How digital power plant tech can help utilities get the most from their assets
GE says its new plant technology can yield up to $230M in annual benefits for a gas plant
Much of the talk last week at the Energy Storage North America conference centered on when storage could replace fossil fuel peaker plants, with keynoter James Avery of SDG&E declaring that a "future where there will be no gas turbines" could be possible in the coming years.
That talk generated excitement for the industry at the industry expo in San Diego, but even the most optimistic analysts know that it won't happen immediately. As Dr. Pratima Rangarajan, storage manager at GE's new venture Current, explained, peakers are "all over our ecosystem, and they're going to be here for a while."
As policymakers push the integration of more variable renewable generation, the need to balance supply and demand on the grid in the most clean and efficient manner will only grow. While storage offers one solution, another big opportunity, especially in the short term, is optimizing existing peaker resources for the grid.
That's where GE's new digital power plant comes in.
At its core, a digital power plant is a combination of software and hardware that interact in a generation plant to make it more flexible and responsive to grid demands.
GE Power and Water, the company's power subsidiary, unvieled its offering last month, and aims to make its digital plants landmarks of the industrial internet as it moves toward fulfilling the promise of its 2014 annual report, “A new kind of industrial company.”
The effort is part of GE's much-heralded corporate shift away from finance and back toward its industrial roots.
“The fundamental concept of a digital power plant is the marriage of the physical and the digital,” explained General Manager Monte Atwell.
Two things are necessary, he said. One is “scalable digital technologies” that leverage real-time operational data. The other is an insider’s understanding of “the true physical boundaries” of the equipment for which those decisions will be made.
Once those two tools are in place, Atwell said, the software simply needs to know the problem that needs to be solved to find "optimal operational modes" in the plant for the solution.
In a white paper released alongside its digital power plant announcement, GE foresees a “convergence of digital and physical innovations, together with advances in energy technologies” that will lead to “higher electricity output from existing assets, as well as enhanced performance from future infrastructure investment."
The digital power plant is a realization of that vision. At that convergence “of people, machines, big data, and analytics,” GE says, is Predix, its cloud-based software platform designed for creating industrial internet applications “that turn real-time operational data into insight for better and faster decision-making.”
Predix is the operating system “across GE, whether it is a locomotive, a wind turbine, or a heavy duty gas turbine,” Atwell explained. “When we talk about the internet of things and the industrial internet, it is based on the Predix platform.”
Apps for Predix, like the unending apps offered to cell phone users, can attack a range of power plant performance issues limited only by the imaginations of their designers.
“It is an open architecture,” Atwell said. That allows GE customers to guide the development of Predix apps because they best know "what they want and the problem they are trying to solve.”
When applied to power generation, GE says its platform can result in big savings and new business opportunities for utilities.
“Digitizing central generation could yield value of up to $100 million for new wind farms (up to 20% higher MW), up to $230 million for a new combined cycle gas power plant, and up to $50 million for an existing combined cycle gas powered plant,” the white paper predicts. “Across the whole power industry, this equates to up to $75 billion of impact for new combined cycle gas turbine and wind turbine orders, with additional value for upgrades to existing assets.”
The first digital power plants
There are no full-scale implementations of the digital power plant concept now up and running. The estimates of value, Atwell said, are based on elements of the technology now in service and revenue models GE has derived from its customers.
With an understanding of power markets and the theoretical value of the enhanced services digital power plants might offer, GE was able to estimate how much more it could deliver with the same hardware.
The wind industry leads the development, with over 4,000 turbines already running an early iteration of GE’s digital wind farm. Leading developer Invenergy is a major proponent.
Turbine-embedded sensors accumulate project-wide performance data and a Predix app analyzes it in real time. The system “learns over time, becoming more predictive,” according to GE. The result is increased production and decreased maintenance issues and maintenance costs.
These capabilities have improved turbine efficiency up to 5%, which translates to up to a 20% per turbine profitability, GE says.
The first full-scale digital power plant pilots will be built into natural gas turbines purchased by New Jersey’s Public Service Electric and Gas (PSEG), and into nuclear plants, wind projects and natural gas plants owned by Exelon.
GE Hitachi Nuclear Energy (GEH) will partner with Exelon Generation to retrofit a portion of the biggest U.S. nuclear fleet with two operational applications.
One, Watchtower, will monitor the real-time operational status of plant equipment and receive proactive notifications of possible machine issues. It will then use the data “to predict the performance of plant equipment, enabling operators to optimize equipment maintenance and improve reliability,” according to GEH Spokesperson Jon Allen.
The other application, Lighthouse, will use advanced analytics to examine historical organizational performance indicators on cost and operations to “drive decisions that enhance organizational effectiveness,” according to Allen.
PSEG will use Predix apps to upgrade the reliability and performance of some of its in-service natural gas turbine fleet. It will also add GE’s 7HA.02 natural gas turbine, one of its world-leading H-class machines, in a new combined-cycle plant.
Along with its commitment to Predix, GE has invested heavily in validating the performance of the H-class turbines. GE engineers have used access to a new “full speed, full load test stand,” Atwell said, that allows them to “push the limits of its operating space.”
The resulting data have allowed the design of a turbine and associated hardware flexible enough for Predix applications to run it to serve base load or peaker operations, Atwell said.
“Once I understand the physical limits of what a machine can do and I understand how far I can push it and in what directions, I can marry that with the digital technologies,” he explained. “Then I can write lots of software to optimize an operating scenario anywhere in that operating space I want.”
New performance options for utilities
The immediate objective is to begin providing GE customers with increased capabilities because, with limited penetration of digital technology, the white paper explains, “challenges of sustained profitability, data deluge and an aging workforce still remain significant.”
“The convergence of digital and physical technologies that is unfolding across industry can turn these challenges into unprecedented opportunities,” it says.
With increased capabilities, GE customers can consider trade-offs they do not now have.
“Managing a business is all about risk and reward,” Atwell explained. With digitally enhanced capabilities, the customer can use a plant built for base load application to seize a revenue opportunity in the capacity or grid services markets.
"The customer may decide to fast-start that plant even if it consumes life from its parts,” Atwell said. GE’s understanding of the plant’s physical limits and its digital control of operations can allow that customer to do an aggressive ramp and to understand the cost in parts life of doing it.
“That risk-reward trade-off scenario in real-time is what we are after,” Atwell said. The customer may lose $1 million in parts life but make $10 million in incremental revenue, he added. GE wants to give its customers the choice.
But that is not the end of what GE’s digital advances can offer.
“The next application could show how to pull some of that life back by doing some incremental maintenance or maybe, for the next period of time, backing the machine down some,” Atwell said.
The digital power plant concept, Atwell added, “allows an almost infinite space of applications we can develop with customers.”
The future of the digital power plant
“Power producers and utilities are embarking on a journey to digitize their processes,” GE's white paper foresees. “This will require not just investment in new technologies, but also a shift in mindset and business models — and the shift will need to be faster than ever before.”
In April, GE announced it would shed the finance business that made it billions each year before the financial crisis, recommitting itself to its industrial roots and an effort to take that business into the digital age. Analysts are not yet convinced.
After buying German competitor Alstom, GE’s main competition in the power sector is from Siemens, with Honeywell and United Technologies trailing. The Siemens core operating margin is about 9.4%, pointed out Hedging Beta analyst Alessandro Pasetti.
The GE Power and Water operating margin for Q2 2015 was “a respectable operating margin of 17%,” he wrote. But with other divisions of the company included, the average operating margin was 18.4%.
Standard and Poors has downgraded the stock in Q2 because "the company could potentially adopt a less conservative financial policy…[or shrink] its industrial focus along with its finance unit.”
“The debate between the bulls and the bears is open and there are reasons according to which GE stock could either rise or fall,” Pasetti concluded after the company’s Q3 earnings report.
But whatever analysts and stock prices say, GE shows no sign of turning back. Its 2 MW next generation Brilliant wind turbines will capitalize on enhanced digital capabilities to operate across the hardware’s operating space. It will have a “digital twin modeling system to build up to 20 different turbine configurations at every unique pad location across a wind farm,” GE says.
Its natural gas plants may not achieve that much flexibility right away. “All the equipment in a traditional power plant was designed for a specific mission,” Atwell said. But, ultimately, “the scalable digital technologies can be used to basically write applications that allow the plants to run any way I want.”
There will be “co-creation” opportunities with the pilot project partners, Atwell expects. “We intend to do a lot of joint learning along the way. We can do anything we want with software, relatively inexpensively. But you have to understand what you are trying to do.”
The new capabilities in wind turbines is important because the heightened visibility into a project’s production will allow grid operators to integrate it more readily.
The new if more limited degree of flexibility in natural gas turbines is important for a different reason. “A customer buys a plant for its mission profile today,” Atwell said, “but two years from now, because of things happening in the market and on the grid and with different kinds of loads and increased renewables, that base load plant might need to be a peaking load plant.”
The vision is for a customer to be able to download the apps required to turn a base load plant into a peaking plant without the necessity of a hardware upgrade, Atwell said. “But just like any other technology developer, we continue to improve our products and there will always be hardware upgrades available along with the digital upgrades.”
Eventually, the plan is for a system that is continually receiving updates to the Predix operating system with “the latest and greatest stuff,” he added. It would also have “an enormous suite of applications from ‘the app store.’”