The day is sweltering, air conditioners are cranked up, and the power grid is straining to meet demand. Today is a "needle peak" day -- on the annual power demand chart, it shows up as a spike. Out of the year's 8,760 hours, needle peaks will occupy 200 hours or less. An extreme day like this is why the grid maintains roughly twice as much power generating and transmission capacity as it uses on an average day. Even though power plants and lines are idle most of the year, this costly overbuilding is needed to cover all contingencies. The grid is built to be there "just in case."
But what if another power resource were available that could dramatically reduce that peak demand, one that involved generating and transmitting no power at all? No, this isn't some weird "zero energy" thing. The paradoxical sounding resource I'm talking about is already in use. It's the demand, also known as the load, itself. The basic idea is that the grid can meet overall needs not only by supplying power, but by adjusting power use. The word for this is demand response, and it's a fundamental aspect of the smart grid.
On the old "dumb" grid, information flow from power users to suppliers consists almost entirely of 12 meter readings a year; from suppliers to users, it is 12 power bills. One of the most profound changes introduced by the smart grid -- indeed, what makes it smart -- is a communications backbone that allows massive two-way information flows. An information network is overlaid on top of the power network. Demand response (DR) employs these information/communications capabilities to engage power users directly in managing the grid. In essence, information becomes a new power resource.
DR is one of the first pieces of the smart grid to emerge. Since the 1980s, utilities have worked with customers to place automatic controls on water heaters, air conditioners, and other electricity-hungry devices to lower demand during peaks. Florida Power is a good example, with nearly half a million customers in its program. Utilities work with farmers to control irrigation pumps during peaks. Many grid operators also engage large industrial and commercial customers to reduce load on those super-hot or super-cold days. Sometimes demand response is directly controlled by the utility, but often someone in the building literally goes around and turns off switches.
Typically, the utility provides credits on bills as an incentive to cut loads. From the utility's standpoint, it's a complex calculus, balancing DR costs with the avoided costs of serving peaks with standard generation and transmission.
What's changing today is that the deep decline in costs for computing technologies and communications bandwidth is changing the calculus. DR is becoming easier and more economical. At the same time, the emergence of Independent Systems Operators (ISOs) to run regional power transmission systems is creating vibrant and large new markets for DR. They provide a level playing field where DR can compete with traditional power resources. Companies such as EnerNOC, Comverge, and Energy Connect are aggregating DR loads from multiple customers and marketing them to ISOs, particularly PJM Interconnect, a notably innovative ISO serving the mid-Atlantic states. These companies are offering demand resources equal to large power plants -- 796 megawatts in the case of EnerNOC, 948 for Comverge.
DR comes in several flavors, and not all are as sweet for utilities. The basic division is between "firm," meaning that customer-end equipment is under a form of direct load control (DLC) that makes it completely predictable, and "non-firm," meaning the load is under customer control. Utilities rely on non-firm DR to reduce costs during peaks and carry them through emergencies. But utility engineers will not cancel a beefed up transmission line in favor of DR unless the load is firm, or "fully dispatchable." They want to be able to basically turn it on and off themselves. And you can't blame them; if the lights go out, they're the dogs.
DR should not be confused with its close relative, energy efficiency (EE). EE seeks to reduce overall power use. DR aims to reduce use at specific hours. Sometimes that means an absolute cut in electricity consumption. For instance, when an air conditioner is cycled down during the day it will not necessarily return to full operation in the evening. But when, say, a hot water heater is turned down, it typically will shift the load to later.
That sets up a potential unintended consequence. DR could actually increase overall pollution. For example, if generation is shifted from daytime hydroelectric or cleaner gas generation to evening coal, then overall emissions could rise. So in designing DR markets, this pitfall needs to be avoided. The New England Demand Response Initiative, a multi-stakeholder group, made sure to take a look at the impacts there and found a small overall emissions reduction benefit on the New England grid from DR.
DR provides some clear environmental benefits. It can serve as a substitute for spinning reserve -- power plants that run ready to supply power on short notice, typically around 10-15 percent of overall power generation. The less spinning reserve, the fewer emissions. And DR could sharply reduce the need for peaker power plants and infrastructure, with all their embedded energy and land-use impacts. Pacific Northwest National Laboratory (PNNL) calculates that moving to smart-grid technology will eliminate the need for between $46 and $117 billion in conventional utility infrastructure. That does not count investments in new smart grid technology. But one PNNL calculation gives an indication of comparative costs: smart appliances that can adjust their demand to grid conditions could, for $600 million, provide reserve capacity equal to power plants costing $6 billion, proving that "bytes are cheaper than iron."
DR has some additional potential for promoting a greener grid, and that potential illuminates key aspects of the smart grid. Advanced DR is built on smart systems that not only control power use by individual devices, but also provide detailed information on power use down to the device. They will let utilities assemble power use data with unprecedented detail. This knowledge can be leveraged for economic value in a number of ways. One is to validate the actual effects of EE programs.
Rob Pratt with PNNL's GridWise Program gives an example. Say a utility pays for efficiency improvements at a number of houses, and seeks to use the resulting emissions reductions to gain carbon market credits. As more regions move to carbon cap-and-trade systems, such credits will become more valuable and important. With today's primitive information flow, these reductions could only be approximated, not verified, so gaining credits could be difficult and the utility would have less incentive to make the investments. But validated information can be taken to the carbon marketplace, so the EE is more likely to be done. This is just one of a number of ways in which DR and EE are synergistic. Future postings will delve more deeply into these interactions.
DR might also provide cost-effective, emissions-cutting means to balance intermittent renewable energy sources such as wind. Wind generation varies with availability and intensity of the resource. So wind farms are partnered with reserve power plants that fill the gap when wind speed diminishes. Often these are natural gas turbines, though in the Pacific Northwest hydroelectric dams are employed.
But what if a smart grid could automatically balance wind with adjustments in power demand? Thinking on this is still at early stages, but it is theoretically possible. Since wind varies minute to minute, the system would have to be quite sophisticated and have a broad diversity of demand resources to tap. Fraunhofer Institute modeled such a system for Germany and found "the costs of additional reserve power could be reduced ... demand response may be a valuable option for integrating wind power into electricity systems." Of course, this would also reduce greenhouse emissions by balancing fossil resources.
The application of smart technologies to the grid makes a largely one-way flow of power and information into a far richer and complex two-way stream. The greater the penetration these technologies achieve, the more power demand will become a power resource as well.
Do we need nuclear and coal plants for baseload power?
Ask Umbra on climate-skeptic teachers, low-flow toilets, and more
To unlock wind power, put a price on carbon
Comments
View as Flat
Gar Lipow Posted 8:17 am
27 Jul 2007
Similarly you are a business. You make widgets. Are you really willing to send your workers home for the day and bring them back to catch up and night on production? That had better be some rebate!
Again storage can take care of most of these. Compressed air, heat and cold can all be stored fairly cheaply. But the study is listing flexibility in addition to that storage. If you've already used the heat or cold or compressed air that you produced when the wind was strong or when production was off peak, I'm having trouble visualizing just doing without services. Please, tell a story; give an example.
Thanks
Gar
Permalink
Gar Lipow Posted 8:24 am
27 Jul 2007
Permalink
Sean Casten Posted 8:31 am
27 Jul 2007
Permalink
Sean Casten Posted 8:34 am
27 Jul 2007
Permalink
rmcleod Posted 8:38 am
27 Jul 2007
May I also politely suggest that you drop the acronyms as well? Defining a new piece of terminology, and then obscuring it with an acronym makes everything a little less clear. You don't need them, they break up the pace of the piece, so why bother?
--
entropyproduction.blogspot.com
Permalink
Sean Casten Posted 8:53 am
27 Jul 2007
Permalink
Patrick Mazza Posted 9:24 am
27 Jul 2007
Sean's response gets to the customer end stuff - In many cases the response will be barely noticeable, and it will be automated. So the clothes dryer might keep spinning but the heat will turn off - pilot smart dryers are already doing this, with web tools that allow overrides. A commercial HVAC system turns temperatures up or down slightly. A buiding energy management system dims the lights just a bit. The key is aggregation of a lot of such responses. Some demand is inflexible and cannot participate.
Aah - acronymns - occupational hazard of energy wonks.
Patrick Mazza
Permalink
Gar Lipow Posted 9:26 am
27 Jul 2007
Right. I'm trying to view the example from the point of view of someone living with it. Am I paying extra for a dryer that will stop drying my shirt right when I have a meeting to go to? Or lets use a smart dishwasher instead which will stop washing my dishes just before I serve a meal? (Cause dryers cost energy; automatic dishwashers used properly save energy compared to sink washing.)
In terms of the business cases. Yeah I guess we have to assume that batch processes will only be delayed if there is other lower energy work that can be done while waiting the extra time. But in terms of CA and EAF I'd feel better with a few details. What are the other processes these industries can deploy workers to. It is interesting that the manufacturing process is that flexible that you can change the order in which tasks are done.
BTW storage also can be part of demand management. And the study does specifically mention it. Heat, and cold can both be stored less expensively than electricity. Possibly compressed air can too, though I have not seeing figures for it. (Upsizing your compressor motors and storage tanks is not cheap - so Im not sure on that point.) So in terms of cooling and air condition you run chillers at night, storing the the results in ice balls or PCM or whatever. (This by the way is demand reduction as well as shifting in any case where day and night temperatures differ by a large enough factor.) Similarly with heat, there are all sorts of low cost thermal mass you can add. And there is no reason a sophisticated storage system cannot be under the control of the electric company as to when demand happens. That is your smart appliance signals "I'm going to need x kWh within the next 24 hours", and the utility choose when within that 24 hour period it is supplied.
Permalink
ffletcher Posted 10:44 am
27 Jul 2007
Gar is correct, storage is a cool approach. There are load side storage approach as well. For example, St. Joesph Medical Center here in Burbank recently built a Cancer center. In that building they incorporated thermal storage so that the utilty could shift the thermal load off the peak. The cost to put that storage in during the design stage of the building was only $125,000. For the 1000 kW of storage that provides no supply side approach can equal such a low cost.
Turning swimming pool motors off during peak load periods is another effective means to reduce peak loads. We also shift water pumping to off peak periods.
The Smart Grid can help better identify the loads during peak periods so those loads can be better understood and reduced. I think the time for smart grids has come. We are begining the development of a Smart Grid system that uses Wi-Fi technology to read the meters and control the devices.
Permalink
Gar Lipow Posted 10:52 am
27 Jul 2007
Permalink
Gar Lipow Posted 11:11 am
27 Jul 2007
I don't think you are going to get a high response with just price signals. I think you will have to require all appliances (at least those with real DR potential) to have ability to implement it, and make DR enabled the default setting for those appliances. People will then be allowed to switch appliances off DR if they want to. In affect - make DR opt out.
Whether this is a good thing to do is another question. But I think if you want large numbers of residences to buy into DR based on usage rather than just that implemented through storage, you will need something on those lines.
Permalink
ffletcher Posted 3:19 pm
27 Jul 2007
I continue to seek supply side storage options, but the demand side options are real as well. It is amazing how many 1970 era air conditioners we have in town. One can easily shave 6 to 7 kW off the peak of a single house (3000 square foot house) by upgrading to current state of the art, and that is without even touching the insulation.
An example of AC storage is the Ice Bear a residential ice storage unit. http://www.ice-energy.com/ , these units allow the air conditioner to be turned off during peak periods and just circulate cool air using air. The cost of these units is about $3000 per kW and it can reduce losses on the lines.
With a smart grid one can determine where these units are most needed and target those places. But one needs the information available via a real time metering system rather than the read once a month method we employ today.
Permalink
trock Posted 5:19 pm
27 Jul 2007
What was learned in California when it had all those Power supply problems. What I think I remember happening was after a year into it, they (Cali. citizens) were able to reduce eletricity usage by 10 to 20 percent. Demand Response or Demand Reduction or just Demand Curtailment because they would be down if they didn't turn lots of stuff off anyway?
How much happens when a Public Service Annoucement is made, like scrolling during TV shows or radio annoucements; or does that piss some off so they turn things on?
Permalink
Pangolin Posted 6:34 pm
27 Jul 2007
Refrigerators-there must be several million obsolete or damaged refrigerators in California alone. First, regulate refrigerator production so there is some modularity and doors, cabinets and compressor units can be swapped out. Then simply replace old units with new units and finance them through metered power bills. This problem is even worse for small retail refrigeration units many of which are 30 years old or older.
Geo-exchange HVAC- require utilities to finance upgrades to geo-exchange units through power bills at the same interest rates they pay on bonds. On a dollar return basis these upgrades are far cheaper than any type of installed power genaration and reduce grid demand AND natural gas demand. Drive around the cheaper section of your town and you will see battered old gaspak HVAC units sitting up on black roofs.
Roofing- Tax the crap out of high thermal mass roofing (spanish tile) and non reflective roofing. The fact that builders every day in California and the South install black roofs is idiocy. In Palm Springs of all places spanish tile is the preferred roofing material. That's several tons of hot rock on top of your house that will never cool down. A metal roof cools down many hours before a tile or asphalt roof does. (thermal MASS) That means your insulation can stop absorbing heat and start dumping it.
Lighting- A residential rental or sale property in should have exactly one incandescent light; the one above the stove. Just this single requirement to transfer residency of property would save enough power to permanently close power plants.
The pro-nuke idiots think we should just go on running our AC's in glass-walled black-roofed buildings and leave a few doors open to boot. Hell we can just build more nukes. It's far easier to just adjust tax rates on certain building materials and replace some damaged appliances. We can get better services from our buildings from smaller power inputs. It's off-the-shelf except for the financial packages to make it work.
Put the Carbon Back
Permalink
trock Posted 11:52 pm
27 Jul 2007
Is there also Demand Delay? I have always thought that Refrigerators/Freezers should have cold storage so that they can go, what, 6, 8 or 12 hours without needing power. They could do it daily from noon to 6pm or 10am to 10pm.
Has someone done the cost effectiveness of storing cold in Refrig/Freezers and that of air conditioning? I imagine air conditioners storing cold would make a bigger power savings, but the sheer numbers of Refrig/Freezers would make a contribution.
Also, has it been studied to do water preheating with the Refrig/Freezer. Run water to the Refrig/Freezer, have a heat transfer from the condenser, and send that to the water heater in the house. Not cost effective? Water going into a Refrig/Freezer at 50 degree better than air at 70 degrees for efficiency? (some Refrig/Freezers get water for ice cubes anyway)
As I've read, Refrig/Freezers can be made more efficient than even new ones now. Is it more cost effective to go for extra levels of Energy Efficiency or put in Demand Delay, neither or both in Refrig/Freezers? Has it been studied?
Permalink
sunflower Posted 12:31 am
28 Jul 2007
We have a power sensor that shuts down discretionary loads, like electric hot water and electric heat when a hair dryer or electric stove increase demand. The house is configured to never draw more than 30 amps.
The cold water coming into the house is preheated in the floor, making cold water warmer, and preheating hot water. The SunFrost refrigerator is built into the wall and surrounded by thick Styrofoam (expanded polystyrene). We have an air to air heat exchanger so that the stale warm air vented from the house (and shower) is used to heat the incoming fresh air. We have shutters on windows to block night heat loss and day heat gain. We also have thermal mass to level the home temperature by holding the cool from night and the warmth from day. And we have 20 kW solar passive gain and NO CLOTHES DRYER. It is just nuts to use coal fired electricity to dry clothes.
All that benefits my power consumption and load requirements. From a power utility perspective, timers and a smart grid would benefit power supply load requirements.
Permalink
spaceshaper Posted 1:53 am
28 Jul 2007
Better services from our buildings (effectiveness). Not just lower heating/cooling costs per square foot (efficiency). We can indeed live better using less energy, and effectiveness-thinking is the key. This is even more true of our transportation issues - we must start thinking of our degree of access to needed resources rather than miles traveled per unit energy as our measure of transportation value.
Sunflower - "Internal demand reduction... The house is configured to never draw more than 30 amps."
This is really a great concept. With buildings configured to even out their electrical power demand internally, the grid, smart or not, becomes much easier to manage. Material savings too - less copper and aluminum in the supply line. And we can do this incrementally, now, one building at a time, while waiting on large-scale grid improvements.
The true meaning of life is to plant trees, under whose shade you do not expect to sit.
Permalink
Gar Lipow Posted 3:47 am
28 Jul 2007
Cutting demand through increases in efficiency (including large increases in effectiveness) is important. But being able to tweak when that demand occurs (without decreasing effectiveness) helps reduce the capital costs of meeting what demand remains.
Permalink
Patrick Mazza Posted 4:16 am
28 Jul 2007
Ice storage in commercial buildings is actually an option for demand response and peak load shaving. It lets building owners store cool by making ice on the off-peak and then use it on peak, also allowing further cycling down when a demand response is asked. I've seen a study done for PacifiCorp that looks at it and finds its still a relatively high cost DR resource. But ideas like this are definitely part of the picture.
Price response probably will not be enough - that's right. A bill working its way through Congress would require appliances to be built with smart chips that can offer demand response. Costs will be nominal. More on this legislation in future posts.
During the 2001 West Coast power meltdown significant demand response saved the grid from collapse. The voluntary response was hugely effective, though this would be classed as non-firm and so is not a substitute for standard utility infrastructure. In the Northwest, Bonneville Power Administration operated a the Demand Exchange system which supplied hundreds of megawatts in demand response. BPA notified customers a day ahead when grid steess was expected - a hot day. Customers who optedc to supply reductions gained bill credit. In some cases here, yes, workers were sent home. But this was an emergency situation.
Sunflower and Pangolin - hugely great ideas and practices. These go under the heading of energy efficiency, and demonstrate the immense potentials we have to reduce electricity demand. To promote more of this we need to re-shape the standard utility model into an energy services business model that gives companies positive incentives to provide customers with efficiency as well as electricity. The Delta Montrose Coop in Colorado, by the way, finances geoheat exchange installations for its customers, similar to the model Pangolin suggested. The old utility model of profit purely or mostly by kilowatt throughput is one of the toughest nuts to crack in bringing on the smart grid and efficiency, and I will post on this in the future
Patrick Mazza
Permalink
spindarella Posted 5:31 am
28 Jul 2007
LOVE MOM AND GRAMMA !!!
Permalink
spaceshaper Posted 11:26 pm
28 Jul 2007
Here is the distinction I would offer. Efficiency-thinking looks at the performance of system components. The arrest record of a beat cop. The successful procedures carried out by a hospital surgical team. The number of miles traveled on a gallon of gas. The number of tons of corn produced on an acre of soil. I expect you can see what I am getting at. The efficiency focus in these various areas, as thoughtful as it may be, gives us no measure of the public safety of a community, or its physical health, its or whether its transportation and nutritional needs are being properly met.
Effectiveness-thinking by contrast looks at total system performance. Effectiveness rates freedom from crime, not the percentage of crimes which are punished. The U.S. leads the world in medical intervention efficiencies, yet its total healthcare delivery performance whether measured absolutely or per unit cost is poor and getting worse. And while our transportation and nutritional failures have been much discussed, even here on Grist the focus has been miserably often on the joys of 200 mpg hypercars rather than on true transportation performance. Looked at from an effectiveness standard, a reasonable transportation effect might be described as ensuring convenient, safe, affordable access for all sectors of the population to the resources they need for such ordinary purposes as employment, education, shopping, recreation, entertainment, and socialization. From that perspective our automobile-based system with its poor safety record, questionable convenience and numerous exclusions from participation looks pretty crude, however efficient the component vehicles might be.
It's worth remembering too that systems which contain only efficient components can nevertheless be ineffective because essential components are missing - the efficient police department with no mandate for community involvement will have no opportunity for early intervention with at-risk kids and thereby fail in its public safety effectiveness. It can also be the case that a system can be extremely effective while containing major inefficiencies - traditional agriculture for example, with its crop rotations, hedgerows and fallowings. I make no apology for mentioning once more the red oak tree in my yard in this connection. Root, stem and branch are a miracle of structural efficiency - the tree "knows" just where to put the right amount of material to achieve with minimal input its tall columns and far-reaching cantilevers. Its photosynthesis operation is efficiently developed fresh every year in the spring and jettisoned in the fall to refertilize the soil. It pumps hundreds of gallons of water a day eighty feet into the air, using little energy and no moving parts except the life-giving fluid itself. And yet its reproductive facility is extravagantly, exuberantly wasteful - a hundred thousand acorns a year, tens of millions in an adult lifespan, rain from its branches so that just a few may germinate into saplings and thus continue the cycle.
If the red oak can afford some inefficiencies and still be effective in continuing its species presence on the earth, so can we (I'd count music, dance, art, sculpture amongst the glorious inefficiencies we can and should permit ourselves). But by nature's iron rule the price of ineffectiveness is, ultimately, death.
The true meaning of life is to plant trees, under whose shade you do not expect to sit.
Permalink
Gar Lipow Posted 3:06 am
29 Jul 2007
Another point is that we (meaning everyone )create the future, but we don't control it. Especially the narrower "we" of people for environmental concerns are a high priority don't control it - nor should we.
Given comfortable, reliable trains along the line of Cybertran, the larger we may choose a society that is close to automobile free. Or they may choose a slightly less automobile dependent society that still uses cars as the primary means of personal transportation. It is better to offer multiple dimensions in solutions, so that whatever mix we end up with is sustainable.
I will add that especially in blog format you can't always be putting everything in larger context. I'm trying to get out of the habit of writing 2,000 word posts - which means not dealing with walkable cities or trains every time I write about cars, not writing about efficient cars every time I write about trains.
Permalink
spaceshaper Posted 8:25 am
29 Jul 2007
And if we want to have a chance of averting that extremely dislocated future by stopping the runaway express of human-induced climate change we face a yet greater challenge within a rapidly shrinking timeframe. In that case it's absolutely essential to think of effective solutions, not just efficient ones.
The true meaning of life is to plant trees, under whose shade you do not expect to sit.
Permalink
sunflower Posted 9:01 am
29 Jul 2007
Permalink