Coal, Emissions, Nuclear

A Revolution Revolution

Issue 8 and Volume 110.

By David C. Wagman, Managing Editor

Mention revolution in a power plant and the first thing that comes to mind may be all the spinning equipment.

But a second, more subtle revolution is gaining steam, if you will. It involves generating equipment, measurement tools, digital monitors, data management and intelligent systems. This revolution holds the promise of changing the way power plants are managed, maintained, dispatched and staffed.

Fully automated systems are already in place that can tell the difference, for example, between how a piece of equipment should operate on a cold day in January versus a hot day in August. Being able to sense the difference enables digital systems to alert human operators to a potential equipment problem well before it can escalate.

Seeds for this digital revolution were sown in the 1970s as computerized operating equipment began to be installed. Data acquisition devices soon followed along with graphical displays showing operations across a power plant. In the 1990s systemwide data historians were introduced, offering a centralized repository for data collected from multiple systems across a plant or an entire company. These were used mainly for regulatory compliance, especially as environmental rules tightened.

What resulted was a more “wired” power plant. More importantly, however, these developments laid the groundwork for the latest wave of change: namely, the use of data management tools and a growing use of data to operate and maintain the plant. For example, current emission control equipment can have thousands of data points that feed the appetites of data historians, but not much else. The objective now is to turn that raw data into usable information. That’s the revolution.

As sometimes happens, revolutionary change is reaching the power industry after it has already visited other industries. Process controls have been in place for years in other sectors.

Increasingly, however, “power plants have become digital workplaces,” Jim Bailes, associate partner of IBM’s Global Business Services business unit told me. “They are a network of computers and data.”

Computers and data excite companies like IBM, which hopes to enter the U.S. power generation business on the nuclear side. It plans to leverage a project recently begun with British Energy. There, IBM is working to build a centralized inventory system, which may help BE’s nuclear fleet reduce parts duplication and increase buying leverage.

John Kerastas of SmartSignal described for me a performance management and diagnostic center that Entergy built. There, some 30,000 MW of capacity across 31 units are monitored for early warnings of equipment problems. Kerastas says that where engineers and analysts spent days digging through data to find the source of an equipment problem, new data management systems cut diagnostic times to hours and even minutes.

Making efficient use of data is only part of the equation, says Kerastas. “The issue is extending asset life.”

At a basic level, data management systems exist to help operators determine whether or not a 15-year-old pump, for example, is running correctly. Benchmarked data help operators determine what is “normal.” Installed on equipment across a fleet of generating plants, the systems can identify equipment that is most in need of repair and replacement. If this sounds familiar, it’s because the power industry is taking a lesson from the way airlines use predictive failure analysis to maintain airplanes, Kerastas says.

One of the beauties of data is that it can be used either in centralized operations or in more distributed settings, depending on the corporate culture.

Rather than having a centralized operation, some companies prefer having a corporate information technology center, which in turn makes data available locally. This sidesteps “Big Brother” issues and respects the plant manager’s role as a primary decision maker.

One of the more digitally advanced power generators is TransAlta in Canada, which has installed wireless technology to monitor equipment in its power plans. And one power plant making extensive use of wireless technology is Comanche Peak, the two-unit nuclear plant near Dallas. In concert with the Electric Power Research Institute, Comanche Peak operators installed 55 wireless sensors on equipment such as pumps, motors and turbines. The plant’s digital network includes several hundred wireless access points. The long-term goal is to phase out manual data collection altogether and rely on continuous data collection and system monitoring.

The digital revolution may be profound, but it need not be expensive. A power generator that needs to install a data collection system from the ground up could pay $100 to $150 million, one source estimates. In many instances, however, investments in monitors and sensors already have been made. This means it’s a relatively inexpensive next step to deploy data management tools. The cost to deploy data management and process control software on top of those systems may run several million dollars.

With their relative low cost, potential for fast payback and promise of making power plant maintenance and operation more predictive and less reactive, this “revolution revolution” seems to be catching fire.