Demand-Side Management Technology Avoids Grid Construction for Bonneville Power (Case Study)
To understand the problem of grid constraints, imagine you're an ice cream maker. It's a hot summer afternoon and you get orders for every last gallon of your cold treats. Now imagine your delivery trucks get caught in a traffic jam in your own shipping yard. You have perishing products, complaining customers, and a retailer revolt -- not to mention the mess. Peak demand management is like keeping your customers from placing their orders all at once.
April 28, 2006
Bonneville Power Administration markets the power generated by 31 hydroelectric stations in the northwestern United States. The electrons produced by water and gravity are whisked along a 15,000-mile network of BPA-owned transmission wires.
Bonneville's wires carry the electricity to 149 utilities like the one in Ashland, Oregon, which resell the power to their residential and commercial customers.
Constrained grid paths cause energy traffic jams
BPA's transmission system is part of the national electric grid. Some areas of the grid are reaching or exceeding their capacity, not just in BPA territory, but nationwide.
Gird constraints create situations similar to an energy shortage. In this case there is plenty of supply, but during peak hours the demand can exceed the ability of the transmission system to carry enough electrons to their destinations. When bottlenecks happen, utilities can't get enough power from BPA to serve their customers.
In the worst cases, grid overloads can lead to incidents like the infamous east coast blackout on August 14, 2003 that turned out the lights on 50 million people and cost an estimated US$6 billion. Blackouts like it affected Chicago, Delaware, Atlanta, New Orleans, San Francisco and Detroit, all in the span of two years.
Alternatives to unwanted new power lines
Expanding the grid is one option, albeit costly and politically charged. The Energy Policy Act of 2005 advocates building more transmission lines. (See sidebar.) There are many alternatives to new line construction, though, and BPA is actively investigating them.
BPA formed a group to study options that might help the agency avoid the expense of building more transmission lines. The group, named the Non-wires Solutions Round Table, identified likely alternatives and began experimenting with them through pilot tests.
There are several alternatives to line construction, based on being smarter about using the electrons we can fit through our existing wires. One alternative is conservation: simply doing without a small portion of electricity and the service it provides. Another is efficiency: enjoying the same services, but with a smaller amount of electricity. Turning off a light is conservation; installing a compact-fluorescent light bulb is efficiency.
A third alternative is based on changing when energy is used, or shifting loads into different times of the day. The problems from constraints in BPA territory occur during hours of peak demand, usually hot summer afternoons and cold winter mornings. Several programs are in place across the nation, designed to level the load on the grid by reducing these peaks. Consumers use the same amount of electricity, and enjoy the use of the same appliances, but slightly earlier or later than usual.
Demand-side management solves the problem -- by remote control
BPA identified a technology called GoodWatts that allows a utility to monitor and control appliances remotely. The technology, from Invensys Controls, was on the market and available, which made BPA's job of pilot-testing easier.
The energy utility in Ashland Oregon participated in a test of this technology. They weren't just interested in helping BPA avoid construction costs. Ashland gets almost all of its power from BPA and pays steeper wholesale rates when demand is high.
"Realistically, if we could control people's appliances to reduce our peak demand and lower our wholesale power bill from Bonneville, then everybody in the city would be better off," says Ashland's Dick Wanderscheid.
BPA recruited 100 residential test customers to participate in the pilot. The utility offered them a $100 rebate for each year of their cooperation. Ashland installed the GoodWatts equipment at each location, and BPA operated it over the internet.
How peak demand management worksThe key devices at each site are a programmable thermostat and a few remote-control switches on appliances. The thermostats allow BPA to remotely raise or lower the temperature setting. Electron-hungry heaters and air conditioners then place less demand on the grid. Using the remote switches, BPA also can turn off the heating elements in water heaters, or turn off pool pumps.
"When the grid starts getting congested on a hot August afternoon, BPA can call a reduction in demand by controlling the energy-intensive appliances for a brief time, usually an hour or two, through a secure Web application," explains Jenny Roehm, the BPA project manager of the Ashland pilot.
During such a call, the home's temperature drifts gradually upward, and the hot water in the tank slowly cools. Small differences in temperature go unnoticed, especially if the home is unoccupied during the day. When the demand peak subsides, BPA can return the home to its normal temperature and operation. But homeowners can override the call sooner, if they aren't comfortable.
Like starting a generator, only cleaner
BPA experimented with their remote control system to see how much energy demand they could shift. In the hot summer days between July 18 and September 23, 2005, they called for reductions in demand 17 times. What they learned, and continue to learn, are indispensable lessons for grid operators like BPA.
BPA found that it could gain just under two kilowatts per home by remotely curtailing loads. Multiplied by a hundred homes, that's like starting a generator the size of a Hummer to supplement the power from the grid. In winter experiments, the gain was closer to 2.4 kW per home.
Consumers had the power to override the remote requests, but rarely did. BPA found they could shift demand for two to four hours at a time, without inconveniencing homeowners.
The BPA experiment in Ashland showed that the peak load management system is technically viable. But is it worth the investment?
Avoided wire construction means savings for communities like Ashland
Shifting peak demand means BPA can postpone the construction of new lines. That saves BPA the interest they would pay on the capital cost of the lines, for as long as they can postpone building them. The actual savings would depend on the line type, length, and location. In a similar pilot on Washington's Olympic Peninsula, the group estimated the savings at $6 million.
The buttons on the Invensys screens at BPA represent a source of energy and transmission capacity when supplies are constrained. Calling a brief reduction in demand is the effective equivalent of starting a local generating plant. Ashland, however, like many BPA-powered communities, doesn't have a local generating plant. The cost of building, fueling, and operating one would be many times the cost of connecting the Ashland's 9,000 homes to GoodWatts.
Data gathering, diagnostics, and psychology
"The test in Ashland did more than confirm the viability of demand-side management technologies," says BPA's Jenny Roehm. "BPA learned valuable lessons we'll use in planning other programs." By using the technology to monitor individual household loads, BPA is learning exactly what contributes the most to peak demand.
![]() |
Participants in Ashland's test can see their electricity usage patterns on a GoodWatts web page. This home's water heater represents 60 percent of its total energy consumption; the HVAC (green) is 16 percent, and the home's other loads (purple) are 24 percent. Other charts include day-by-day energy consumption and cost. Awareness often leads to conservation. |
In addition to being educational, technology like this can be diagnostic. GoodWatts often can spot a faulty air conditioner or water heater before the homeowner notices it.
Wanderscheid says the test is giving utilities insight into consumer behavior, too.
"If we're going to require customers to be involved in a program like this, we have to know whether they'll buy into it," Ashland's Dick Wanderscheid says. "It would do us no good at all to have a system that can deliver 50 megawatts of demand response, if everyone overrides the curtailment on a 105-degree afternoon, when we need it most. Ultimately we have to test them, and that's what we're doing here."
When our phone bills are unusually high, we can examine every call. The typical power bill doesn't offer that much detail, but what if it did? Roehm explains that homeowners like being able to see their power usage.
"Some people really get into it, watching their usage correlated with the time and temperature, right on their PC screens." As we've all learned at one time or another with cell phone roaming charges, paying attention to usage often reduces it. This corollary conservation effect supplements the benefits of shifting peak loads.
Beyond the field test
The trial program is scheduled to continue through September, 2006. Similar programs are in place in other states, including Nevada and California. Invensys has ten utilities and 1,000 customer premises on its GoodWatts system, representing about 4,000 devices.
For deployment on a larger scale, Wanderscheid says the biggest return on investment would come from new homes with heat pumps. That number is growing, as builders trend away from gas heat due to fuel prices.


Comments
Your comment: "Several programs are in place across the nation, designed to level the load on the grid by reducing these peaks" is slightly misleading. Demand Response programs shifting residential hot water heater and air conditioning load during peak times are responsible for over ten gigawatts of peak capacity nationwide, and make a huge impact on the reliabiltiy of our national grid almost on a daily basis during the summer months. The word "several" in your statement actually refers to hundreds of programs across the US & Canada.
In fact, many utilities are considering a fully-outsourced model for peak load management programs - where the utility essentially purchases peak capacity from Comverge, just as they would from a peaker plant. In some cases, the contract is structured exactly like a a PPA.
Price response is also a key element of demand response and full-scale programs are in place allowing customers to reduce their energy bill by controlling appliances themselves via a thermostat home automation portal or via the web. Customers are saving an average of 15% on their monthly electricity bill.
Posted by: Kelly Ziegler | May 3, 2006 07:18 AM