You could call it utility-scale renewables 3.0. 
The previous two phases focused on getting large solar arrays and wind farms up and running (1.0) and then boosting their output (2.0). Today, developers are looking beyond just adding more rows of panels or bigger turbines to their plans. They are looking to new technologies that consider solar and wind opportunities in new ways and add performance-­stretching intelligence to existing installations. In the process, they are bringing these admittedly intermittent resources closer to performing like base-load generating assets.


Humble beginnings to big business


The first large-scale solar plant, located on California’s Carizzo Plain, went into operation in 1983 as the largest solar array in the world with a peak output of 5.2 megawatts (MW). It was decommissioned just 11 years later in 1994 when oil prices, which had been on the rise, stagnated and fell, making the plant’s output too expensive to maintain. The first large U.S. wind farm, at Crotched Mountain in New Hampshire, powered up in 1980 with 20 30-kilowatt (kW) turbines. Due to equipment malfunctions, that installation failed after just a year or so.


Things have changed more than a little since those early days. In 2012, wind and solar became the two fastest growing electric-generation technologies in the United States, and production capabilities have climbed just as dramatically. The Topaz Solar Farm, in the same general Carizzo Plain region as the failed first installation, went online in late 2013 with 5 million panels and a rated output of 550 MW. 


Wind has grown even faster. As of 2013, the Alta Wind Energy Center in Kern County, Calif., was the largest wind farm in the world with an installed capacity of 1,320 MW. Planned additions are expected to boost that figure to 3,000 MW by 2019.


In fact, we may be close to reaching the outer limits of solar- and wind-farm production with current approaches. The photovoltaic panels of the Topaz solar farm are spread over 9.5 square miles—yes, miles—and the Alta wind project’s turbines, many of which have nameplate production capacity of 3 MW, occupy approximately 9,000 acres. Getting rights or ownership of that much property is an expensive and contentious process, so researchers and developers are looking at new ways to get more output without simply adding more panels or bigger turbines.


Solar’s game of concentration


For solar developers, boosting performance means looking beyond traditional PV panels to more advanced means of extracting energy from the sun’s rays. Concentrated solar power (CSP) approaches, for example, are making inroads in the United States with the launch of the Ivanpah plant in California’s Mojave Desert and, perhaps more important, the Solana plant near Gila Bend, Ariz., about 70 miles southwest of Phoenix. Both plants produce electricity the old-fashioned way, through a steam-driven turbine. What’s unique is that they combine traditional thermal generation with heat from the sun instead of from combusted fossil fuel. Solana takes this technology a step further by storing solar heat in molten salt and extending production for up to six hours after the sun goes down.


Utility-scale concentrated solar plants were first constructed in Spain. More than 50 such facilities were brought online during that country’s boom years in the late 1990s and early 2000s. However, the international financial downturn hit Spain hard, and subsidies that supported construction have disappeared. Today, the United States is seen as a bright spot for concentrated solar innovation.


“The plants that are coming online are U.S. plants,” said Mark Mehos, CSP program manager for the National Renewable Energy Laboratory (NREL) in Boulder, Colo. “But in the U.S., it’s been more of a one-off. I think that building these plants at scale will begin to help the utilities become more comfortable with the technology.”


Instead of photovoltaic panels, which produce electricity directly through sunlight-stimulated electrochemical reactions, CSP operations use mirrors or lenses to concentrate and direct solar energy to a collection point, where it heats steam for a traditional turbine. Designs for these plants vary. Some use mirrors to focus sunlight onto boilers located on central lighthouse-style “power towers.” The 392-MW Ivanpah station, the world’s largest solar-thermal power plant, uses this approach. Other CSP designs, such as the Solana plant, incorporate rows of curved mirrors. The mirrors are connected by tubes filled with a heat-transfer fluid that passes through a heat exchanger to create steam.


Of course, both of these options on their own face the same challenge as traditional PV panels. Plant developers only pay off their investment when the sun is shining. This is why Solana’s developers added some salt to spice up their plant’s production. Specifically, a molten-salt thermal storage system means Solana is still contributing to system supplies during the peak-demand period just before sunset when utility customers come home from work and turn on air conditioners.


Mehos sees storage as a major differentiator for Solana and other plants like it. Such plants are expensive to construct, but thermal storage allows system operators to treat them more like base-load natural gas, coal and nuclear generators. Such plants’ output is steady and dependable and doesn’t suffer from the momentary voltage drops that challenge PV panel operations (and the grids to which they are connected).


“It becomes more significant when you start [estimating] the value of PV,” Mehos said. 


At higher penetration levels, the value of each added PV panel drops for the whole system because of the technology’s intermittent output and the fact that its production drops off just as demand is hitting its peak.


Building better blade runners


The 30-kW output of the original Crotched Mountain wind turbines would barely register in today’s commercial market. Today, manufacturers, such as Alstom, GE, Siemens and Vestas, market equipment capable of producing up to 7.5 MW each (though most U.S wind farm turbines max out closer to 3 MW). Reaching this performance level, however, requires manufacturers to think big—really big. The blades of a Siemens 3-MW turbine are 180 feet long. Add that to a tower that could reach 400 feet tall, and the result is a structure that stretches 580 feet from blade tip to the ground.


In addition to climbing ever higher, turbine experts are investigating what they can do to push more energy out of this equipment by helping it operate more intelligently. GE leads this field, having recently announced two new offerings intended to boost productivity and profitability at existing wind installations by looking at individual turbines and entire farms.


The company’s PowerUp Analytics package enables technicians to assess and fine-tune specific categories of turbine operation, such as blade pitch, torque, speed and the orientation of the nacelle—the housing for turbine mechanics at the point where the blades come together—for an incremental increase in performance that can have an outsize effect on a wind farm’s bottom line.


“Turbines are designed for certain wind conditions. However, where they are actually sited is different than what the designed conditions might be,” said Keith Longtin, general manager wind products in GE’s renewable-energy division. “We are always thinking about how we can get even more power out of the asset.”


The company estimates the PowerUp program can boost a wind farm’s overall performance by up to 5 percent, which can lead to a 20 percent boost to the owner’s bottom line.


“The figure is the result of the turbine being more profitable without increasing costs or depreciation,” Longtin said. “[This] provides our customers with enhanced economics and higher efficiency from their farms.”


Utility crystal ball


Despite improvements that can help solar and wind facilities act more like base-load generation, their intermittent nature can still wreak havoc at all levels of the electricity grid, from regional transmission systems down to neighborhood-level distribution transformers. Big data is beginning to help utilities and transmission-­system operators predict such system anomalies before they happen and present that information in a way that has meaning to the person sitting in front of a computer terminal.


“We take data from different sources; we analyze and visualize it,” said Steve Ehrlich, senior vice president of San Mateo, Calif.-based Space-Time Insight. The company, which launched in 2008, has seven of the top 20 U.S. utilities as customers, and it is helping improve insight into grid operations, from the level of the independent system operator (California’s independent system operator is a client) down to individual customer meters.


For large-scale solar and wind integration, Space-Time Insight can tie near-real-time weather forecasting to specific wind and solar farms to provide updated forecasts of how much actual electricity those facilities can be expected to produce.


“It’s what we call translating big data into little data,” Ehrlich said.


The Sacramento Municipal Utility District (SMUD) is another customer, and its planners are taking Space-Time Insight’s translation skills down to the customer level to help understand the effect of rooftop PV panels, and they have learned how electric vehicles (EVs) can affect the local distribution grid.


“It turned out that, the first time anyone plugged in their Tesla, they used many more cycles than expected,” Ehrlich said. 


So, SMUD began tracking EV households, even accessing Department of Motor Vehicle information on owner addresses and vehicle type (because every EV has different charging characteristics) to better understand potential capacity problems.


“They need to balance the grid. If they don’t have enough capacity in a neighborhood, they need to add to that,” he said.


With a vast range of specific data on assets—such as transmission and distribution substation locations, high-­voltage line availability and other information—the software also can help planners understand where new capacity, including solar and wind installations, makes the most sense and where it doesn’t. Users can simply mouse around a data-populated map, and sweet-spot locations will show up as green on their screen.


All of this is leading to the capability to, essentially, predict the future. With enough information about equipment condition, system demands, upcoming weather events and other critical factors, grid operators working in regions with a broad array of intermittent and base-load resources might be able to head off outages or other problems before a single lamp flickers off, Ehrlich said.


“The big thing right now is predictive analytics,” he said. “If I can analyze it and tell someone what’s about to happen, that’s the star that everyone’s reaching for.”