Risk and Cost Reduction in Real Time

Jason Van Velsor, Matt Freeman, Ben Ruchte

Installed sensors and continuous online monitoring are revolutionizing how power plants manage assets and risk by facilitating the transformation to condition-based maintenance routines. With access to near real-time data, condition assessments, and operating trends, operators have the opportunity to safely and intelligently reduce operations and maintenance costs and outage durations, maximize component lifecycles and uptime, and improve overall operating efficiency.

But not all data is created equal and determining what to monitor, where to monitor, selecting appropriate sensors, and determining data frequency are all critical decisions that impact data value. Furthermore, sensor procurement, installation services, data historian/storage, and data analysis are often provided by separate entities, which can lead to implementation challenges and disruptions to efficient data flow.

To provide our clients with a simplified implementation option that expedites the transition of data into intelligence, SI has developed SIIQ, a turnkey monitoring solution consisting of:

• Assessing the most appropriate (e.g. highest risk) locations and method for monitoring
• Multipurpose wireless sensor network
• An independent data transmission infrastructure
• PlantTrack™ visual database integration
• Customizable automated alerts
• Automated engineering insight

While there are many applications in which effective monitoring can be used to more efficiently manage the operation and maintenance of passive assets, such as high energy piping, attemperator management is one specific application that clearly demonstrates the value that can come from an effective monitoring program.

Industry Issue
Attemperators (or desuperheaters), which reduce steam temperature using a water spray, are one of the most problematic components in combined cycle plants. There are several attemperator designs and configurations, but all are potentially vulnerable to damage. If the causes of damage are not addressed early, cracking and steam leaks can occur, leading to costly repairs and replacements. As is typically the case, currently installed data transmitters (pressure taps and thermowells) are located far downstream/upstream and cannot detect local transients that would suggest events like spraywater impingement, pooling, etc. The main challenge is that these events can lead to damage that often goes undetected until it is too late because the damaging temperature transients are not detected by standard plant control instrumentation. Without this local temperature data, it can be hard to predict when re-inspections/other mitigation steps should be pursued.

Monitoring Equipment/Capabilities
To better characterize local temperature events and provide early indication of non-optimal attemperator operating conditions, SI offers a combination of software and hardware components that can be implemented with a range of services from monitoring, detecting, through diagnosing. At the root of these services is the need to collect data from locally installed thermocouples. While some plants chose to run the signal through the data historian and then transmit to SI for processing, an alternative is to use our wireless sensor network to collect and transmit data. SI’s wireless sensor network consists of two primary components: (1) a sensor node that collects the sensor data locally and transmits it wirelessly to (2) a gateway that transfers the data to the cloud.

Figure 1 shows an image of SI’s data collection node, highlighting several of its features. Each node has multiple sensor channels and is capable of collecting data from a variety of sensor types. For the case of temperature monitoring, up to nine different standard thermocouples can be connected to a single node. Additionally, each node is battery powered and is available with an optional solar charging kit for outdoor installations. Furthermore, the data acquisitions nodes are weather-proof and designed to be installed in exposed locations.

As shown in Figure 2, the data acquisition node is installed locally and all thermocouples are hardwired to the node. The node then transmits the data wirelessly to the installed gateway using a proprietary 900MHz wireless protocol. The data collection and transmission frequency is adjustable based on the requirements of the application.

The data from all installed nodes are transmitted to a locally installed wifi/cellular-enabled gateway, which stores the data on a local database until the data is successfully transmitted to a cloud database. Serving as the edge connection to the web, the gateway can be configured to use a cellular network, eliminating the need to connect to any plant networks, or it can be configured with a plant-wide wifi network, if available and accessible. The location of the gateway enclosure is flexible as long as it is within ~1000 ft of all installed data collection nodes.

Figure 1 – Wireless data collection node.

SIIQ/PlantTrack App
Once transmitted off-site, data can be accessed through SI’s PlantTrack software (part of the SIIQ. PlantTrack provides a suite of real-time event and damage tracking applications for common plant components: piping, headers, tubing, attemperators, etc. These applications interface to common DCS / Historian systems allowing for easy implementation, including review and analysis of historical data where that exists.

For Attemperator damage, tracking of temperature differentials with strategically placed TCs provides a means to quantify the number and accumulation of thermal transient events. The signals from the TCs are analyzed to log temperature differential events exceeding some threshold, providing valuable data that can be used to target inspections and plan outage scopes more efficiently. Our software can be configured to provide email alerts when certain magnitude events occur or based on trends in temperature events. Optionally, if PlantTrack Online is connected to the site data historian, SI can fully implement the PlantTrack Attemperator Damage Tracking module, which uses additional sensor data to aid in diagnosing and trending attemperator damage. Actual diagnoses and recommended remediation involves one of SI’s experts reviewing the data. This is made much easier with all the necessary data being compiled automatically within the PlantTrack system. Typical service includes reviewing the data on a periodic basis (e.g. quarterly) and providing a summary of damage events, likely causes, and recommended actions.

Figure 2 – Wireless sensor network configuration.

Figure 3 – Thermocouple data display in PlantTrack.

Figure 4 – Example automated email alert.

To provide some context, the following information provides two (2) case studies of continuous monitoring where some value was realized.

CASE 1:  Bypass spray water stations (Maryland)

Finding: Noted variances in warm-up line functionality and changes to the circumferential temperature differentials/upshock and downshock of the piping.

A select combined cycle plant (2×1) recently experienced a through-wall leak at a girth weld on one of the HRSG’s hot reheat to condenser bypass line. A ring section containing the failed girth weld was removed and submitted to SI’s Materials Lab in Austin, Texas for review. The examination indicated that the crack was consistent with typical thermal fatigue damage, which is the expected damage mechanism for the area considering the proximity of a spray water station. SI recommended that the plant install local thermocouples (TCs) to assess the magnitude of transients experienced during load change events and normal operation – recommendation was made to instrument all 4 areas (2 hot reheat bypass, 2 high pressure bypass). SI also implemented our proprietary wireless sensor network where a node collects the TC data and transmits it wirelessly to a gateway that transfers the data to the cloud (Figure 5). Understanding the transients is the necessary first step, then evaluating/changing the logic, and follow-up with pertinent NDE inspections to ensure there is an understanding of the potential geometric factors here that could exacerbate any issue.  If follow-on inspections find damage then the plant may also consider FEA/fracture mechanics to assess the timing of run/repair/replace options. It is also important to mention that the failed hot reheat bypass girth weld prompted the installation of a new spray water probe assembly to be completed at a later outage.

Figure 5 – SI’s wireless node at the select combined cycle facility that has TC sensors connected.

SI performed a high-level review of the TC data pre- and post-installation of this new spray water probe assembly during a particular outage and also examined all of the bypass location temperature data:

• Pre-outage (data in Figure 6 shows 10/28/2020 @ ~5:20AM EST)
  • Warm-up line doesn’t appear to be operational – data is similar to post-outage data for the other hot reheat bypass line (has not failed)
  • Several ‘transient’ periods show steady rates of temperature change
    • Sides of the pipe – ambient to ~750-850F and back down to ambient over 20-30 min period
    • Top/bottom – ambient to ~275-400F and back down to ambient over 20-30 min period
    • >400-500F differentials around the circumference

Figure 6 – Pre-outage data for the hot reheat bypass system that experienced a failure.

• Post-outage (data in Figure 7 shows 12/17/2020 @ ~10:50AM EST)
  • Warm-up line appears to be operational – now differs from the other HRSG hot reheat bypass (warm-up line appears to be malfunctioning/not in operation)
  • Several ‘transient’ periods show much more prominent upshock and downshock (~275F/min in the plot below)
  • Sides of the pipe – steady from ~700-750F
  • Top/bottom – steady from ~700-750F, but then experience differentials after prominent upshocks and downshocks before settling out
    • >250-300F differentials around the circumference
• Consensus on the pre- and post-outage data
  • Temperature differentials for the hot reheat bypass that failed appear to have improved from pre- to post-outage with a new probe assembly, but now with a functional warm-up line there are periods of more prominent temperature transients
    • Differentials around the circumference still exist
  • Spray nozzles can still be optimized

Figure 7 – Post-outage data for the hot reheat bypass system that experienced a failure.

CASE 2:  Reheat interstage spray water stations (Texas)

Finding: Identified unevaporated spray water is present during cold starts and load changes. Resulting inspections identified prominent cracking of the piping in the vicinity of the spray water probe assembly.

A select combined cycle plant (2×1) has a reheat interstage line that was previously identified by plant personnel as having a prominent sag with the low point located near the desuperheater in 2017. A liner is indicated on the drawing, which should protect pipe ID surface from spraywater. However, SI performed a high-level operating data review and did some localized NDE of this region (January 2018).

This initial data review considered existing transmitters (pressure, temperature, valve positions, combustion turbine loads, etc.) and found that there is some indication that the reheater desuperheater spray control valve is not fully closed, or may be leaking under some conditions. A leaking spray water valve could contribute to pipe bowing as that would make the bottom of the pipe colder than the top. Normally, if the desuperheater piping is able to flex, then when it is cold on the bottom and hot on the top it will hog (bow up). If, however, the piping flex is constrained so it cannot hog, then the pipe remains horizontal and a significant tensile stress is developed in the bottom of the pipe. This causes the pipe to effectively “stretch” on the bottom so the bottom is longer, and over time this can lead to a bow down. During the warm start there are a few minutes where the desuperheater pipe is at or below saturation temperature, which could result in condensation in that line. There could also be spray water that has collected in the line prior to startup that takes some time to evaporate. In either case the result would be a top to bottom temperature difference in the pipe.

From the inspection side there were no major issues noted, but a recommendation was made to install surface-mounted thermocouples (TCs) at pertinent locations to assess the magnitude of thermal transients experienced during load change events and normal operation. Plant personnel installed 5 TCs (2 upstream of the liner/2 downstream of the liner at the top and bottom of the piping; 1 at the extrados of the downstream bend). Plant personnel routed the TC sensors to their data historian (PI) for continuous monitoring.

SI was then requested to perform a review of this second dataset to determine if there are problematic temperature differentials within this line (October 2020). The data indicated that during the cold start and at low load operation (Figure 8 and Figure 9), the spray flow is not fully mixed and saturated steam is impinging on the top of the pipe downstream of the spray. This prompted another inspection (January 2021) now that saturated steam was identified and also prompted a review of the liner/probe assembly port.

This particular inspection identified circumferential indications consistent with ID-initiated thermal fatigue noted within liner boundary. This damage started at the downstream side of nozzle port and continued axially for ~5’ before dissipating and was located from 10:00à2:00 (top circumference). The through-wall depths were prominent – through-wall failure and several other locations with 40%+ (some rough measurements of ~60-80% through-wall noted, as well). It appeared that possible condensate may be inadvertently leaking around the nozzle assembly and entering the reheat interstage line through the nozzle port/flanged connection – once it hits a little bit of steam flow in the line it may push this into the void between the liner and ID of the pipe.

In areas that were originally slated for inspection (exit of the liner, downstream extrados of the bend), no findings were noted. The unevaporated spraywater that was identified by the secondary data review is obviously not ideal, but damage development is driven by the magnitude of the temperature transient and the cycle count and does not appear to have manifested in service damage at this stage. Continuous monitoring is advisable.

Figure 8 – TC and existing transmitter data for a cold start that revealed unevaporated spray water in the reheat interstage line.

Figure 9 – TC and existing transmitter data for a load change that revealed unevaporated spray water in the reheat interstage line.

(Summary)

The crucial aspect in assessing the performance of these systems with spray water stations is being able to determine the magnitude and frequency of thermal transients. With the nearest temperature transmitters (thermoelements) typically located far downstream, local thermal transients at the conditioning valve and desuperheaters are often not detected. Surface-mounted thermocouples routed to the data historian/digital control system (DCS) or collected wirelessly help to evaluate temperature differentials around the pipe circumference and at geometrical impingement points. This, in conjunction with existing transmitters, allows for early detection of potentially damaging events so that appropriate mitigations (maintenance, logic updates, etc.) can be performed before costly repairs are required.