Risk-based inspections, planning on the minds of users at AHUG annual meeting

The Australasian HRSG Users Group annual meeting (AHUG2016), held November 15-17, 2016 in Sydney, Australia, covered a full range of topics, including:

      • Results of risk-based inspections vary across fleet (go)

      • Long-term layup (go)

      • Oxygen control (go)

      • Applying RBI to chemistry (go)

      • Transitioning from baseload to flexible operation (go)

      • Reheater tube failure (go)

      • Modify procedures to support flexible operation (go)

      • Economizer tube failure (go)

      • Findings illustrate value of inspection (go)

Each annual conference, combined with its associated workshops, seems to develop an over-riding theme—one that threads its way through the presentations and on-floor discussions. For AHUG2016, the recurrent theme was risk-based inspections (RBI).  

By better managing the inspection process, so the discussions went, owner/operators can perhaps extend outage schedules without harm to key equipment, reduce downtime, and increase revenues. But such an effort requires both precision and management.

Results of risk-based inspections vary across fleet. Mark Utley, Contact Energy (CE), New Zealand, and a member of the AHUG steering committee, led off Day One with a detailed look at risk-based inspection programs for boilers, pressure vessels, piping, and protection systems (such as safety valves) at three plants. In Australia/New Zealand, the base inspection period (starting point) for equipment under pressure is one year as required by national codes. Extending the inspection interval can save both time and money.

Piping example: Risk-related features for high-energy-piping (HEP) can include weld condition, flow accelerated corrosion (FAC), corrosion under insulation, and system chemistry. To develop the RBI approach, personnel must identify all relevant features of the system, then collect and record clear and comprehensive baseline data. And sometimes, stated Utley, “when you start the inspection you begin to find completely unexpected problems—such as inaccurate construction data.”

He then offered the following sample priorities based on perceived risk: areas of highest stress, areas not previously recorded (undocumented welds, for example), areas of less-than-ideal geometries, and areas repaired during fabrication and/or operation.

Throughout the RBI process, Utley suggested that his colleagues focus on these fundamental questions:

      1. 1. Do we understand what can go wrong?

      2. 2. Do we know what systems we have to prevent this from happening?

      3. 3. Do we have information to assure us that our systems are working effectively?

Turning to the three subject plants, Stage One risk-based assessments for HEP considered materials of construction, QA during construction, and NDE methods used. The conclusion: Baseline inspection was needed in the high-energy Grade 91, 22, 11, and HT (heat-treated) stainless piping systems at two of the three plants.

Stage Two assessments focused on pressure equipment, and evaluated the following:

      • Mechanical features that can develop into defects over time, and/or

      • Process mechanisms that can change the condition of the pressure equipment and threaten its integrity.

This led to several subsets of inspection criteria for the same two sites. Precise details were given by the speaker.

Next, an overall inspection and test plan (OITP) for a nominal 10-year period for pressure equipment was developed, to be reviewed at least annually. The OITP must be all-inclusive and documented, with a clear audit trail. Application is then made to the New Zealand regulators.

For these three plants, regulatory reviews resulted thusly:

      • Plant A moved from a one- to three-year internal inspection interval to match major outages.

      • Plant B is moving toward a three-year interval.

      • Plant C is shut down because of market conditions.

To adjust to the new schedules, Contact Energy cycle-chemistry improvements are now taking place related to layup and storage procedures, improved instrumentation levels and equipment, and full compliance with IAPWS-recommended online chemistry monitoring of feedwater, evaporator, and steam circuits. AHUG Chairman Barry Dooley and Committee Member David Addison would cover IAPWS Technical Guidance Documents on cycle chemistry on Day Two.

During questions and comments (some focused on local regulatory bodies and responsibilities), Utley offered a sharp and direct summary: “The owner/operator is ultimately responsible,” he affirmed. “If you need to push plants harder, then an RBI process can give you confidence that you have the right information to do so.”

Long-term layup. Stanwell Corp’s 375-MW Swanbank E Power Station in Queensland was removed from service in December 2014, primarily to benefit from the increasing world-market value of its gas entitlements. The unit’s return to service (originally 2017) has now been extended to 2018, increasing various long-term storage requirements and risks.

Stanwell’s John Blake, a member of the AHUG steering committee, explained how the site continues to implement and expand comprehensive cold-storage and preservation techniques for all systems. This was consistent with his 2015 presentation that explained how the small caretaker team does not simply set and forget. It remains alert and looks for improvements.

Some of the 13 dehumidifiers have experienced component failures, so crossover ducts are now installed for both redundancy and to maintain air turnover during repairs. Also, initial attempts to make an air-inlet duct balloon seal completely were not successful. HDPE sheets offered improved seals. Specifically, 600 1.5-mm sheets were installed between the pocket frames and filters, holding relative humidity (RH) below 10% back to the GT exhaust.

The presentation and discussion generated many questions and comments on a range of topics.

Blake’s summary statement fit the venue: “Long-term storage has a number of unknowns. It is good to come to these user group meetings to share our experiences and learn from others.”

Oxygen control. David Williams, APA Group, followed with a user case history on oxygen control for new HRSGs, describing the company’s 240-MW Diamantina Power Station in Queensland, which opened at the end of 2014.

After commissioning, cycle chemistry was within acceptable limits and controllable, except for feedwater dissolved oxygen (3 ppb versus the specified 5 to 20 ppb). Cycle chemistry control was achieved through condensate and feedwater dosing and blowdown. Unit chemistry was designed to operate as an AVT(O) system. (Proper oxygen in the feedwater economizers is critical to allow formation of protective layers, minimizing FAC.) After 12 months of operation, a 3-mm wall-thickness loss had occurred in the HP and LP economizers. Focused and comprehensive investigations began.

A critical discovery: The deaerator venting arrangement had not been installed as recommended by the manufacturer. Modifications were made and dissolved oxygen brought under control. Investigation after 12 months showed no evidence of FAC. Iron-transport studies revealed a decrease in total iron levels (soluble and particulate), and magnetite levels through the condensate system were noticeably reduced. The inspection interval then was extended using a risk-based approach.

Applying RBI to chemistry. One of the largest industrial projects in the world is a major gas/LNG export facility in Australia that includes a 500-MW combined-cycle plant. Hayden Henderson, a member of the AHUG steering committee, described this generating facility.

The complex features five GE Frame 6 machines with HRSGs, and three 100-MW steam turbines fed by a common steam header that connects to three isopentane utility boilers. Overall plant design features N+1 redundancy so that turbine trips do not affect LNG production. The plant supplies power only to the LNG complex and is not grid-connected.

The task discussed at AHUG2016: A collaborative effort to design a risk-based inspection program for all pressure vessels, piping, and utilities within the boundaries of the combined-cycle plant, and to extend inspection intervals from 12 to 36 months. The goal:

      • Highlight all relevant degradation-mechanism likelihoods and consequences in all piping and pressure equipment, and

      • Create inspection plans, drawings, and written schemes of examination for all piping and pressure equipment.

Regulations would allow the extension if the proper RBI program were used, the boiler had adequate water-treatment facilities, and the equipment had a demonstrated history of reliability.

Henderson’s RBI presentation focused on station chemistry to define the inspection plan.

At project launch, data were limited, power demand and operating modes were not known, cycle-chemistry guidelines (from the EPC contractor) were not adequate, and co-owner (UK-based Petrofac Corp and Japan’s INPEX Corp) methodologies focused on oil and gas facilities, not combined cycles.

Chemistry guidelines were critical. IAPWS and EPRI guidelines were reviewed, leading to fixed integrity operating windows (IOWs) for both feedwater and boiler chemistry. All decisions were made through workshops in a controlled collaborative process, targeting international best-practice cycle chemistry. NDE inspection would be completed at least twice within the first 36 months.

The IOWs are now being developed into a cycle-chemistry guidance document. Provided the IOWs are met, the inspection plan is executed, and no defects are found, the boiler internal inspection frequencies can be extended to match hot-gas-path (HGP) inspections.

During discussions, Chairman Barry Dooley offered the following Rule of Thumb: “In these types of plants, corrosion is either on or off; it’s off if you get the chemistry right.”

Transitioning from baseload to flexible operation. Ivan Currie, Energy Australia, then presented cycle-chemistry activities at the 435-MW Tallawarra Power Station where he labeled the pursuits “a kaleidoscope.” The facility was commissioned in 2009 for baseload operation, but is now transitioning to flexible operation. It operates with an oxidizing feedwater chemistry and trisodium phosphate in the evaporators. The plant is saltwater-cooled.

Unstable chemistry (especially evaporators) was making proper dosing difficult, and control alarms were frequent. Causes included low sample flows, suboptimal pH analyzers, frequent load changes (impacting conductivity-based dosing), inappropriate alarm set points, and insufficient sample cooling under thermal-shock conditions.

Monitoring improvements included these:

      • Replaced the solenoid valve in the temperature protection system with a mechanical thermal shutoff valve (TSV).

      • Upgraded pH analyzers.

      • Improved the instrument maintenance schedule.

      • Implemented pH-based trisodium phosphate dosing.

      • Verified proper use of sample conditioning components.

      • Modified alarm set points to align with new chemistry specification.

      • Implemented trim cooling system for sample streams.

 The monitoring improvements were conducive to the following positive results:

      • No nuisance alarms.

      • Stable chemistry and dosing.

      • Confidence in instrumentation.

Also, the plant implemented various instrument upgrades, modified blowdown procedures, and lowered drum level by about 4 in. Some issues remain such as tiger striping in the LP condenser, and discoloration adjacent to the HP feedwater line into the drum. Work continues.

Reheater tube failure. Luke Mosele, site chemist, and Jason Spencer, operator/maintainer HRSG, NewGen Power, discussed tube failures at the 330-MW Kwinana plant in Western Australia. The 1 × 1 combined cycle has a dual-pressure HRSG (with 80 MW of duct firing) coupled to a 160-MW steam turbine. The system includes a seawater-cooled condenser and demineralized water plant (ion exchange).

Investigations into excessive water usage began in February 2016, with tube failures the most probable cause. The plant could not be shut down at the time, and there were no external signs of leakage. Online checks showed a drop in reheater (RH) pressure.

An earlier inspection had identified cracks at several header-to-manifold links on the reheater. However, a mid-February 2016 force-cooled shutdown inspection found that the leak was not originating from the lower or upper header-to-manifold links. The site then filled the reheater section with water via the RH attemperator sprays, to source the leak.

For commercial reasons the plant was returned to service for three days. Then, it was shut down and scaffolding installed. A leaking tube was found four rows back. Several tubes (eight T23 and four T91) had to be removed.

There were challenges. Contractor experience was minimal. Initial heat-treatment methods were slow. Window welding (versus mirror) was used. The initial time estimate of four days grew into seven. Access was difficult and repair costs were high. Removed tubes were then fully analyzed.

Kwinana returned to cycling service for the next three months. But water usage increased again, and a new test found another leak in a similar location. Commercial operations had to continue until the October outage.

Major cracking was discovered in four tubes, and staff decided to plug the tubes at the headers.

The plant returned to service in November 2016, with 68 thermocouples installed to monitor the operational modes (rapid load changes) which could be causing thermal fatigue. Specialized analyses continue.

Modify procedures to support flexible operation. Yarnima Power Station operates on an island grid in remote Western Australia. Challenges include frequent load variations (both daily and seasonal) and a requirement to maintain minimum spinning reserve. Also, original duct-burner design allowed firing only when the respective GT was above 90% capacity, and the plant rarely operates at this level. Following OEM reviews, new duct-firing options were tested, then reviewed and approved by regulators.

The Yarnima complex, which supports iron-ore operations, was described by Mark Watkinson, principal chemist, TW Power Services, includes:

      • Three SGT-800 industrial gas turbines from Siemens.

      • Three HRSGs from RCR Energy.

      • Two SST-400 industrial steam turbines from Siemens.

      • Three black-start diesel/generators.

Newly sunk boreholes supply raw water to the site’s treatment plant and storage tanks. Bore water has high alkalinity and hardness, high silica, and is intermittently contaminated with diesel oil (up to 1 mg/l as TPH, total petroleum hydrocarbons). Also, sources and water quality may change over time or suffer short periods of interruption. Wastewater is routed to two large tanks, then to acid-rock-drainage evaporator ponds.

Water treatment is complex, and no untreated source waters are suitable for cooling-water systems. Treated permeate is used for cooling-tower makeup and for supply to the GT air-intake evaporative coolers. Cooling-water chemistry is operated at four to six cycles of concentration. Many water issues were identified in the presentation. A project to upgrade online instrumentation is in progress.

HRSG cycle chemistry is:

      • AVT(O) for feedwater.

      • Emergency dosing of tri-sodium phosphate (TSP) in the HP drum and evaporator.

      • OEM oxygen scavenger never used; converted to ammonia dosing at deaerator outlet.

One of three HRSGs often is in layup/standby mode, but must be ready to return to service quickly when needed. Nitrogen capping is used for offline periods of up to four days. Beyond that the HRSG is drained and nitrogen blanketed at a positive pressure above 7 psig (no economizer protection).

Ongoing issues include these:

      • CO2 pickup from storage tanks.

      • Lack of monitoring for degassed conductivity and conductivity after cation exchange (CACE) on all units.

      • Corrosion noted in economizers and boiler blowdown.

FAC has not been identified.

In keeping with the risk theme, Watkinson noted that “successful allocation of funds is based on risk profile and sound risk assessment.” Many questions and helpful ideas followed from the participants—including emergency phosphate dosing, possible pitting from poor layup, membrane biofouling, iron testing, fast-response condenser instrumentation, and flame monitoring.

Economizer tube failure. Anita Zunker of PEI Pressure Equipment Integrity explained a recent HP economizer tube failure at the Stratford combined-cycle power station at Taranaki, NZ. Commissioned in 1999, the plant was in baseload service until moving to more flexible operation in 2013, with extended out-of-service periods. “Robust preservation procedures are in place,” stated Zunker.

Current out-of-service programs include:

      • Short term (few days only): wet storage.

      • Longer term: nitrogen capped, drained hot, piping dehumidified.

The tube leak was detected at the end of March 2016. The pinhole leak occurred in HP economizer 7, top tube-to-header weld (sixth tube from east end). Material is carbon steel. The repair method: Six tubes were demolished and the header was cut shorter.

A pressure test then found another leak (seventh tube from west end). That leak was plugged, and metallurgical analysis of the first leak showed oxide-filled cracks and corrosion on both tube and header surfaces (within crevice), along with minor pitting of the internal tube surface.

The cause was a corrosion fatigue crack, with the crevice between tube and header acting as a concentration site for corrodents, and as a stress riser.

Data from 2006 also were reviewed. The conclusion: original welding defects are an influence.

The solution: Concentrate on what can be controlled. Maintain operating and preservation procedures to the highest standards to minimize ongoing risks. Be aware of the issues.

Findings illustrate value of inspection. On Day One, Contact Energy’s Utley discussed extending inspection intervals at three plants. The company’s Rachelle Meijer was at the podium on Day Two to discuss current outage activities at one of those facilities.

A full external inspection had been conducted before the last outage. Hot spots were found around the diverter door and transition duct. A follow-up comprehensive internal inspection included hardness and dye-penetrant testing of superheater tubes. Findings were:

      • Duct liner plate buckling.

      • Duct burner 2 (of 12) had sagged; DB 3 was beginning to sag.

      • Duct-burner final elbows found highly corroded.

Soon after restart, water usage increased and IP steam production decreased. Shutdown followed. A leak was found between the economizer and evaporator headers. Tube samples sent for analysis revealed corrosion fatigue.

Contributing factors were:

      • Thermal shocking during non-optimal trip test.

      • Poor layup and storage in past.

      • Poor design of header and tube attachments.

Moving forward, actions will include:

      • Conduct further inspections at next outage.

      • Install thermocouples and strain gauges.

      • Install nitrogen generator for layup.

      • Update procedures for trip testing.

      • Improve mass balance data.

Posted in HRSGs |

Comments are closed.

Categories