings. To do that, you have to create effectively duty-cycled buildings.”
To improve the value of data in the UC San Diego Energy Dashboard, they are also working with a private company on a less expensive plug-level meter. Today individual meters that can monitor energy use remotely cost approximately $200 each; Agarwal thinks that if they can get that price down to the $30-$50 range, individuals wanting to track their own carbon footprint will be happy to invest in a meter that would transmit its real-time data to the Energy Dashboard, where the user would be able to use the portal’s tools to track their own usage – and even compare it to the energy profile of a colleague in the next office. “Working with a set of very creative and intelligent students, and leveraging their talent to address some of the energy issues of today, is also immensely satisfying since it feels like you are solving a real-world problem in the end,” said Agarwal. Among the graduate students working on the Energy Dashboard project: Ph.D. student Thomas Weng, a co-author on the November 2009* paper with Agarwal and Gupta.
According to Agarwal, his group is now working on an Energy Dashboard API that will make it possible for anyone at UC San Diego to integrate their own power meter into the dashboard and take advantage of its visualization and comparison features. In the longer term, the researchers are looking into ways to release the API to the larger community outside of UC San Diego, so that anyone with the appropriate energy meter can post, visualize and compare their energy use data on an externally available Energy Dashboard.
* “The Energy Dashboard: Improving the Visibility of Energy Consumption at a Campus-Wide Scale,” Yuvraj Agarwal, Thomas Weng, Rajesh Gupta, First ACM Workshop on Embedded Sensing Systems For Energy-Efficiency In Buildings, November 2009.
“Somniloquy: Augmenting Network Interfaces to Reduce PC Energy Usage,” Yuvraj Agarwal, Steve Hodges, James Scott, Ranveer Chandra, Paramvir Bahl, and Rajesh Gupta. In Proceedings of USENIX Symposium on Networked Systems Design and Implementation (NSDI ’09), April 2009.
“SleepServer: A Software-Only Approach for Reducing the Energy Consumption of PCs within Enterprise Environments,” Yuvraj Agarwal, Stefan Savage, and R. Gupta.
To Appear at the USENIX Annual Technical Conference (USENIX '10), June 2010.
Doug Ramsey, 858-822-5825, firstname.lastname@example.org
A UCSD/UCI PARTNERSHIP > California Institute for Telecommunications and Information Technology Contact Us
home : about us : people : research : partners : education : newsroom : events Calit2 is one of four California Institutes for Science and Innovation
Tuesday, March 30, 2010
Why Cloud Computing Leaders Need to Demand Clean Power http://bit.ly/9jw6Nd
Why Cloud Computing Leaders Need to Demand Clean Power
By Katie Fehrenbacher Mar. 30, 2010, 12:00am PDT No Comments
The launch of Apple’s (a AAPL) iPad this weekend represents a lot of firsts for the tech industry: a device with some of the most media attention of all time, and the start of an $8 billion tablet application market. But the iPad also represents one of a wave of media-consuming mobile devices that increasingly depends on “the cloud” — basically the Internet and data centers — to deliver hosted services and digital content, and will help contribute to a massive growth in energy consumption and carbon emissions associated with so-called cloud-computing over the coming years.
According to a report published Tuesday from the environmental researchers at Greenpeace , the energy consumption and carbon emissions of cloud computing are already significantly higher than previously thought. Using data from The Climate Group’s Smart 2020 report, which came out in 2008 and relied on carbon emission projections from McKinsey, Greenpeace added in the energy consumption info for data centers reported by the Environmental Protection Agency. The result is that Greenpeace says that the energy consumption of cloud computing in 2007 was 622.6 billion kWh, which is 1.3 times larger than reported by the Smart 2020 report.
This new, larger estimate of energy consumption associated with cloud computing emphasizes just how big the problem will be as the sector grows over the coming years. Cloud computing is a trend that has just started (see our Structure 2010 conference) and business-focused cloud computing initiatives like Microsoft’s Azure platform have recently launched. Using the more aggressive cloud computing energy footprint, Greenpeace says that cloud computing will consume 1,963.74 billion kWh of energy by 2020.
All of this isn’t to say that cloud computing companies need to curb their growth. Rather, they need to focus on making data centers and servers more energy efficient and increasingly look to source more clean power. Greenpeace points to Facebook’s decision to build its first-ever data center in Prineville, Ore., which will primarily be powered by coal (GigaOM Pro, subscription required), as a major missed opportunity.
Instead, Internet giants like Google, Yahoo, and Apple should use their energy buying power to demand more access to economic clean power and to support policies that will help drive the proliferation of low-cost renewables. Greenpeace says:
The potential of ICT technologies and cloud computing to drive low-carbon economic growth underscore the importance of building cloud infrastructure in places powered by clean renewable energy. Companies like Facebook, Google, and other large players in the cloud computing market must advocate for policy change at the local, national and international levels to ensure that, as their appetite for energy increases, so does the supply of renewable energy.
We’ll be looking at the issues of energy consumption and the carbon footprint of information technology, data centers and servers at our Green:Net conference. Google’s Green Energy Czar Bill Weihl will be discussing some of the search engine’s industry-leading green data center work, and Greenpeace’s Casey Harrell, one of the authors of the report, will be discussing how the Internet leads to dematerialization, or replacing atoms with bits.
Tuesday, March 23, 2010
Australian govt sets out ICT carbon reduction targets
Australia’s Finance Minister, Lindsay Tanner, has reportedly laid out a target to cut roughly 13% of the carbon emissions from its data centre operations over the next five years.
According to this report by ITwire,Tanner told a conference at CeBIT that the Australian government is the largest data centre operator in the country - larger than the country’s four big banks combined.
The goal is to reduce the estimated 300,000 tonnes of emissions annually today by 40,000 tonnes on an annual basis in five years, Tanner said.
Under a 15-year data centre strategy announced by Tanner, all departments and agencies will have to measure and report the energy consumption of their data centres and ICT infrastructure annually.
Tanner added that future government procurement of data centres will put a major consideration on the ‘green credentials’ of the site and infrastructure. The locations of data centres, as well as other contributing factors, such as free air cooling, and access to telecommunications and power infrastructure would also play key parts in the decision making process. The new procurement parametres will come into effect in the second half of the year.
Wednesday, March 3, 2010
Joulemeter: VM, Server, Client, and Software Energy Usage
Joulemeter is a software based mechanism to measure the energy usage of virtual machines (VMs), servers, desktops, laptops, and even individual softwares running on a computer.
Joulemeter estimates the energy usage of a VM, computer, or software by measuring the hardware resources (CPU, disk, memory, screen etc) being used and converting the resource usage to actual power usage based on automatically learned realistic power models.
Joulemeter can be used for gaining visibility into energy use and for making several power management and provisioning decisions in data centers, client computing, and software design.
For more technical details on the system here is their paper.
Virtual Machine Power Metering and Provisioning
Aman Kansal, Feng
Zhao, Jie Liu
University of Southern
Virtualization is often used in cloud computing platforms for its
several advantages in efficient management of the physical resources.
However, virtualization raises certain additional challenges, and
one of them is lack of power metering for virtual machines (VMs).
Power management requirements in modern data centers have led
to most new servers providing power usage measurement in hardware
and alternate solutions exist for older servers using circuit and
outlet level measurements. However, VM power cannot be measured
purely in hardware. We present a solution for VM power metering.
We build power models to infer power consumption from resource
usage at runtime and identify the challenges that arise when
applying such models for VM power metering. We show how existing
instrumentation in server hardware and hypervisors can be
used to build the required power models on real platforms with low
error. The entire metering approach is designed to operate with
extremely low runtime overhead while providing practically useful
accuracy. We illustrate the use of the proposed metering capability
for VM power capping, leading to significant savings in power provisioning
costs that constitute a large fraction of data center power
costs. Experiments are performed on server traces from several
thousand production servers, hosting Microsoft’s real-world applications
such as Windows Live Messenger. The results show that
not only does VM power metering allows reclaiming the savings
that were earlier achieved using physical server power capping, but
also that it enables further savings in provisioning costs with virtualization.
Note there will be a desktop and laptop version available soon.
Download: A freely downloadable version of the Joulemeter software that measures laptop and desktop energy usage will be be available in a few weeks. Watch this space!
- November (4)
- October (6)
- September (2)
- August (4)
- June (3)
- May (6)
- April (2)
- March (5)
- February (3)
- January (2)
- December (3)
- November (5)
- October (6)
- September (8)
- August (5)
- July (10)
- June (3)
- May (5)
- April (9)
- March (4)
- February (4)
- January (3)
- December (7)
- November (10)
- October (7)
- September (9)
- August (10)
- June (4)
- May (9)
- April (6)
- March (4)
- February (4)
- January (10)
- December (3)
- November (4)
- October (7)
- September (4)
- August (3)
- July (3)
- June (2)
- May (9)
- April (5)
- March (4)
- February (4)
- January (7)