AHUG: Sharing knowledge globally

The HRSG User’s Group is well known to readers as the “go-to” organization in the US for information on heat-recovery steam generators and other Rankine Cycle equipment used in combined-cycle plants. The group meets annually and attracts about 350 participants to its conference and exposition. Some attendees come from nations outside North America, but not many.

So Chairman Bob Anderson and Dr Barry Dooley of Structural Integ­rity Inc decided to do some mission­ary work (Sidebar 1). Their first off-shore conference was held in Brisbane, Australia, last Decem­ber, and coincided with the launch of the Australasian HRSG Users Group (AHUG). There were over 50 attendees at the two-day meeting, half employed by owner/operators of generating assets powered by gas turbines (Sidebar 2).

Not surprisingly, most of the issues faced by the AHUG participants are the same as, or similar to, those encountered in North America. The meeting summary that follows is based in large part on notes taken by David Addison, principal, Thermal Chemis­try Ltd, Horsham Downs, Hamilton, NZ, who was representing CS Ener­gy, a Queensland government-owned corporation and one of Australia’s top electricity sup­pliers. Its stations can generate more than 3000 MW, relying on a mix of coal, natural gas, and landfill gas.

Program highlights

The meeting offered a mixture of focused open-discussion periods sepa­rated by prepared presentations—much the same format as the US con­ference. Discussion topics included heat-transfer components and pres­sure parts; cycle chemistry; piping systems; ductwork, dampers, stacks; materials issues, flow-accelerated corrosion; risk-based inspection; life-cycle costs.

Anderson got the group’s adrena­line flowing quickly by opening the conference with this question: How many attendees have experienced tube failures in their HRSGs? About a half-dozen hands shot up immedi­ately. But judging from the ensuing discussion on evaporator and super­heater tube failures and tube-to-header failures in economizers, many others in the room could have raised their hands. It can take time for con­servative engineers to get used to Anderson’s style.

1. Meet the AHUG leadership

AHUG leaders Robert W Anderson and Barry Dooley, PhD, DSc, are among the world’s leading experts in the design, construction, operation, and maintenance of heat-recovery steam generators and other Rankine Cycle equipment.

Anderson leads Competitive Power Resources Corp, Palmetto, Fla, a consulting firm he established in 1994 to assist clients in the management of powerplant assets (anderson@competitivepower.us). He is a 33-yr veteran of Florida Power Corp/Progress Energy, where he served as a plant manager of several thermal generation assets and later as the corporation’s fleet-wide HRSG solutions provider. Anderson stands tall among the industry’s technical discussion leaders.

Dooley is one of the world’s foremost authorities in the fields of powerplant water chemistry and metallurgy (bdooley@structint.com). He is best known to readers for his work in improving both cycle chemistry and the reliability of HRSGs, and in reducing boiler tube failures and flow-accelerated corrosion. Educated in both the UK and Russia, Dooley has authored or co-authored more than 260 technical papers as well as books on boiler-tube failures, steam turbine damage mechanisms, and FAC. He is a senior scientist at Structural Integrity Associates Inc.

The chairman may walk miles up and down aisles during the day—microphone in hand—firing ques­tions, offering solutions, keeping everyone involved. Active partici­pation is critical to the meeting’s success and Anderson is expert at keeping attendees focused on the subject at hand. If the group becomes lethargic, he may ask everyone to stand and run through a regimen of in-place exercises to get more oxygen in the collective bloodstream.

Tube failures. Here’s a summary of Australasian experience on tube failures from the first discussion ses­sion:

  • Thermal fatigue common in the area of tube-to-header welds on both the hot and cold ends of the unit.
  • Flow-accelerated corrosion (FAC) found in many HRSGs at tube bends.
  • Corrosion under insulation found most often in the petrochemical industry where water ingress typically is a major issue. Piping at temperatures less than about 265F is at greatest risk, as are drain lines operated infrequently. One way to prevent drain-line corrosion is to replace insulation with expanded mesh for person­nel protection when the drain is active.
  • Attendees said root-cause analyses (RCAs) were conducted on most failures, which Dooley applauded as a best practice.
  • Users who experienced corrosion of casing seals and have switched to designs that permit flow-through of warm air reported eliminating this nuisance.
  • Some attendees reported “infant mortality” of seam-welded tubes. Most often these failures were experienced during commission­ing.
  • Quenching of economizers attrib­uted to flow starvation during startup because of drum swell generated some discussion. What happens is the economizer acts as a heat sink until steam starts flowing and drum level drops; thermal shock occurs when water flow through the heat-transfer section restarts. Two suggestions to avoid the problem: Establish flow via drains or install recircula­tion pumps to minimize tempera­ture gradients.

Question from the floor asked by Lester Stanley, HRST Inc, who recently returned to company head­quarters in Eden Prairie, Minn, after completing a three-year assignment in Australia: A few HRSGs in North America are struggling with stress-corrosion cracking in carbon-steel economizer tubes—a problem thought linked to the presence of nitrates in gas-side deposits. What has been the experience in Australia?

2. AHUG II, December 7-8

Shortly after this article was written, the Australasian HRSG Users’ Group announced a formal structure for the organization as well as the dates and location of its next meeting.

AHUG’s mission is to provide a forum for owners, operators, manufacturers, service providers, consultants, and others with an interest in heat-recovery steam generators and associated plant processes and equipment to share knowledge and experiences.

The group is chaired by Bob Anderson, principal of Florida-based Competitive Power Resources Corp; Barry Dooley of North Carolina-based Structural Integrity Associates Inc is vice chair. Other members of the Steering Committee are the following:

  • Gary Joy, CS Energy.
  • John Rickerby, Contact Energy.
  • Lester Stanley, HRST Inc.
  • David Addison, Thermal Chemistry Ltd.
  • Daniel Cole, Origin Energy.

The 2010 meeting will be held December 7-8 (summertime in Australia) at the Brisbane Convention and Exhibition Centre. Workshops on flow-accelerated corrosion and P91 piping are scheduled for the day after the meeting and require separate registration. For more information, visit www.ahug.co.nz.

Only one user reported stress-corrosion cracking in his boilers and that occurred at the bends in hori­zontal economizer tubes of a refinery HRSG. An RCA had not yet been performed.

Dooley stepped to the podium after the opening discussion period and made the first of his two presen­tations, this one on trends in cycle-chemistry performance. Here are some industry best practices from that paper:

  • Use of amine blends is ill advised.
  • Regarding phosphate blends, use only trisodium phosphate.
  • Best location for HP evaporator sampling is close to the outlet header of the highest-temperature evaporator tube.
  • Keep the steam turbine dry dur­ing outages to avoid pitting of turbine blades. For details on how to do this, access www.combined­cyclejournal.com/archives.html, click 1Q/2010, click “Preservation program. . . .” on issue cover.
  • Tube samples are important when dense, tenacious deposits are pres­ent. Nondestructive examination (NDE) techniques available for deposit analysis are not sufficient­ly accurate.

Dooley has written extensively on cycle chemistry. For an in-depth look at industry practices, return to the journal’s archives and click 1Q/2009 to access “HRSG assessments identi­fy trends in cycle chemistry, thermal transient performance,” which he co-authored with Anderson.

More discussion followed Dooley’s presentation. One “thread” focused on problems associated with gas-side deposition and fouling in the cold end of HRSGs. A user reported sulfur and rust deposition and was concerned about long-term effects although none were in evidence—yet. Another attendee reported fin corro­sion, yet another said airborne dirt was sticking to heat-transfer surfac­es which had become coated with oil during construction. Surprisingly, by show of hands, no one was monitor­ing backpressure.

Next topic: Long-term inspection plans. First user volunteered that past focus had been on pressure parts, but her company was undergo­ing a philosophical change to a risk-based inspection program for predic­tive maintenance. Work in progress included identifying what/where to inspect and preparing equipment for ease of inspection.

Another attendee offered that a risk rating system had been devel­oped over time for inspections and replacements of virtually all com­ponents. This program melded both the experience of his company and of others. Latter often came from pre­sentations, group discussions, and personal contact at user meetings.

One user spoke about the installa­tion of Riskwise, a risk-based assess­ment program, at his plant. Expecta­tion was that additional NDE would be required on critical steam-pipe welds, etc. Group think was that risk-based assessment is a learning process driven by safety and manage­ment concerns.

Diligence and commitment are key elements of any inspection plan. The point was driven home by Chair­man Anderson who recalled a shop using a heat-treatment procedure for P22 on P91 pipe bends. This segued into an exchange on the importance of verifying materials during bench­mark inspections. Another article in this issue, “P91 commands respect,” illustrates the point well.

Dealing with upsets

Thermal Chemistry’s Addison offered a series of operational best practices in his presentation, “Major HRSG cycle-chemistry upsets: Practical management steps and avoidance strategies.” Addison’s suggestions are based on a combination of back­ground knowledge (a bachelor’s degree in chemistry and a masters in materials) and 13 years of hands-on experience on the deck plates of combined-cycle, conventional steam, and geothermal plants.

His core message was “be pre­pared” to deal with a wide range of possible upset conditions, noting that “things can go bad very, very quick­ly.” Failure to respond immediately with the proper solution can be very expensive, Addison stressed.

The chemist/metallurgist’s “post­er” example was seawater ingress, which can cause significant damage to boilers and steam turbines within minutes. Even repetitive small leaks, he continued, cause cumulative dam­age over the long term, increasing the likelihood of component failures—boiler tubes and turbine blades, for example.

“While you cannot eliminate all chemistry risk,” Addison said, “you can minimize it to a large degree.” His recipe for minimizing the risk of a chemistry “disaster” includes good:

  • Systems. Clear procedures acces­sible to all employees.
  • People: Operators, chemists, and managers.
  • Plant equipment: Online instru­mentation, proper dosing, conden­sate polishers.
  • Confidence: Decisions sometimes are required quickly. Empower qualified people to make such decisions.

Contaminated conditioning chemicals—such as ammonia, phos­phate, sodium hydroxide, etc—can be a source of cycle contaminants (chlorides, organics, for example). Consider the possibilities:

1. Storage-tank contamination and failure to properly flush new tanks and connecting piping.
2. Use of incorrect dilution water—perhaps raw water in place of demin­eralized water.
3. Manufacturing errors—use of a wrong chemical or grade.
4. Pump-over errors to wrong tank through contaminated transfer lines.

Contamination often is the cause of a sudden deterioration in evapo­rator/steam cation conductivity and a drop in pH soon after a chemical delivery. If this occurs, plant proce­dures might suggest that operators stop the chemical feed, switch to alternative supplies, and/or increase blowdown; or in an extreme case, shut down the unit.

Water-treatment-plant upsets can be caused by the failure of ion-exchange resins; regeneration issues, such as wastes finding their way into demin tanks; membrane failures; tuning difficulties following a main­tenance outage, etc.

A sudden change in evapora­tor/steam cation conductivity and decrease in pH linked to makeup system operation indicates an upset condition. First action should be to check operation of the water treat­ment plant and dump contaminated water from storage tanks; then con­sider unit shutdown.

Addison said you can minimize the likelihood of future upsets by review­ing the design of the water treatment plant and making judicious improve­ments, pre-testing new resins, pro­viding for online analysis of makeup water, improving interlocks, and increasing the level of training for water-plant operators.

A condensate polisher is strongly recommended by Addison for most plants. However, it can be a double-edge sword. When operat­ing properly it assures top-quality feedwater; when not, it can be a source of cycle contamination. Pos­sible issues include (1) leaching of organics because of resin failure, (2) contamination resulting from poor regeneration, (3) incorrect ammonia cycle operation with standard mixed beds, and (4) anionic leaching as a result of running to exhaustion.

Polisher upsets are identified by poor quality of outlet water, ele­vated evaporator cation conductiv­ity, depressed evaporator pH, and increased cation conductivity of steam. Suggested first response: Shut down the polisher and increase HRSG blow­down. Then consider unit shutdown.

Consider the following to avoid upsets in the future: (1) Tighten pro­cedures for resin testing and commis­sioning, (2) extend recycle on startup, and (3) establish stringent quality limits for automatic trips.

HRSG contamination from out­age activities. Failure to remove waterside debris at the end of an outage can cause tube blockages and resultant overheating failures. Con­taminants from boots, NDE solvents, weld repairs, etc, are other concerns.

Elevated levels of silica and cat­ion conductivity in the evaporator, unusual organics, and an increase in steam cation conductivity are indica­tive of contamination. Suggested corrective action: Heavy blowdown, maintain drum pressure below nor­mal operating pressure until the steam/water circuit is clean, then shut down the unit and dump/flush.

Eliminating contamination, or min­imizing its effects, is relatively simple and all about good procedures and discipline. First and foremost, enforce clean internal-access procedures; fill, flush, and dump the HRSG after each outage requiring internal entry; maintain heavy blowdown when the boiler is returned to service until sys­tem chemistry is within spec.

Condenser tube leaks happen, which is why a polishers should be designed into the condensate/feedwa­ter circuits of all water-cooled E- and F-class combined-cycle plants. Addi­son told the group that the risks asso­ciated with a condenser tube leak in a plant without condensate polishing are major. Even titanium-tubed con­densers fail, he reminded.

Cooling water’s high concentra­tions of total dissolved solids are of greatest concern—especially those high in chlorides. Operators must be prepared to respond quickly if online analyzers go “off scale.” Unprepared personnel can become confused and delay action necessary to minimize equipment damage. A characteris­tic of condenser tube failures is the sequential increase in cation con­ductivity through the cycle. Keep in mind, too, that contamination of steam may be virtually instanta­neous because condensate is used for desuperheating.

Recommended response: Blow­down, drop load, increase evaporator pH, stop attemperation, and shut down the unit. Addison said there are two schools of thought on what to do next:

1. Drain the HRSG as soon as pos­sible to remove the contaminants. However, this course of action comes with the risk of “baking in chlorides.”
2. Allow the HRSG to cool, ammo­niate heavily to form highly soluble ammonium chloride and then drain and flush repeatedly.

Consider chemical cleaning, too, unless damage is so extensive that it is necessary to replace the HRSG. An alternative to chemical cleaning: Restore the unit to normal operation and deal with pressure-part failures as they occur. Automatic unit trips based on conductivity readings are something to think about going for­ward. So, too, is the installation of a condensate polisher.

Summing up, Addison stressed that poor operating practices and operational mistakes either cause or exacerbate many of the chemistry upsets plants experience. He sug­gested that each plant conduct a site-specific cycle-chemistry risk assess­ment and create a living document that provides clear and robust O&M guidelines to mitigate risk. The docu­ment should be updated annually or more frequently.

Development of your plant’s guide­lines should be a multi-disciplined process for best results—one involv­ing inputs from operators, engineers, managers, and chemists. At the end of the process, all participants will know their roles, responsibilities, and decision-making authority.

Perhaps the most important sec­tion in your guidelines will be the one with procedures for responding to a cycle-chemistry upset. It should stress the importance of initiating remedial action as soon as possible. Some points to remember:

  • Blowdown almost always is a good initial response, the heavier the better.
  • Reduce load and drum pressure to buy time.
  • Protect the steam path and tur­bine. Halt attemperation and begin steam-turbine bypass opera­tion quickly.
  • When in doubt, shut down the unit.

Thermal transients

Anderson replaced Addison at the podium and addressed thermal tran­sients in HRSGs. Much of what he covered can be found in the 1Q/2009 article referenced earlier that he co-authored with Dooley. However, a couple of O&M “pearls” are worth repeating:

  • On startup, it’s better to let steam temperatures go high rather than over spray. One second of over spray does far more damage in terms of thermal fatigue than the creep damage associated with operating for one hour at a steam temperature above design.
  • Attemperator maintenance is important for plants starting daily. Focus on isolation valves, to prevent water in-leakage. The latest edition of the ASME Boiler and Pressure Vessel Code requires a drain downstream of the attem­perator.

Allan Evans of Origin Energy took his turn at the podium with a presentation on HRSG controls. He opened with an entertaining history of boiler controls from the pneumat­ics era forward that included vintage photos of Bailey boards. But the real purpose of Evans’ presentation was to provide guidelines on HRSG control system arrangement and alarm management. Regarding the latter, he noted that the number of configured alarms per operator has grown from less than 100 in 1960 to about 8000 today—clearly ridiculous. Think of alarms as one area where less probably is more.

23 skidoo. Anderson started the next discussion session by gaug­ing interest in advanced materials, such as T23. End users have identi­fied problems with T23 and ASME recently issued Code Case 2199 to tighten up its chemistry. Attendees expressed little interest in the sub­ject. One metallurgist recently told the editors, tongue-in-cheek, that the higher number—referring to the industry standard T22—didn’t make the material better.

Manway gaskets. The chairman quickly moved to a more practical problem: Sudden failure of steam-drum manway gaskets under pres­sure. Anderson is familiar with this issue and suggested the following actions to minimize the possibility of occurrence:

  • Be sure the correct gasket mate­rial is installed; use OEM parts.
  • Verify proper closure and sealing of the drum door before restarting the unit.
  • Do not make any change to the drum door that is not approved by the OEM.

91 roulette. The ensuing discus­sion on 91 material created a buzz. One user said that based on his com­pany’s experience, 183 HV (Hardness- Vickers) is the minimum acceptable hardness. Below that creep-life issues emerge. Another attendee cautioned against condemning P91 based on hardness measurements alone; met­allographic analysis is necessary for accurate decision-making, he said.

According to the calculator at www.tribology-abc.com, 183 HV is equivalent to 183 HB (Hardness- Brinell). In the article on P91 in this issue (see reference above), Steve Gressler, a metallurgist with Struc­tural Integrity Associates Inc, sug­gested 190 HB as a lower limit. So, it seems that experts east and west are in general agreement.

Another attendee said he had heard that some plants in the US were considering the replacement of P91 with P22 because of the uncer­tainties associated with the 91 mate­rial. The group agreed that tight 91 specifications and diligent enforce­ment of those specs were critical to successful application of the mate­rial. A suggestion offered: If you run into a problem with P91, seek advice from people and companies with rel­evant experience. There’s no reasonto repeat mistakes.

No one disagreed with the thought that leak-before-break is valid for P91 external to the HRSG. Ander­son picked up the microphone at this point and recommended always shutting down and depressurizing the system in the event water is seen dripping from piping. Only then should you remove insulation. Safety first! A representative of a New Zea­land utility agreed and said his com­pany treats all leaks as major until proven otherwise.

The subject of 92 was raised briefly because only one plant owner represented at the meeting had the material installed—T92 tubes weld­ed to P91 headers—and its experi­ence was limited. Oxide formation associated with 92 was raised as a concern because it was not well understood.

Oil ingress into the condensate/feedwater circuit was a subject of interest to plant representatives. Two of the three main contributors to this dialog said they had to chemi­cally clean and flush their HRSGs as a result. Two of the three said the source of oil was a turbine bearing leak, the other said it came from air-blow compressors in use during com­missioning.

FAC guru

Dooley’s second presentation was on FAC in HRSGs, a subject he could have lectured on for two days. One of the world’s top experts on the subject, Dooley began with a simple statement: “FAC involves rapid mass transport, magne­tite dissolution, and wall thinning with local condi­tions becoming more reducing.” He stressed that FAC is not a one-time issue and that the cure requires addressing the root cause with a corpo­rate/plant program.

Circuits sus­ceptible to FAC in combined-cycle plants include the feedwater system, deaerator, econo­mizer tubing, LP evaporator tubing, LP drum internals, lower drums in the LP section, and air-cooled condens­ers. Dooley went on to identify locations where FAC is known to occur in HRSGs and other equipment, showing dozens of photos so attendees knew exactly what to look for when they returned to their plants (Figs 1, 2).

Next he ran through the causes of single- and two-phase FAC and the parameters that influence their occurrence—such as pH, tempera­ture, water velocity, etc—and the benefits of alloying elements such as chromium. Perhaps the most impor­tant thing to remember from a Dool­ey lecture on FAC: Never use reduc­ing agents in your steam/condensate system—not even for layup/storage.

The latest thinking and experi­ence on this subject will be covered in a special report next issue based on the presentations and discussions at the recent international conference, “Fossil FAC,” developed and chaired by Dooley and one of his colleagues from Structural Integrity.

Air-cooled condensers. The pre­sentation on ACCs by Ian Richardson of CS Energy was based on experi­ence at a supercritical coal-fired unit, but much of the material dissemi­nated also was applicable to conven­tional subcritical steam boilers and HRSGs. The big benefit of ACCs is that absent cooling water they facili­tate the siting of generating facili­ties in arid areas. Plus, plants with dry cooling avoid many of the cycle-chemistry risks associated with wet cooling as described earlier.

However, ACCs have a negative impact on plant heat rate, and are prone to FAC, air in-leakage, and production of significant amounts of corrosion products because of the huge amount of surface area required to condense turbine exhaust steam.

FAC is something every ACC owner/operator wants to avoid because the metal loss (1) contributes to high particulate loading in conden­sate filters and polishers and high concentrations of iron in feedwater, (2) is conducive to air in-leakage, and (3) increases the potential for fouling of, and damage to, boiler heat-trans­fer surfaces and turbine steam-path components (Fig 3).

Richardson stressed that tight pH control is critical to minimizing corrosion, which typically is found in the turbine exhaust section, steam ducting from turbine to ACC, and condenser headers and tubes (Fig 4). While the rate of corrosion can be controlled, it’s not possible to eliminate it altogether and conden­sate treatment is necessary. Options include filtration and/or polishing.

Owner/operators wanting to learn more about the operation and main­tenance of ACCs, and share their experiences with others, should attend the second annual meeting of the ACC Users Group, September 28-29, at Xcel Energy’s Comanche Station in Pueblo, Colo. Contact San­dra Brown, systems chemist, today (sandra.k.brown@xcelenergy.com, 719-549-3784).

Operating experience. Several attendees from CS Energy’s Swan­bank Power Station reviewed the plant’s experience with its unfired triple-pressure HRSG. The 1 × 1, single-shaft, 385-MW, combined-cycle facility reached 48,000 equiva­lent operating hours in June 2009 and conducted a 55-day outage. It included the second C inspections for the gas turbine and HRSG, the first B inspection on the generator, and the steam turbine’s first IP/LP C inspection. The F-class facility, commissioned in March 2002, oper­ates on both natural gas and coal-bed methane.

The HRSG inspection revealed corrosion both in the back end of the unit and the stack, and cracking of the following components:

  • Desuperheaters—HP nozzle and liner (Fig 5); hot-reheat (HRH) piping.
  • HP steam-flow venturi.
  • P91—HP drains, HRH thermocou­ple pocket and fixed-anchor-point saddle (Fig 6).
  • Casing—HRSG transition immedi­ately ahead of HP superheater 3.
  • HP-drum support saddle (Fig 7).

Not surprisingly, perhaps, large-bore bends in the HP and HRH P91 piping were found “soft.” The com­pany’s position is that parent and weld material must be at least 183 HV. The lowest recorded readings on HP elbows were between 132 and 142 HV.

The soft pipe triggered an imme­diate and extensive nondestruc­tive examination of HRSG piping. Lifetime implications still are being assessed. Accelerated creep testing is underway and a monitoring program has been implemented. Creep tests are being conducted at elevated tem­perature for times ranging from 1000 hours to six months. Ccj