News & Analysis
Smart grid hits snag over powerline standard
Rick Merritt
11/30/2009 12:01 AM EST
SAN JOSE, Calif. — The U.S. government is grappling with one of the first big snags in the wake of awarding $4 billion in grants to build a smart electric grid. It could take years before there are any low cost appliances for Joe Consumer to plug into an intelligent power network.
The vision of a smart grid includes smart appliances that automatically turn on or off in response to fluctuating energy prices as electric demand peaks and troughs. Intelligent fridges, dryers and other energy hogs could help utilities reduce their needs for new power plants, help consumers save money and ease stress on the environment.
To enable this so-called demand response application, appliance makers need an easy, low cost way to plug into the grid. Today they face as many as a dozen wired and wireless choices, most of them far too expensive and high bandwidth, focused on carrying digital music and video around the home.
In an effort to fill the gap, a senior government technologist laid down the gauntlet before a recent gathering of powerline networking engineers: create a standard to plug appliances into tomorrow's smart grid soon--or Uncle Sam may do it for you.
Industry representatives at the meeting said the government is over-stepping its boundaries. Forcing the handful of powerline technologies in the market to converge makes no sense, they added.
Government planners knew a lack of standards was one of the big issues preventing the move to a digital, networked power grid. Before its first economic stimulus grant went out, it spent $10 million to launch a new smart grid standards effort organized by the National Institute of Standards and Technology.
NIST is now reviewing industry feedback on a first draft framework for smart grid standards. As a next step, NIST convened in Denver in mid-November a so-called Smart Grid Interoperability Panel of diverse stakeholders to drive the standards work forward.
SGIP includes fifteen Priority Action Plan committees focused on some of the thorniest standards issues ahead. At the first meeting of the PAP-15 group focused on powerline home networks, the national coordinator of smart grid standards at NIST, George Arnold, gave the group his ultimatum.
"It was a pretty competitive group to put in one room," quipped Arnold. "I almost thought I would have to hire a security guard."
In the meeting Arnold said NIST could use its experience selecting the AES security standard as a model for creating a technology bake-off for a low cost powerline standard for smart appliances.
"The moment where he most needed a security guard was when he said if you guys can't get together [on a standard], NIST will decide," said Stefano Galli, a lead scientist at Panasonic R&D Co. of America who attended the meeting. "There was an uproar.
"Big companies feel this weight of a government decision is not beneficial," said Galli who also co-chairs an IEEE task force exploring communications standards for the smart grid. "A premature convergence of the home network could create more problems than it solves," he added.
"I would rather let the market decide, and I think the 70 companies in the HomePlug Alliance would rather have the market decide, too," said Rob Ranck, president of HomePlug which has released multiple generations of home powerline networking standards. "We've met with [regulators], the Department of Energy and Congressional staff members and there doesn't seem to be a clear consensus that [picking a standard] is NIST's role," he said.
"If you go back to the legislation that supports NIST, it talks about interoperability, but it's not clear whether it sets up the federal government to pick a single winner," Ranck added.
The industry has tried and failed for years to set a single powerline home networking standard. After four years of work, the IEEE 1901 group is about to finalize a standard that essentially blesses multiple powerline physical layers and media access controllers.
The stakes are as high as the difficulties. The government is keen to ensure its $4 billion in recent smart grid grants is money well spent. Many of the projects include pilot projects in demand response systems.
White goods giant Whirlpool Corp. made a high-profile promise last fall that it will ship in 2011 a million dryers ready to plug into the smart grid—if there is a suitable networking standard the company can use.
"The last thing I want to see is a Wall Street Journal article a year from now saying Whirlpool had to renege on its promise because of the lack of a standard," said Arnold.
From NIST's perspective, "the ideal would be to have one wired and one wireless standard," Arnold told EE Times.
Members of competing Wi-Fi and Zigbee trade groups also attended the meetings in Denver. Some expressed they disagreed with the idea of a single wireless standard—even if their approach was picked.
Comments
JMWilliams
11/30/2009 2:40 PM EST
There is a major problem with using the power grid itself to transmit control information: Most service entrances collect noise, and users requiring quiet power will very conveniently remove the noise at the entry point by means of a line filter such as a ferrite core clip-on.
However, removing noise from the line also would remove the power control signals discussed in this article.
A better plan would be to use an FCC reserved frequency band to broadcast the control information independently from the power. This would not require any capital expenditure or other accommodation by the power provider and could be incorporated by electrical equipment manufacturers at their leisure. Furthermore, immediate adoption of smart-grid Federal standards could be implemented; if changes later were found necessary, a different frequency band or encoding scheme could be defined.
Decoupling of the control from the power supply obviously has many advantages, as would decoupling of the control hardware at the receiver end from the specific frequency and encoding standard being used for broadcast.
In addition, RF can have unpleasant biological effects on humans such as tinnitus; there would be some liability incurred by any power company adding RF, especially digital RF, to its power lines.
If the control was broadcast separately, sensitive individuals and equipment could be shielded easily, perhaps at the cost of having to ignore the available "smart-grid" features.
Globbing everything in the "smart grid" into the power lines doesn't seem to be a very smart idea.
Sign in to Reply
KenKrechmer
11/30/2009 3:18 PM EST
The desire of NIST for a single wireless and wireless standard reflects a very old fashioned view. When interfaces are computer controlled and memory is very low cost, there is no technical reason to have a single standard. All that is needed is an adaptability mechanism that allow each side of the interface to identify, negotiate and select which compatible interface is to be used in common.
NIST should be focusing on standardizing adaptability mechanisms not compatible interfaces.
Sign in to Reply
jonsmirl
11/30/2009 9:16 PM EST
There is no need to pick a physical layer. Pick them all and then implement IPv6 over them. Put a USB stick into a wifi router or have a little bridge device. If people won't pay $20 for a bridge they probably don't care enough to make use of the data either.
Sign in to Reply
embeded
12/1/2009 4:43 AM EST
I live in Boulder, CO, were the power utility (Xcel) has just installed Smart Grid to the home. In effect, they just wired the whole city with fiber to the home. With the proper adapter, my toaster has the ability to talk to Xcel's Smart Grid network at a faster rate than my home network (which is only cable, no fiber Internet connection is yet available in Boulder and probably won't be for years).
I'm not sure why fiber bandwidth is required. How often do they intend to sample the power usage of each appliance? If each appliance just sends a simple "I'm on/off" message, the total bandwidth should be minimal, even when aggregated over all the homes in Boulder. Does Xcel want to know every time the appliance changes its power usage even by a little? What's wrong with just recording the aggregate power usage for the home? The fact that fiber is needed implies they will be collecting a *lot* of data, and possibly sending out a *lot* of control messages. Just the idea that Xcel will have a database of every major appliance in Boulder and its usage pattern is not a pleasant one -- especially since they have demonstrated a willingness to share that data with the city government by already giving the city each home's total electricity usage for the month so it can be taxed for the city's "Climate Action Plan".
As I understand it, powerline will only be used inside the home to the smart meter connecting your house to the Smart Grid. I've used powerline adapters before, and found that they are indeed somewhat quirky. In some rooms I've gotten 100 Mbps, in others just 13 Kbps (yes, K). These were $150 units which presumably could afford some
means to minimize noise.
Right now, the city's Smart Grid efforts are stalled by a lack of devices to connect appliances to the Smart Grid. I understand there are a few clunky prototypes used in demo homes, but nothing commercial yet. I am *very* reluctant to invest in something that will support an early, obsolete standard. If the industry settles on a standard that Boulder's preliminary Smart Grid installation does not support, nobody will be supplying appliances that support Boulder's Smart Grid. I cannot emphasize enough how important it is for the industry to standardize before a lot of Smart Grid is installed. Appliance makers cannot support a bunch of different standards in a cost effective manner. The truth is that homeowners operate on much smaller budgets than businesses which can support the higher costs of flexibility.
Some have said that converters should be software upgradeable, but that just adds significant cost. Ideally, every major appliance will have a converter, and if it's not around $20 each, the cost will be prohibitive. This is especially true when you have to overcome the cost of installing the Smart Grid infrastructure itself. Boulder's installation cost is around $1000 per residence. Another article on this site says Smart Grid will provide only 4% savings when fully implemented. That means you don't break even on installation costs until you use $25,000 worth of electricity, which can be 20 years or more. Using similar reasoning, even a $20 Smart Grid adapter for a major appliance won't pay off until that appliance uses $500 of electricity. For most appliances that will be several years, or perhaps not even over the lifetime of the appliance.
Sign in to Reply
Rick Merritt
12/1/2009 4:37 PM EST
So that's three strong votes for generic adapters.
Any other options?
Sign in to Reply
KenKrechmer
12/1/2009 6:34 PM EST
Generic adapters can work if there is a protocol to negotiate which of the different modes of operation is to be selected/downloaded. Such a protocol I term an adaptability mechanism and it requires standardizing.
Without such an adaptability mechanism the user is stuck trying to figure out which of the different modes of communications work in a specific installation.
Sign in to Reply
Ginger45
12/2/2009 3:26 PM EST
Honestly, I'm not willing to invest one cent in extra gear to support smart grid. Projected savings of 4% nationally by 2030 does not warrant any extra investment on my part. We, the consumers, are already paying for deployment through stimulus funds (taxes) and I'm sure the utilities will find a way to charge us for the cost of upgrading the grid. In today's dollars I would save about $5/month in electrical bills (based on the 4% estimate), but the net change in my month costs will likely be positive. Hardly worth the effort!
Sign in to Reply