Business, Controls and Automation, North America

Hydro Review: The Quiet Revolution That’s Upending the Utility Industry

If there’s such a thing as a quiet digital revolution, its field general is New York Power Authority President and Chief Executive Officer Gil Quiniones.

As head of largest state-owned utility in the U.S., Quiniones was the visionary for transforming NYPA into the first all-digital utility in North America. It was an audacious plan, even for a utility accustomed to spending hundreds of millions of dollars annually to update its infrastructure. The technology exists to make it happen. So, Quiniones asked a simple question: “Why not us?”  Why not, indeed.

Consumers may not notice the difference when they turn their lights on, but just about everything else a utility does to make that possible is about to undergo radical change. We take being connected for granted. Our offices, cars, homes and appliances are all linked to devices we can readily access so that we can analyze and act on consolidated information in one form or another. It should be no different for a utility.

NYPA’s approach

At NYPA, enterprise-wide operational data is fed into the Integrated Smart Operations Center (ISOC) at the utility’s White Plains, N.Y., headquarters, about 20 miles north of New York City   The center, which opened in December 2017, uses predictive analytics software called Predix from GE Digital that collects information from 26,000 data points (with more to come) throughout NYPA’s 16 generating facilities — nine of them hydro. NYPA’s substations and transmission assets will also be connected via this software by 2024.

The data provides real-time information about signs of aging equipment or mechanical anomalies that could signal potentially crippling equipment issues, such as high vibrations or insufficient stator cooling. Thanks to the rapidly evolving technology of predictive analytics, which uses modeling algorithms coupled with trending and regression analysis, potential issues can be identified before they turn into full-blown problems.

The system effectively raises its hand to let an ISOC engineer know if something is trending in the wrong direction. NYPA’s Variable Similarity Based Modeling software — developed by Smart Signal, now a wholly owned subsidiary of GE — has enabled personnel at the ISOC, in just its first year of operations, to detect numerous conditions that have led to at least $3 million in avoided costs due to such conditions as erratic thermocouples, frozen transmitters, and operational deviation resulting from normal wear and tear.

The establishment of NYPA’s ISOC was complemented by an engineering reorganization that now enables a seamless and comprehensive collaboration between plant maintenance and engineering staff and engineers working at NYPA headquarters. The notion of connecting NYPA’s operations technology and information technology platforms may have once been unusual to the point of being downright radical. But this technological convergence has altered the paradigms of how NYPA performs asset health management. Enterprise data is now converted into actionable information. Among other things, that means providing personnel at its plants with detection of emerging anomalies to avert forced outages.

For example, at the 2,755-MW Niagara Power Project, the largest power plant in New York, there are a total of 25 turbines. Collectively, the 2,525-MW Robert Moses Power Plant and the 240-MW Lewiston Pump Generating Plant generate 2.6 million kW of power using 748,000 gallons of water a second. The project is critical to ensuring an adequate power supply for western New York and beyond. Actionable information from the ISOC will help ensure its reliability.

In January 2018, the ISOC software generated an advisory, leading to an investigational case that indicated the thrust bearing cooling water flow at one Lewiston unit was tracking at 37 to 41 gpm, or about 19 to 22 gpm higher than required to provide adequate cooling given the operating conditions. Trending analysis confirmed the advisory was correct. This analysis that resulted in this “case” prompted plant engineers to check the position of the flow valves. They were found to be in the normal position for the required flow. However, further investigation revealed that the calibration on the cooling water flow transmitter had drifted.

Why is this important? Adequate cooling water flow must be maintained to prevent damage to the thrust bearing. An erroneous reading that indicates a water flow error 50% higher than normal would also have provided a deceptively normal reading should the flow fallen below normal. Had the calibration issue not been revealed, it could have caused damage to the machinery, leading to expensive repairs and possibly a prolonged outage.

Ask anyone who’s worked on the floor of a hydro plant and they will tell you how labor-intensive it is to track unit outages and failures for the purpose of identifying their individual root causes. It could take six to eight months of mining data nonstop, if it was done at all. That same data can now be processed in six or seven minutes by a computer. Engineers can only look at so much data at once, while a computer can look at all of the equipment on a continual basis and draw attention to a piece of equipment that is not functioning normally.

How the system works

Within the GE Predix software, NYPA has created what are essentially digital twins of its assets using historical and real-time data, relying considerably on NYPA’s enterprise-wide time-stamped PI data historian from OSIsoft and supervisory control and data acquisition (SCADA) historian from Oracle. Together, these systems form the basis for the predictive analytics and an important byproduct — scheduled maintenance. The software uses the plant operational data to model an asset’s performance when it is running normally and compares this “training data” to how the asset is running at any given moment. Essentially, the computer is trained to recognize what “good” looks like.

Financial benefits

Having a predictable schedule of when to take units out of service for routine maintenance can cut costs considerably. Using these analytic tools enterprise-wide shows up in the bottom line in other ways, especially at hydro plants, where increased efficiency sometimes comes with an unwanted byproduct: vibration. Vibration can often be traced to problems with bearings and wicket gates and failures of shear pins. Hydro plant operators will tell you that regardless of the make and model, vibration can be a persistent problem. In addition, if generation equipment is shaking, it can also signal the bane of hydro operators — cavitation and the erosion of metal in draft tubes, which can lead to turbine failure if left unchecked.

However, if caught in time, the end result is likely to be a shorter and less costly outage. The Predix software in use at the ISOC can trend anomalies tied to vibration and the units can be programmed to run accordingly. Tackling vibration is more than just about avoiding the cost of repairs. At the end of the day, vibration is little more than wasted energy that results in unwanted heat and wear. Reducing it enables NYPA’s units to run more efficiently. On an individual unit basis, those savings might feel slight, but they quickly add up when spread over all of the units at the plant. Even if that increased efficiency allows for only a penny more per kilowatt-hour in revenue, that still adds up to about $3 million annually.

Lower operating costs also mean NYPA can bid its power into the market at a lower rate to be chosen by the New York Independent System Operator and still realize a healthy margin. Conversely, if there’s an unplanned outage, customers will still receive power at their contractual price. But that power may need to be acquired on the market at a higher cost than if NYPA had generated the electricity.

Digitizing will help keep costs low or even lower costs further for NYPA customers, who already pay among the lowest electric rates in New York. Power from Niagara and NYPA’s 912-MW St. Lawrence-FDR Power Project in Massena helps support ReCharge NY, a program created in 2011 by Governor Andrew M. Cuomo that provides low-cost power to New York business enterprises in return for a commitment to retain or create new jobs and make new capital investments. More than 369,000 jobs and $30.3 billion in business investments are linked to ReCharge NY.

Promoting economic development is at the core of NYPA’s mission. So is being at the forefront of shaping New York’s digital energy future. The ISOC will help fulfill the goals of Governor Cuomo’s Reforming the Energy Vision strategy to build an energy system that is cleaner, more resilient and affordable for New York’s 19 million residents. It will also help meet the governor’s Clean Energy Standard, which calls for half of all electricity used in New York to come from renewable sources by 2030. About 71% of the power NYPA produces come from hydropower.

The human factor

Digitizing NYPA’s operations is an exciting prospect. At the same time, the human adjustment factor must still be weighed. NYPA has hundreds of dedicated and highly skilled people working at its plants, headquarters and transmission substations. For them, the benefits of going digital are clear when it comes to plant maintenance. But that does not mean the adjustment is always going to be easy. At Niagara’s Lewiston Pump Generating Plant, the governor system had an old-school fly-ball assembly. It was a sturdy work horse, but when it acted up, it could take weeks of sorting through the hard-wired system looking for a problem deep within the old electromechanical relay system to find a root cause that might never be revealed. Now there are simply two positioning servo-motors fed by a programmable logic controller that provides continuous data for troubleshooting and real-time health analysis.

For veteran NYPA mechanics and engineers who have been working with the “hard wires” for decades, this new approach will take some adjustment. Conversely, newer hires are coming to NYPA with a digital mindset. Asking them to master the electromechanical equipment installed 60 years ago is a tall order. But they are ready for an all-digital future. Where once plant operators may have started in other departments, like security, and worked their way up, now they come to NYPA armed with electrical engineering degrees and MBAs.

As game-changing as this digital asset-management program will become, the human touch is still needed. Despite the apparent benefits of interactive software that transforms machine sensor data into actionable intelligence, those analytics must still work hand-in-hand with expert ISOC operators and on-site expertise about a plant’s overall operations.

Whether they have been at NYPA for three months or 33 years, staff at the plants and ISOC are now digitally empowered to rapidly turn collaboration into action. Operating system data is now available to every operations manager, plant engineer and headquarters engineer who needs that information to do their job. Why have monthly reports when you can instead have real-time dashboards and constantly updated key performance indicators?

It’s not data for the sake of having data. It’s having the right kind of data that can be acted on to provide insightful information that will make NYPA  hydro plants more efficient, less expensive to maintain, and more competitive in an increasingly challenging energy marketplace.

In 2018, it’s how the digital revolution is won.

Richard Gaines is director of the Integrated Smart Operations Center at New York Power Authority headquarters in White Plains. Christopher Carey is operations superintendent at the Niagara Power Project.