Bernard Norris is the steam turbine services product line leader GE’s legacy fleet. Based in Atlanta, where GE houses its remote Monitoring & Diagnostics Center, he is part of the team involved in the company’s first-ever ESP installations for GE D-11 steam turbines at combined-cycle power plants in Johnson, RI, and Elk Hills, CA. He spoke with Plant Services about what an ESP upgrade really means, the value of remote monitoring as a decision aid and a cost-savings tool, and what the future of monitoring looks like.
PS: What’s the significance of ESP for steam turbines? What value do customers get from having someone in Atlanta keeping an eye on their equipment?
BN: As these D-11 units are monitored, the team is looking at different parameters. They’re looking at performance, output, heat rate, they’re able to calculate or identify vibration issues or different load swings or decrease in the output. The team is proactively looking at the data and then seeing if there are any issues with any unit in particular.
With the project in California, we saw some potential rotor-bow issues, and then that rotor-bow issue could impact operability or availability of the unit. We get involved and we say, “This is how much your unit is vibrating or there’s a rotor bow that could lead to vibration issues.” And then we proposed to do a repair. In some cases customers will take the repair option. But in other cases, the customer will make a different decision and say, “I don’t want you to repair my rotor, I want you to replace it.” The benefit of replacing it is to reset the clock of the high-pressure and the intermediate-pressure section of the steam turbine, extending the life of the unit as well as mitigating future risk of unplanned or forced outage.
It’s critical that we have units that are monitored, but more importantly we have the ability, the analytics, and the algorithms to assess a unit to tell a customer, “Your unit is experiencing rotor bow; here’s a way that we can mitigate this risk, either through a repair (which essentially straightens the rotor, removing the bow) or replace.”
PS: Do the algorithms reliably indicate whether a repair or replace is the best option?
BN: The M&D algorithm will tell identify the severity of the rotor bower, you whether (a bow is) 10 mils (0.01 inches), or it’s 2 mils or 3 mils; we have the ability to measure to that level of accuracy.
I won't say that the repair solution is a quick fix. There is a repair solution that will address the rotor, but the cycle is outside of the normal outage. Essentially it straightens the rotor and removes the bow so therefore when the rotor is installed in the unit, it helps drives the right level of eccentricities and clearances to mitigate any rubs or unplanned events.
We have a value economic model that we’ll run where we’ll say:to point out “Here’s (costs and payback associated with). Here’s the option for repair; here’s what the price is for two years with payback, but it also will show the cost to replace the entire rotor. And when you replace the entire rotor, you also get an additional 1.5% of output, as well as you get up to a 1% increase on efficiency of the unit.”
PS: Can you go into a little more detail about that process?
BN: I'll walk you through the genesis, which would be around the M&D data. We have around 200 units in our D11 fleet. About 70% of those are monitored through our facility here in Atlanta. An engineer is pretty much on call 24/7. There's always someone in the center looking at data. The teams are running the algorithms, and they look and see, OK, there's a particular unit that's having some issues around operability or let's say vibration issues and we assess, this is due to eccentricity, and most likely this unit has a rotor bow. We then dispatch our regional application engineering team. These are engineers that are assigned to specific customers, so in the case of Elk Hills in California, there's an engineer out in Utah who's talking to them quite a bit, visits them, etc., along with the salesperson. So you kind of think of these folks as technical salespeople.
The regional application engineer will take a look at the data, based on the M&D, assess the unit configuration and unit history to understand what's the history of the unit, if the unit has experienced any operability issues in the past beyond what we've seen within that data analysis. They'll then tie that back to our sales and commercial process. They'll engage the sales manager and present a high-level summary of findings to the sales manager stating, "Here's what we believe the issue to be; here's the remedy." The sales manager looks at it and says, "OK, I know what the value drivers are for my customer, so in this case that's availability, reliability, and more around life extension.
So this customer says, "I plan on running this asset for another 20 or 30 years." The salesperson will take all that into account, have some discussions with the application engineering team, and then that's when we make the decision to say, is this a viable solution for our customer? And then they have a discovery session with the customer. It's not, "Hey, we want to sell you the latest whiz-bang technology, and you have to buy it." We want to put all the options on the table. Here's the cost, here's the cycle, here's your payback.
PS: Is this the future of plant maintenance, where you’re going to see certain expertise moving outside the plant walls, and you get this collaborative, remote-based team effort?
BN: Absolutely, that’s exactly where the industry is headed. GE is really big committed to developing the plant of the future, the “digital power plant.” Everything is going to be digitized, with the ability to monitor, the ability to analyze.