Top concerns of HRSG owner/operators haven’t changed much in the last several years. Emissions control, drain issues, attemperator problems, etc, remain etched in the minds of users. Some of the key takeaways from the 2017 meeting follow.

HRSG drain systems improperly designed and/or operated continue to torment personnel at some plants. Glen Wilson, Nooter/Eriksen, offered a timely project-based look at installation of an EPRI/Flexim-developed ultrasonic steam/water detection system for control of superheater/reheater drain valves.

The overall objective: Automatically detect and remove condensate from superheaters and reheaters without excessive release of steam, thereby preventing damage to coils and other steam-path equipment. This would reduce damage from tube-metal failures, stretching and bowing of tubes, and a host of related issues.

 “Simply put,” noted Wilson, “the industry has vast experience using ultrasonic meters to measure flow. The application to detect condensate uses the same sensing method, now calibrated to distinguish condensate from steam.” 

Next came a discussion on condensate removal. This is a severe service involving large pressure drops and flashing liquid; flow capacities vary with operations.

Continued throttling reduces valve life, and discussion turned to the master/martyr drain-valve arrangement in which the master valve is opened first and closed last. This maintains reliable shutoff tightness of the master valve by using the sacrificial martyr valve.  In this case, the master metal-seated ball valve opens and closes in two seconds with an unobstructed flow path.

The martyr valve, for throttling, can adjust to fairly low flows. As Wilson explained, “The martyr valve will open to a preset position based on current system pressure and then modulate based on presence of condensate at the ultrasonic sensor. Modulation rate may be lower prior to the purge and faster during the purge and after the CT is fired.”

The all-inclusive goal is to operate the master valve as little as possible.

Installation and calibration should be by experienced technicians the group was told. Although more expensive than the commonly used thermocouples, “ultrasonic sensing element signals can be directly correlated to the presence of liquid or vapor, with fast response to changes in state. Thermocouples respond more slowly and cannot distinguish water from steam at a saturated condition,” Wilson noted.

Many questions followed, some stressing again the importance of proper installation. Chairman Anderson added that a downward piping slope is necessary, as water must be able to flow toward the sensor. Therefore, it is best to specify this equipment before construction. Retrofits can be challenging with high and low spots in the drain pipe, but there are now retrofits in full-time service.

Catalyst technology. Brian Helner, Cormetech Inc, presented on Meteor™, a multi-pollutant catalyst technology patented by Siemens. This is a homogeneously extruded honeycomb catalyst in one layer with both oxidation and SCR functionality. It has been “optimized and fully developed into commercial production by Cormetech,” stated Helner.

A principal benefit is reduced pressure drop, along with accessibility (less space requirement). 

He then presented a case study, Ennis Power Co in Texas, a G-class combined cycle. Results after one year of operation showed “capability to operate at lower loads while maintaining CO emissions compliance,” plus faster CO compliance during startup. Field test data verified 99% CO oxidation at the 36% GT load point (previous was 55% to 60%), low ammonia slip, and NOx within the target value. Inspection showed excellent catalyst condition with clean and open cells.

Further installations are scheduled. One general point noted was possible permitting obstacles with substitute technologies.

Online monitoring. Through its acquisition of Alstom’s power and grid businesses, GE can now supply and service HRSGs. (GE also has acquired the Doosan’s HRSG business.) With GE’s digital background, the presentation by Pascal Decoussemaeker concentrated on turning data into actionable intelligence to improve plant performance. His overall theme was asset performance management (APM).

Typical HRSG failure modes and locations were discussed (Fig 1).

Decoussemaeker then compared traditional single-sensor monitoring with real-time, multi-sensor analysis systems with dynamic alert band, to more easily detect early stages of damage.

Next, he concentrated on the monitoring of some of the main life-limiting failure modes for HRSGs—such as fatigue, creep, and flooding events. The last, he explained, include desuperheater overspray and condensate flooding of the lower headers or manifolds during startup. However, it is not sufficient to just monitor. GE uses these insights to optimize the outage scope in an application called the Outage Advisor (Fig 2).

“Operation optimization,” he explained, “uses actual versus expected values for thermal performance.”  This is traditionally applied in monitoring systems to proactively manage plant performance. However, advances in IT also allow using artificial intelligence in completely new ways. In one application, a newly developed signal called “maximum likely demand in the next 30 minutes” was created to optimize the increase of duct-firing activity to free up gas-turbine capacity for rapid response, supporting the ancillary reserve market. This led to an annual fuel saving of $1 million.

“This example shows,” Decoussemaeker said, “how the exchange between plant operators and IT developments can lead to differentiators in a changing market environment.”

Various questions and discussions followed on how to retrofit life-monitoring systems, factory versus site installation methods for instrumentation, adjustments for cycling operation, and incorporation of site-specific data such as measured material properties.

Attemperators. One of the headliners on the Forum program was a panel discussion on HRSG attemperator concerns. Panel members were:

• • • Tom Freeman, GE Power.

    • Joe Schroeder, JES Consulting.

    • Ory Selzer, IMI-CCI.

    • Justin Goodwin, Emerson Fisher.

Chairman Anderson moderated the panel after introducing the topic and panelists. Ensuing discussions centered on the following topics:

• • • Tube and pipe failures, steam-pipe distortions, and weld cracking.

    • Performance problems, including high steam temperatures and overspray.

    • Attemperator hardware damage.

    • System controls.

Superheater and reheater tube bowing was voted a common problem by the panelists. Anderson began by asking how many people with this problem operated GE 7FA machines. Most indicated “yes.” Discussions then centered on reheater and superheater piping length, master/martyr valve combinations, and turndowns.

The panel was asked about attemperator design features (developments) to address fast starts. Selzer opened the discussion with a summary of inter-stage and terminal units in different thermal and operating environments. Goodwin continued, explaining that current operating ranges are not consistent with original attemperator designs. This led to discussion of two-stage setups.

One participant discussed a site’s move to Inconel spray nozzles; benefits were noted. Others indicated this might be a good form of “cheap insurance.” The option of reheater bypass for attemperation also was reviewed.

The gas-turbine perspective. Anderson set the stage with a summary of common experience: GT exhaust temperatures can go very high, and the attemperator cannot spray enough water. So, operators turn down the set point to prevent outlet-temperature overshoot. This increases the probability of overspray.

Freeman discussed the original GT design for nightly turndowns, and the differences inherent in cycling operation. “The world has changed,” he said. “Maintenance people want to protect the hardware; dispatch people want to make money and worry about parts life later.”  He then added, “The money-making piece plays heavily.”

Freeman examined three topics. He began with the market perspective, noting that when most existing powerplants were commissioned no one conceived of the world in which they function today. He described this as a global market and mission shift. 

Next, he extended this shift to the CT operational profile. Most existing plants were designed with part-load paths that were merely transients on the way to baseload, where units were expected to run for most of their lives.

Often, part-load curves were designed for what was then the new generation of low emissions combustors. But before long the North American market would see dramatic operational shifts. Turndown was becoming a significant factor. And sites were beginning to feel new pressures on HRSG operation and maintenance.

Freeman continued: “GE is rethinking the gas turbine load path [GT load versus exhaust temperature]. The way the load path was configured,” he continued, “was to hold temperature rise as a constant as long as possible. Some GE units are 100 deg F hotter on exhaust than counterpart GTs.”

A few years back, he said, his gas turbine team was largely unaware of HRSG constraints. “When information began to emerge, it became clear that there were two primary regions of interest: startup and turndown. The startup regime is generally related to independent GT/plant controls and often can be solved with a simple feed forward loop. In GE, that is called OpFlex Advanced Attemperation.

“The more intriguing issue is turndown,” he noted. “If you look closely at the operation, the concern often is not at low turndown but rather in the high-energy portion of the turndown, as the gas turbine approaches the isotherm while air flow (inlet gas velocity/IGV) remains fairly high.”

He then relayed some more history.  “A few years ago, GE was experimenting with a much broader load path range. The low end of the range could be called cold load path. Essentially, it becomes a trade-off on giving up air flow more quickly than T-fire (uprating the base firing temperature). Old timers would call it a simple-cycle load path. Interest grew because the modern DLN systems were now fairly robust and could provide greater flexibility.

The HRSG high energy corner could be bypassed. Attemperation valves that were running to 100% stroke could be held down around 40%. Thus, GE began to include the optimized load path in the later implementations of the advanced gas path (AGP)” (Fig 3).

He cautioned, however, about locking into any one path.

Freeman summarized the industry situation this way: GE has some methods to mitigate near-term operational boundaries. Yet, he sees the industry as needing to make uprates to the HRSG subsystems. Looking at industry direction, Freeman predicted increased exhaust conditions and the need to make an improved integrated system decision focused not on status quo, but on where the industry will likely go.

And his summary message to the HRSG community: “Scale up what you are doing. Don’t solve for today; solve for tomorrow. If you do an upgrade without understanding where the market is going, then you might end up paying twice.”