Control system helps to gather maintenance data

TMPA serves the 450,000 residents and business owners in the four Texas cities of Bryan, Denton, Garland and Greenville with 3.5 million megawatts of energy each year.

The electricity from the generator increases to 22,000 V before it reaches the Gibbons Creek Switchyard, where a transformer steps it up to 345,000 V and feeds it into the Texas power grid via transmission lines. These transmission lines are connected to smaller distribution centers that deliver power to homes and businesses throughout the four member cities.

To produce the 470 MW/h necessary to provide power to the four municipalities, TMPA needs its system to be online and running as much as possible. “If our station goes off-line for any reason, we have nothing to supply, and the distribution centers we provide power to are forced to buy electricity on the open market,” says Bruce Partlow, chief operations engineer, TMPA. “That can cost the cities millions of dollars a day, depending on how much electricity costs at that time. It’s critical to our customers that our unit remain on-line and in good working order.”

Unfortunately, the 20-year-old steam turbine’s supervisory protection system — the part of the machine tasked with monitoring the health of one of the plant’s most vital assets — was outdated. Spare parts were increasingly difficult to track down, and the analog system could no longer provide the level of detail Partlow and his team required. Sometimes, the system would catch a small issue that could be addressed at any time, but due to the lack of visibility into the cause of the trip and a lack of trained personnel on-staff able to diagnose the problem the plant wouldn’t be able to continue operating.

“When an alarm went off on the old system, we would have to call in an expert, often from the OEM that sold us the machine, to come out, hook up a handheld analyzer, take readings and then diagnose the problem,” explains Partlow. “If we were lucky, someone could come out in a day. But sometimes it took several days, which meant we ran the risk of unplanned and unnecessary downtime.”

When the risk of nuisance trips and unplanned downtime proved too costly for TMPA and its customers, the company decided to upgrade the turbine supervisory instrumentation system. It considered another solution from the original system’s manufacturer. However, that solution came in a single box and didn’t provide the level of information or flexibility TMPA required.

The TMPA team decided instead to install customized turbine supervisory instrumentation based on the Rockwell Automation’s Allen-Bradley XM-series condition monitoring modules. The team first installed the distributed system on the steam turbine and the two turbine-driven boiler feed pumps and then expanded to other key boiler components, such as the primary air, forced and induced draft fans.

The modules feed vibration, temperature and positioning information gathered from the turbine, feed pumps and fans to Rockwell Software RSView32 human-machine interface software running on Allen-Bradley VersaView computers. “The software offers our operators easy access to basic machine health information and alarms,” explains Partlow. “If the modules identify a problem with the turbine, it’s much easier to determine whether the issue is severe enough to warrant a shutdown, or whether it can wait until our next scheduled maintenance period.”

The modules also feed asset health information to the plant’s distributed control system from an Allen-Bradley ControlLogix programmable automation controller. Communicating via the DeviceNet network, the system provides continuous updates on the turbine’s health to the central system, which helps keep the plant running at maximum uptime.

The team also used Rockwell Software RSNetWorx software to design and configure TMPA’s EtherNet/IP and DeviceNet networks, and installed Allen-Bradley Emonitor software to extract live asset health data and provide system vibration analysis capabilities. “We can use the information to move from reactive, preventive maintenance to proactive, planned maintenance,” explains Partlow.

Operators need access to machine health information to make better decisions.

Partlow and his team now have access to critical, real-time information about the health of their assets. “We’re operating smarter,” says Partlow. “We have access not only to more data, but better data, which allows us to make more informed decisions about our equipment. We’re able to determine when a problem is truly critical, and when we can work around the problem until there’s an opportune time to address it.”

The system also allows Partlow and his team to be more self-reliant. “We no longer have to pay for the turbine OEM to come out and fix our equipment, which saves us time and money” he said. “Our system is far more reliable now, and we have the expertise to identify and prioritize maintenance concerns in-house.”

For example, the team recently completed some maintenance work on the turbine, but when they brought it back on-line, the new supervisory system caught some inconsistencies in the vibration levels. Partlow was able to diagnose an alignment problem, rather than a more serious problem, such as a bearing failure or rub. “We were able to avoid another system shutdown, which meant more uptime and increased access to wholesale power for our customers,” says Partlow.

Indeed, the cities of Bryan, Denton, Garland and Greenville are perhaps the biggest beneficiaries of the new system. When TMPA is on-line and providing power for the municipalities, each city is avoiding the potentially millions of dollars it can cost to buy electricity on the open market. “We saw a return on our investment in less than two years, and our forced outage rate is far below the industry average of 7%,” says Partlow. “We’re able to keep the system on-line 88% of the time, generating at 85% of the capacity of the unit.”