HRSG USERS GROUP

HRSG USERS GROUP

New tools for tube repair, analysis of risks posed by safety/relief valve clinic
ver the past 13 years, the annual conference and exhibition of the HRSG Users Group (HRSG UG) developed a reputation for great venues, exceptional food, andmost importantlya solid technical program that provided participants the practical information needed to reduce outage time and O&M costs, increase plant efficiency and availability, and lower emissions. The 2006 event, held last March at the Broadmoor Hotel in Colorado Springs, certainly enhanced that reputation, exceeding expectations both in creature comforts and in program and exhibition content. Delegates were treated to the organizations most impressive venue
32

and culinary service yet: The Broadmoor is one of only three resorts in the country to have earned the AAA Five-Diamond rating every year since the awards were established in 1976. This years information exchange comprised two jam-packed days of insightful presentations by industry experts, plus the traditional user-driven discussions. The latter were skillfully moderated by Chairman Bob Anderson, principal, Competitive Power Resources Corp, Palmetto, Fla (anderson@ competitivepower.us), and summarized in a special report for attendees by Rob Swanekamp, executive director of the HRSG Users Group ([email protected]). A relaxed, but productive, exhibition enhanced collaboration among

users, manufacturers, architect/ engineers, consulting engineers, and service providers in attendance (see montage above). Another bonus. All attendees received a copy of the HRSG Users Handbook, a new 500-page reference work that contains a wealth of practical knowledge on the procurement, design, operation, and maintenance of heat-recovery steam generators. The editors of the COMBINED CYCLE Journal consider this handbook a must-read for the combinedcycle/cogen community (to order online, access www.hrsgusers.org). From the users vantage point in the Broadmoor Hall, it was clear that the groups Steering CommitteeAnderson, Swanekamp, Yogesh Patel of Tampa Electric Co, and Paul

COMBINED CYCLE JOURNAL, Second Quarter 2006

HRSG USERS GROUP

welding best practices, prolonged low-load operation, highlight annual meeting

Fernandez of GE Contractual Servicesworked hard to produce an enjoyable event, while compiling a program that delivered high-caliber technical information on HRSGs and their associated steam systems.

Constant format, fresh info

Through the years, the meetings format has remained relatively constant: A series of Open Forum Discussionsself-help clinics, if you willdominate the agenda, with only a handful of prepared presentations sprinkled in to cover particularly challenging issues. While this years presentations were interesting and on-point as usual, the things that really separate

1. Steam-plant issues raised by owner/operators at the 2006 HRSG Users Group conference included methods to monitor HRSG performance, the pros and cons of all-volatile water-chemistry treatment, and the challenges associated with controlling superheater and reheater outlet temperatures while at low loads
33

COMBINED CYCLE JOURNAL, Second Quarter 2006

Plant Proven Solutions

Anytime. Anywhere.

(tm)

D.L. Ricci Corp. Pipe Cutting & Beveling Clamshells

Portable, easy to set-up and safe to operate All materials to include P91 & P92 On-site machinists ready to come to your plant

D.L. Ricci Corp. has a Full Line-Up of Insitu Tools

+ Portable Boring Bars + Portable Milling Machines + On-Site Heat Treating + Manway Grinders + Keyway Cutters + Turbine Coupling Milling Machine + Journal Turning Lathes

Our Plant Proven tools are available for:

Sales Rental On-Site Support

D.L. Ricci Corp. Proudly Made in the USA

www.dlricci.com 651-388-8661
this event from the rest of the conference pack are the number of attendeesupwards of 350and the type of people actively contributing to the discussion topicsmostly end users (plant managers, O&M supervisors, operators, maintenance techs, chemists, and so on), with a significant and vocal contingent of plant-support authorities (HRSG OEM engineers, water-chemistry specialists, plant service contractors, engineering consultants, etc). This technically skilled assembly, unleashed to raise whatever steam-plant issues they currently face, ensures that the discussions are fresh and timely year after year. Note that the delegate assembly was particularly robust this year, thanks in part to the active participation of the dog that wags the tail of this industry: Several leaders from GE Energys team that specifies and delivers complete combinedcycle plants participated in the 2006 event. Their technical contributions, and the customer feedback they received, were of significant value to the industry. The formal presentations delivered at this years conference are summarized in the large sidebars included with this report. What follows are some highlights of the Open Forum Discussions, the centerpiece of the HRSG UGs annual conference (Fig 1).

Heat transfer opens the meeting

The conference opened with a question from a refinery operator concerning the tools and methods available to predict HRSG thermal performance, given gas-turbine (GT) and ductburner theoretical heat input data. This design-related discussion then evolved toward practical methods that operators can apply to monitor HRSG performance, given the limited instrumentation installed at the typical plant. For instance, operators can trend the pinch, approach, and

2. Many HRSG users experience tube fouling caused by ammonium salts. But one user at the conference was experiencing this fouling only on the side tubes of his HRSG. Possible causes and cures for this localized deposition were discussed by the 350 attendees
34

COMBINED CYCLE JOURNAL, Second Quarter 2006

CMI AD

2/10/06

2:56 PM

Page 1

stack temperatures, one consulting engineer explained to the crowd. Discussion of this topic might have gone on longer if one of the attendees hadnt noted that Chapter 5.2 of the HRSG Users Handbookdubbed the Blue Book by that individualcovered this topic in great detail. Next came a good example of a user bringing a vexing question to the meeting and getting a definitive answer on the spot. A steam specialist from a southeastern utility displayed a photo of salt deposits that had formed in the cold end of one of his HRSGs, and asked of the audience why these deposits seem to be forming only along the sides of the unit, rather than uniformly across the HRSG (Fig 2). After studying the photo for a minute, one experienced consultant suggested that the pattern was typical of cross-flow tube coils where cold inlet water starts at the side of the coil, and moves across the tube bank progressively picking up heat. The end result is some tubes being warmer than others. The salt deposition selectively builds up on the colder tubes, the consultant explained. The user thought he had his answer, until an engineer from that HRSGs OEM interjected that this particular design did not use any cross-flow coils. The OEM engineer proposed that, instead, these deposits might be forming during shutdown when a slipstream of colder air along the sidewalls is created because of natural draft. Often the corrosion and deposition phenomena are worse when a unit is not operating, than when online and up to temperature, he explained. The proverbial light bulb went off in the users head, as he realized that other evidence he had observed supported the OEMs suggestion. The user had found a very similar, localized pattern of ammonium salts collected around the HRSGs manway door gasketsgaskets that were known to be experiencing air in-leakage. Unfortunately, not all questions during the Open Forum Discussions can be as definitively answered as this one, even by the large group of experts and operators gathered in the room. For example, a user from the Midwest with experience in fossil-fired boilers asked how many years of service the typical HRSG can be expected to deliver before it requires chemical cleaning. The response: It depends. As other attendees explained, there are many variables affecting this intervalvarious design factors, the operators execution of the water chemistry program, and base-load versus cyclic operation, to name a few. The answer also is not definitive because of the lack of data within the industry on tube-deposit formation. The fleet of high-pressure HRSGs in power-generation service, compared to the fleet of fossil boilers, is still quite young, and little, if any, empirical data on tube deposition have been collected and analyzed. As a result, discussion of this question centered around when, where, and how to take evaporator tube samples to analyze the rate of deposit formation in your HRSG. While a definitive answer for the HRSG community awaits such data, a consensus seemed to emerge among the veterans in attendance that a well-designed and well-run HRSG can be expected to go without chemical cleaning for 10 to 15 years. Of course, it only takes a few hours of operation outside accepted water chemistry limits to create a situation that could necessitate immediate chemical cleaning, Chairman Anderson pointed out. Continues on page 39
COMBINED CYCLE JOURNAL, Second Quarter 2006

HRSGs from 30MW to +250MW References with all Combustion Turbine Manufacturers Successful references with Combined Cycle Plants used in Daily Cycling Operating Mode since 1974 Over 500 References Worldwide

5300 Knowledge Parkway, Suite 200 Erie, PA 16510-4660 USA Tel: + 1 814 897 7000 Fax: + 1 814 897 1090 [email protected] www.cmiepti.com

Avenue Greiner, 1 B 4100 Seraing - Belgium Tel: + 32 43 30 24 28 Fax: + 32 43 30 22 00 [email protected] www.cmigroupe.com

35

HRSG USERS GROUP

Welding techniques for a successful plant outage

ver since riveted pressure vessels went the way of the horse n buggy, welding has been an essential and common activity in the construction and maintenance of boilers and piping systems. Most users see it going on during plant outages and some may even get involved in its planning and execution. Most also can list some of the well-known problems associated with weldingsuch as the frustrating length of time required (often on the critical path) for fit up, welding, and post-weld heat treating (PWHT) o f l a rg e p i p e ; Newell rework of rejected welds; and finding sufficient numbers of skilled welders when needed (Fig A). While these are significant issues worthy of the maintenance managers attention, there are many other less obvious considerations associated with planning and executing a welding job that can mean the difference between an on-budget, on-schedule project and the proverbial train wreck. At the 2006 HRSG Users Group meeting, William F Newell, Jr, PE, IWE, a vice president of Euroweld Ltd, Mooresville, NC (wfnewell@ pobox.com) shared with the audience several of these less obviousbut

still importantwelding pitfalls, covering both technical troubles and financial flops.

Measure twice, cut once

Newell opened by emphasizing the importance of planning your outage well in advance. Key planning steps include collecting the applicable specifications and drawings, identifying predictable trouble spots based on previous problems and fleet history, deciding how to work with dissimilar metal joints, obtaining sufficient materials inventory, and so on. He pointed out that these fundamental planning activitiesthough perhaps a bit tedious and unexcitingcan prove highly valuable to a successful outage. Following the review of outageplanning fundamentals, Newell discussed some of the planning tasks that are specific to welding projects. These include becoming familiar with codes, NDE techniques, and acceptance criteria; selecting the optimum weld processes and consumables for each job; procuring the right materials; creating and implementing qualified procedures, engaging qualified and productive welders, andlast but not leastsatisfying the authorized inspector and/or regulatory bodies so all repairs are acceptable for a plant restart. In Newells experience, observing

ASME Boiler & Pressure Vessel Code (Code), specification, and jurisdiction criteriacoupled with hiring the right personnelcan lead directly to enhanced equipment reliability and higher plant availability. But someone at the top must appreciate and understand the job, he said. Thats not always easy given the complexity of welding with advanced alloys. Older materials were very forgiving, compared with the alloys in todays combined-cycle plant, Newell pointed out. And just as welding has gotten more complicated, fewer and fewer people are taking the time to truly understand the discipline.

Material shortages
Another very timely aspect of his presentation dealt with todays challenge of procuring sufficient materials for welding projects. Such items as welding rods, wires, fluxes, and special-alloy piping are in short supply. To a large extent, Newell said, these shortages are attributed to the significant amount of industrial construction going on today. That may seem surprising to power generation professionals in North America and Europe, because new construction in this industry and in these locales is quite sluggish. But there is so much work going on in other parts of the world, particularly in China, Newell said, that it is literally sucking up the

A. Welding projects are a part of virtually all combinedcycle/cogen plant outages. Keeping them on-schedule and within budget can be a serious challenge
36

B. Successful welding jobs begin with proper preparation and fit-up

COMBINED CYCLE JOURNAL, Second Quarter 2006

HRSG USERS GROUP

raw materials we need for our power good fit-up and youll save time in To increase productivity without projects. the long run (Fig B). sacrificing quality, Newell said you Compounding that problem is the n For every welding dollar spent, should cull the welders in the weld fact that most of the weld metal for figure labor and overhead will cost test shoprather than on the job. chrome-moly pipingused exten86 cents, equipment (manual) six High reject rate in the test shop sively in combined-cycle/cogen cents, and filler metal and shield- always equals lower reject rate on plantsis being made overseas ing gasses eight cents. the job, and vice versa. Newell recTo minimize time, Newell sug- ommended requalifying each welder these days. This adds perhaps four to six weeks to the delivery timefor gested that users: on important projects rather than shipment across the oceanand n Coordinate weld-groove geom- accepting CommonArc or UA certifioften keeps much of that piping in its etries with welding process selec- cations unless you have prior positive country of origin. Why should steel tion. experience with these group qualifiproducers ship it to far-flung reaches n Use gas tungsten arc welding cation efforts. of the world, when they can sell it at (GTAW, TIG, heliarc) when posAs evidence, he discussed a case handsome profit to a nearby consible. Although considered slow, it study where all of the above steps struction project? is faster in the long run than other were considered during the planning Its a similar situation even for the processes on small-bore or stan- stage. The project competed ahead base metals, Newell said. of schedule with a reject Reportedly there is only one rate of less than 1%. Of US vendor that can make course, there were other seamless pipe, and that supdetails during execution of plier is very busy. We put the work that also had a in an inquiry for strip cladgreat impact on this sucding in 250-ton increments, cess. For example, the and we were laughed at, best planned welding job Newell reported. They said, wont be successful if you Dont bother us. Were makdont keep the arc on and ing money selling the cheap hot, he said. stuff. We dont want to make No welding process anything sophisticated. And permits deposition of weld this was a premier melter! metal 100% of the time. Newell cautioned HRSG Some things that require users that this situation interruption of welding could become very critical include changing elecif they have an emergency trodes, the welder reponeed for a particular fitting sitioning, chipping slag, or special piping. Trying to grinding, preheat/PWHT, C. Post-weld heat treatment is needed for many welding get that special fitting, or joint preparation and perforging, or valve you need projects, and is essential if youre dealing with P91/T91 sonal time. material [in an emergency] could be Selection of the welding very problematic, he said. process also can have an Moral of the story, as Aesops ant dard-wall piping. impact on welding productivity. For understood but his grasshopper did n Implement processes to be used example, typical operating factors for not: Better lay in some stores while based on who shows up. In some common welding processes you can! other words, coordinate the selec- when used in the shop are 30% for tion of welding processes with the manual GTAW and shielded metal arc available skilled labor pool. welding (SMAW or stick), 40% for Cost and schedule n Pre-select or pre-qualify your semi-automatic gas metal arc welding Estimating and controlling costs and heat-treatment contractor. (GMAW, MIG) and flux cored arc weldschedule challenge plant managers Plan how the heat treatment on ing (FCAW), and 50% for fully autoduring most welding projects. Thats each job will be conducted. Is it best matic submerged arc welding (SAW). because there are many variables to preheat and weld, then to perform All of the above values are at that can enter into the equation. PWHT on a number of joints? Will least 10 percentage points lower for Newell offered a few rules of thumb you heat-treat single versus multiple field application. This 10 percentage that hes found handy: components? Should you cluster the point reduction correlates with only n Regardless of the job, assume pre- heat treating activities in one area, six additional minutes each hour heat and PWHT will be required, or heat-treat individual components when weld metal is not being deposand budget your money and time in remote locations? Maybe its best ited, but it can drive a 21% to 50% accordingly. to use a combinationperform indi- increase (dependent upon electrode n Regardless of the job, skilled craft vidual preheat, then batch PWHT (Fig size and process) in the cost per and welders are necessary. Dont C). pound to deposit the weld metal. think you can save money by hirAlso plan the heat sources to be One of the things you can do to ing on the cheap. used: flame (not on the latest alloys), increase welder productivity is to n Regardless of the job, certified electrical resistance, electrical induc- select welding processes and conmaterial test reports for weld metal tion, or portable furnaces (gas or sumables that appeal to the operashould be required because theyll electric). Today it is not uncommon tor. This means those that produce reduce your project headaches. for the alloys and configurations to a smooth arc, good control of the n Regardless of the job, demand dictate the methods available. puddle and bead, reduce weld cleanCOMBINED CYCLE JOURNAL, Second Quarter 2006 37

The pneumatically powered ESCO Millhog Tube Fin Removal Tool removes fins from the tube O.D. quickly and easily. The standard tool removes 4 in. of fin in less than two minutes and bevels the tube end all in the same step. Other features of the tool include: no reaction torque to operator, easy to use, helps reduce outage time.
Esco Tool, 50 Park St., Medfield, MA 02052 Tel: 800-343-6926, 508-359-4311 Fax: 508-359-4145 E-mail: [email protected] Please visit our website at www.escotool.com

ing and grinding by producing slag that is easy to remove, produce no spatter and produce good bead shape. Its also important, Newell said, to take steps that reduce the welders personal time requirements (distance to water cooler and rest room) and eliminate non-welding tasks and diversions. As for keeping the arc hot, Newell explained that reducing the current 10% below optimum when using the FCAW process will result in a decrease in weld metal deposition rate of 13% and an increase of deposition cost of 15%. Another factor that can have a significant effect on the cost of the welding project and its schedule is the choice of welding wire when using the FCAW process. For example, a low-alloy wire with an efficiency of 60% may cost $10/lb, while a wire with an efficiency of 90% may cost $14/lb. Therefore, to get one pound of metal deposited requires 1.7 lb of the less expensive wire, but only 1.1 lb of the premium material. If thats not already bad enough, most of the lost metal ends up as spatter that requires grinding. You can see where this is going. If you do the math, the cheap wire actually costs $17/lb of deposited
38

metal while the more expensive wire can be deposited for $15.40/lb. Which one would you choose?

Using Code as king

One of Newells concluding messages dealt with the Code, and the propensity of todays users to rely on it as some sort of infallible construction handbook. Many purchasers of combined-cycle and cogeneration plants are under the impression that their specifications will be sufficient as long as they require compliance with the Code. But the sole purpose of the Code, Newell reminded the audience, is to ensure safetyto prevent catastrophic failures and the resultant loss of life, injury, and property. Its a set of minimum requirements only the bare-bones minimum. If you think you need more, then its up to you to do more. The comment, triggered by a question from a user in the audience, prompted Newell to regale the crowd with a short anecdote based on his many years of experience in welding and on Code committees. We were sitting in an ASME Strength of Weldments Subgroup meeting, Newell recalled, and three Fellows present

reasoned that the problem with the Code and the allowables and everything else is that no rule-of-thumb is provided as a guide and young designers sitting in front of their computer screens today actually believe what theyre seeing. In the past, Newell continued, designers would determine what the Code required, and would then add a margin of safety. If I need an inch, Ill use an inch and a quarter. If I need an inch and a half, Ill use an inch and three quarters. If I need five inches, Ill use seven. But those designersand their conservative design philosophyare no longer in charge. Newell also pointed out that the advanced materials being used in HRSGsand in the next generation of fossil boilersare simply not as forgiving as the older, traditional steels. Plus were running these components at higher stress levels. Net result, Newell said, is that users do need to be more diligent in their specification, than to simply say comply with ASME Code. Meanwhile, Newell, along with many other materials and design experts, are working within ASME committees to set more prescriptive standards when appropriate.

COMBINED CYCLE JOURNAL, Second Quarter 2006

WATER TREATMENT

3. Water-chemistry authoritieslike Rich McKee (left) and Jim Witherow, of Nalco Co, Naperville, Ill, and Scientech LLC, Clearwater, Fla, respectivelyresponded to users questions about moving away from their conventional phosphate treatment programs and toward all-volatile treatment Continued from page 35 At early HRSG UG conferences, Water chemistry, as leakage around drum-door gaskets and their catastrophic failure were always revealed to be relatively common Discussions surrounding water problems in the industry. Hence, the chemistry are always popular at reasons causing the leaks and the HRSG UG meetings, and the 2006 methods to avoid them usually gar- event was no exception. This year nered considerable discussion. At the the topic that seemed to garner the 2006 event, drum-door leakage was most interest is the desire, if not an raised yet again as a serious issue, outright trend, for operators to move suggesting that even after more than toward all-volatile treatment (AVT) a decade of US experience, this prob- and away from their conventional lem remains. As usual, some excel- phosphate treatment programs (Fig lent tips were bantered about, but 3). Discussion on this topic was so this years discussion of drum-door lively, that an excerpt is presented gaskets didnt consume a lengthy below. It was extracted directly from time period because this is another the meeting notes with permission of topic now covered in detail in the the HRSG Users Group. HRSG Users Handbook User: One of our consultants has COMBINED CYCLE JOURNAL, Fourth Quarter 2005
COMBINED CYCLE JOURNAL, Second Quarter 2006

recommended going to AVT. Were having issues with phosphate hideout; result is that every time we start up, pH drops down in the 8-range for probably the first one or two hours that were operating. If people are not switching to an AVT program, Id be curious as to why theyre not. What are the concerns? Chemical vendor: To answer the first question, about experience in changing to AVT-O: It is happening right now in many plants. Theyve had, generally, a good degree of success. But there are a number of qualifiers that you have to meet at the plant level, and at the chemistry level, in order to make that change successfully. Issues associated with dissolved oxygen in the hotwell; with air in-leakage; with what happens when youre cycling, etc, all can be addressed mechanically and chemically. When people have done that, theyve had good results, in terms of iron transport, and generally theyre very satisfied with the switch to AVT-O, and also with the change to no phosphate in the drums. So its been effective, and from my point of view its an industry trend that well see more of. To make the transition it requires taking care of the mechanical, as well as the chemical, things. Chairman Anderson: I want to ask a couple of follow-up questions. When you say AVT-O, are you talking about true all-volatile treatment, or are you just talking about very low phosphate levels to avoid the hideout. And if youre talking all-volatile treatment, are you using a condensate polisher? Chemical vendor: The terminology can be a bit confusing. Of course, 1
39

we always use all-volatile treatment in the condensate/feedwater system, so even if you use phosphate in the boiler, you are on an AVT program in the condensate/feedwater. The AVT-O specifically refers to no use of passivatorsso no hydrazine, no carbohydrazide, no other chemical

passivator for an all-steel system. And that has been a very successful change, whether youre still using phosphate in the drums or not. Now, going completely to a nonphosphate treatment in the drum is something we also see people doing, without condensate polishing, as

long as they have tight condensers. Many plants recently commissioned that dont have condenser tube leaks have gone to all-volatile treatment with no polishersuccessfully. Some go to a backup of phosphate in case they do have a condensate cation conductivity issue. Others have

Workshop to explore water chemistry, outage practices

The HRSG Users Group, an organization that welcomes as members all combined-cycle/cogen professionals, recently announced dates for its next two events. Furthest away is the 15th Annual Conference & Expo, which will be held Mar 26-28, 2007, in Kansas City, Mo. Preliminary details are available at www.hrsgusers.org. This fall, the organization will conduct another of its popular Steam Plant Workshopsthis one co-located with Power-Gen International, in Orlando, November 2830. Workshops, conducted periodically, are intended to supplementnot replacethe annual conference. The latter encourages free-ranging discussion on a broad array of issues facing users; the workshops are in-depth seminars that focus on one or two topics. First day of the fall 2006 workshop will address the latest advances in boiler water/feedwater chemistry. Presentations include: n Understanding AVT-O (oxidation) and AVT-R (reduction). n Pros and cons of feeding amines in HRSGs.
40

n New guidelines on phosphate treatment programs. n Caustic treatment: An overlooked program option. n Online particle counters for boiler water control.

Day Two will focus on best practices for outage maintenance, with presentations on: n HRSG cold-end corrosion: Causes and cures. n User experience with gas-side tube cleaning. n SCR life cycle: Minimizing fouling, degradation testing, and refurbishing/replacing. n Planning and executing an FAC survey. n User experience with FAC monitoring. The HRSG Users Group is one of the fewperhaps the onlyuser organization serving the combined-cycle community that is approved by a state board of professional engineers as a Certified Education Provider. Steam Plant Workshop attendees will receive two continuing education units. For more details on this, and all other HRSG Users Group activities, visit www.hrsgusers.org or email or call Executive Director Rob Swanekamp (info@ hrsgusers.org, 406-582-8655).
COMBINED CYCLE JOURNAL, Second Quarter 2006

PACIFIC DIAZO PRODUCTS

Your Source for Ammonia
930 S. Mt. Vernon Ave., Ste 900 Colton, CA 92324 (800) AMMONIA (266-6642) Fax: (909) 357-0264

Safety First

24-Hour Emergency Service Radio-Dispatched Drivers Deliveries from 100 to 6500 gallons Pump-Out Services Removal of Ammonia Temporary Storage Safety Training

gone to a caustic treatment, instead of phosphate, in the drumsin fact I was in some plants last week that did thisand thats been successful as well. So there are many variations going on, but I do see a trend in the industry. Independent consultant: I disagree completely. There is a large trend, and its a continuing trend over the last decade, of getting cleaner and cleaner in the evaporator sections until you cant get any cleaner because something stops youwhether thats air in-leakage or condenser contamination from the circulating water. If youre on a cooling tower and chlorides and sulfates start butting up against your controls, then you have to start putting some sort of buffering chemistry in there. In the entire loop, both high and intermediate pressure (HP, IP), AVT is becoming a very viable process. Im concerned about the recommendation to summarily removing oxygen scavenger from non-ferrous systems. Just like some others here, I came back from a conference all hot-to-trot about knocking out the scavenger/reducing agent, in a unit with a copper condenser, and it ate our lunch in a heartbeat. EPRI says thats not supposed to happen.

EPRI says you can have 90/10 copper/nickel-tubed condensers and still get away with it. But for us, that was not the case. So I would be interested in speaking with people who are making these AVT-O swaps, getting rid of the reducing agent. Gosh, were carrying 9.4, 9.5, 9.6 pH levels, which is 10 times the ammonia we used to have when we were on an AVT-R program with a copper condenser. Im curious if anybody is having any success doing that, because it really addresses the question: Why would you not go to AVT-O? Thats two parts of ammonia and 25 or 50 ppb of dissolved oxygen, and thats deadly. Chemical vendor: I think the devil is in the details. Reading the EPRI guidelines carefully, they say that if there is a copper condenser then lets move the pH up very slowly. The new pH range for copper, for mixed-metal systems, is 9.0 to 9.3. They suggest that if youre adopting AVT-O, try going to 9.4 to 9.6, but monitor the copper levels in the condensate both before and after the change, and see what your system will tolerate. Copper pickup from the condenser is also going to be tied to the air or oxygen level, so its not nec-

essarily a blind mans step to a high pH. Its a careful evaluation of what your system can tolerate. Chemical vendor: I agree with my colleague on that. I think that the EPRI representatives comment in other sessions about completely eliminating the reducing agent was directly related to the fact that most of these HRSG systems are all-steel. They do not have any copper alloys in them. So AVT-O is obviously a good option to consider. I want to make a comment on the one fellows statement that hes thinking of switching because of phosphate hideout problems. That is a big driver for this switch. But dont forget that you also have in your arsenal the ability to go to caustic treatment, or to what I call the lowlevel alkaline-phosphate program where you carry the two to four parts of phosphate, but you carry some free hydroxide alkalinity. You try not to run equilibriumphosphate chemistry, because one of the big drivers of hideout in these HRSGs is the dependence on duct burners; it can be the high energy input from them that affects the solubility of the phosphate if youre running at too high a level. We also know from this forum that one of the major
41

COMBINED CYCLE JOURNAL, Second Quarter 2006

HRSG USERS GROUP

New technologies to facilitate HRSG tube repair

he most common location for leaks to occur in HRSGs is at tube-to-header welds. Sometimes, the leak will be in the front or rear tube row of a module, along a lower header, and quite accessible for repair. More often, however, the leak is located in a tube row in the interior of a module, and is therefore Gandy extremely difficult to access. To reach the leaker, many undamaged tubes must be removed and replaceda process often described as cutting your way in and welding your way out. Not only is this time-consuming and expensive, but it often leads to future failures of the tubes that had to be disturbed for access. Another repair method sometimes used on inaccessible tube-to-header welds is to cut a window in the back side of the header, on each end of the failed tube, and then weld a plug into the tube openinga process of abandoning the tube in place. Most HRSG coils can handle a few abandoned tubes without serious thermal performance impact, but there is a limit to how many times this method can be applied.

Tube-to-header leaks
In his presentation at the 2006 HRSG Users Group meeting, David Gandy, senior project manager for materials and repair, EPRI, Palo Alto, Calif failure mechanisms in HRSGs has been over-feed of phosphate, because of hideout, that causes acid-phosphate attack. This is a critical issue to consider if you want to move into AVT-O. If you have an all-steel system, theres no problem. If you have a system with a copper-alloy condenser, then you must consider the earlier comments by my colleague. Chairman Anderson: Are there any users here who are using AVT in their evaporators? If so, can you share with us why you are, and how that program is working out for you?
42

the back side of the header. The proprietary process and specialized tooling was developed for EPRI by Encompass Machines Inc, Rock Hill, SC. Like tube plugging, EDM discussed above, this technique assembly avoids the need to cut and replace many otherwise good tubes. It offers the added advantages that it leaves the repaired tube in-servicehence no reduction in HRSG thermal performanceand it requires only a Base plate window weld in the header adjacent to the leaknot in the headers on both ends. According to Gandy, the new equipment provides precise control of tube and tooling alignment via a precision base-plate strapped to the header, and mandrels to provide close tolerance weld fit-up and high quality full-penetration welds. The result, Gandy reported, is considered a permanent repair. Heres the process: The header A. One new tube-repair method window and the internal header intended for leaks at tube-to-header weld prep are precisely and rapidly jointsgains access to the damaged cut using specially shaped carbon area via a precisely machined window electrodes, and electrical discharge cut into the back side of the header. machining (EDM) technology (Fig A). Applicable to header diameters rangNext, a precisely measured length ing from 3 to 12 in., the technique of the original tubing (containing should reduce the need to plug leaking the crack) along with the original tubes or disturb nearby good tubes tube-to-header weld is removed by remote machining through the ([email protected]), described header window. A custom-machined new, specialized tooling that can replacement oversize tapered stub improve on these traditional meth- is then pressed into the header, and ods. The new tooling allows removal remotely welded on each end, to the and replacement of the leaking weld tube and header respectively (Fig B). using field machining tools and a All tube and stub welding is automatremote welding device. The new tools ed, and performed from the inside. gain access to the damaged area via Finally, the header window is manua precisely machined window cut into ally welded back into place, and any User: Yes, weve switched to AVT and only using ammonia. One of the big reasons we could make the switch is because were air-cooled. We dont have a cooling tower, so we avoid the issues that come with it. Its been three years since we made the switch. We find it easier and more cost-effective to control pH. We no longer use any other chemicals regularly; however, occasionally well use caustic for additional pH support on the drumsif we start getting high on the condensate. Copper transport is not an issue for us. And we have been seeing steadily decreasing iron. Were now six years old, and originally we had a lot of iron transport when the air-cooled condenser was breaking in. I dont know how much of our decrease in iron can be attributed to the switch to AVT, and how much is attributed to the air-cooled condenser finally getting passivated. Regarding dissolved oxygen in the feedwater: Were real close to 10. Weve got the deaerator working quite well. Ive heard a lot of people talking about demin makeup. Sometimes

COMBINED CYCLE JOURNAL, Second Quarter 2006

HRSG USERS GROUP

necessary post-weld heat treatment is performed. The supplier provides tooling and trained technicians as a contract service for EPRI members and nonmembers alike.

Leaks along the tube length

When water chemistry is not managed well, under-deposit corrosion can cause tube leaks along a tubes lengthrather than at the more common tube-to-header weld. If located some distance from a header, the leaks cannot be repaired using the tooling described above, Gandy pointed out. In these cases, traditional tube plugging or cutting in/welding out probably are the only options. But Gandy described to the group another recent innovation that might be useful for some of these leaks along the tube length. Like the technology described in the first half of his presentation, this technique also was developed for EPRI, this time by Carolina Energy Solutions, Rock Hill, SC. It requires removing a precisely measured length of original tubing via conventional non-flame cutting methods. The tube ends are then prepped and a prefabricated replacement section is installed. Why is that innovative? Because the welding is performed remotely from the inside of the tube, typically in only one pass per end. This is accomplished using new remote welding tooling through a precisely machined oval window in the replacement tubes wall. Once the tube ends are welded and inspected (both inside and out), a precision-machined section of tube material is manually welded into the window opening. This is considered a permanent repair, Gandy reported, since the carefully controlled geometry of the window produces no large stress we have some tube leaks and the condensate dissolved oxygen will be high. Weve got a pretty tight HRSG, I think right now were down to the 20s or so. Ammonia level at condensate discharge: I dont have any idea right now. Chemistry savings: Obviously weve had some, because the chemicals are a lot cheaper. And its easier for the operators to test the chemistry. I dont know if weve had more tube leaks, or fewer, directly related to the AVT switch. Regarding the oxide layer: Its kind of red in the drums, and its black in the air-cooled

B. A precisely measured length of the original tubing (containing the crack) along with the original tube-to-header weld is removed by remote machining through the header window (left), after which a replacement oversized stub is pressed into the header, and remotely welded on each end

Window

C. Leaks along the tube length might benefit from another new repair technique, which removes a precisely measured, ovalshaped length of original tubing via conventional non-flame cutting methods (left). The tube ends are then prepped for welding and a prefabricated replacement section is installedall perWindow formed remotely and from the inside diameter of the tube (right)
Tube

concentrationsas are often found in field-fabricated window welds. According to the supplier, use of this technique is faster than conventional external tube welding, and it provides welds of superior quality. It has the minor disadvantage, Gandy conceded, that tube sections of appropriate dimensions and materials must be prefabricated with the oval window and window closure plug. Although the technique was developed for use in fossil boilers, Gandy expects it to find use in HRSGs condenser and LP turbine exhaust. Chairman Anderson: Thanks. Its very helpful to hear directly from users with their experience on this, particularly since youre out there at the leading edge of the technology. Those interested in learning more about transitioning away from phosphate treatment to other chemistry programs, as well as what the latest guidelines are for phosphate programs, should consider attending the HRSG UGs upcoming Steam Plant Workshop, where a full day will be dedicated to presentations

where access for conventional welding to the back and sides of tubing is limited by the close spacing of adjacent tubes. It also may be put to good use in quickly and reliably removing tube samples for determination of when to chemically clean an HRSG. Premachined tube sections, the remote welding equipment, or turnkey field service using this technique is available from Aggressive Equipment, Lawrenceville, Ga. EPRI membership is not required to purchase or contract its use. and discussion of water chemistry (see sidebar, p 40).

Confounding controls questions

Control systems was the third of 10 Open Forum Discussions in Colorado Springs. Several users reported difficulty in controlling superheater and reheater outlet temperatures while at low loadsparticularly with HRSGs operating behind a 7FA gas turbine (GE Energy, Atlanta). Continues on page 46
43

COMBINED CYCLE JOURNAL, Second Quarter 2006

HRSG USERS GROUP

Risks to HRSGs in low-load operation

any of todays combinedcycle plants are compelled by market demands to shut down and restart frequentlywhether the plant was designed to do so or not. These repeated start/stop cycles, as readers of the COMBINED CYCLE Journal and attendees of HRSG Users Group conferences have come to understand, imposes substantial wear-and-tear on heatrecovery steam generators. But market demands also are forcing many plants to spend long periods operating at low loads. At the 2006 conference, Scott Wambeke, systems engineer for HRST Inc, Eden Prairie, Minn, (swambeke@ hrstinc.com) discussed the less well-known risks of this low-load operating mode. For purposes of his presentation, Wambeke defined lowload as being below 80% of base load, although some gas turbines (GTs) permit operation as low as 30% of the nameplate rating. There is a long list of potential steamplant problems caused by low-load operations, not all of which are

Gas flow

Gas flow

A. Tubes are restrained between rigid top and bottom headers, in a horizontal HRSG. Differential thermal expansion can then produce the stress necessary for fatigue
44

into those tubes directly opposite the panels inlet nozzle, while dry steam makes the turn and enters the other tubes. The tubes receiving the wet steam are quenched significantly and therefore experience significant tensile stresses, compared to the other tubes. The tensile intuitive. While Wambeke stresses can be so significant focused on the first four that these tubes often are of these, he listed the found elongated and warped many potential low-load out of position during visual problems as: inspections of HRSG intern Economizer and prenals. In severe cases, the heater fatigue. tubes can be cracked and/or n Desuperheater overpulled away from the header. spray. What does low-load operan Failure of non-pressure tion have to do with desupartssuch as perfoperheater overspray? As GT Wambeke rated plates, turning load is decreased, Wambeke vanes, liners, and baffles. explained, steam flow through the n Operation below the design basesuperheaters and reheaters also ment pressure (also called floor decreases. Depending on the exhaust pressure). profile from the GT, this may shift n Insufficient catalyst temperatures. heat absorption in the HRSG to favor n Elevated cold reheat temperathe superheaters and reheaters. At tures. the same time, reduced steam velocn Control valve throttling. ities around the desuperheater may n Economizer steaming. result in a less-than-optimal mixing n Off-design duct firing. environment for the steam and water. n Vibration. Making matters worse, some GTs n Water chemistry. notably the 7FA (GE Energy, AtlanE c o n o m i z e r a n d p re h e a t e r ta)actually produce higher exhaustfatigue can occur during low-load gas temperatures at low load than at operation if feedwater flow instability base load. The result is that very high exists. Flow instability causes some desuperheater spray-water flow is tubes to fluctuate in temperature needed (to prevent exceeding design relative to their neighboring tubes. steam temperatures) at a time when The resulting differential thermal spray-nozzle performance is probexpansion between these tubes ably degraded. which in many horizontal HRSGs are How can you tell if your HRSG is restrained between rigid top and bot- experiencing desuperheater overtom headersproduces the stress spray? Wambeke pointed out that necessary for fatigue (Fig A). one obvious indication is superAll types of economizer/preheater heater or reheater tube failures near panels (multiple row, single row, and the header inlet nozzles. Warped return bend) are susceptible to this tubes can be another indication, condition. Wambekes case in point: but thorough investigation may be A feedwater preheater that was suf- necessary to eliminate other posfering repeated tube leaks at spe- sible sources of tube warping. If the cific tube-to-header welds. A study steam temperature downstream of performed by HRST indicated that your desuperheater has less than 35 low water velocities were occurring deg F residual superheat, you probwhen the unit was operating at low ably have high risk of an overspray load, resulting in flow instability (Fig condition. Likewise, if the desuperB), hence differential thermal expan- heater is installed near upstream or sion. The solution? HRST worked downstream pipe bends, you probwith the client to modify the internal ably have poorly mixed and poorly header flow baffles in such a way vaporized spray and overspray-like as to increase flow in the affected symptoms (Fig C). tubes. What to do? Depending on the Desuperheater overspray occurs situation, here are a few possible when more cooling water is sprayed options that Wambeke presented: into the steam flow than can be n Modify steam piping around the completely vaporized. When the desuperheater. resulting steam/water mixture enters n Install a different style desuperthe superheater panel, the heavier heater less prone to performance water droplets take a straight path loss during low steam flow periods.
COMBINED CYCLE JOURNAL, Second Quarter 2006

HRSG USERS GROUP

Tubes below inlet nozzles are quenched and experience a large tensile load Thermocouple reads at or near saturation temperature Steam from desuperheater Upper header

Feedwater outlet

Feedwater inlet

Lower header Stagnant or reverse-flow circuits Steam outlet Quenched and stretched tubes Superheater tubes at uniform high temperature

B. Operating at low load can cause flow instability in econ- buckle as they equalize temperatures with neighboring tubes omizers and preheaters, which leads to differential thermal expansion and potential fatigue damage in HRSG tubes

C. Desuperheater overspray is another potential problem induced by low-load operation. As the load is decreased, steam flow through the superheaters and reheaters also decreases, causing superheater panels to over-perform and reducing the turbulence needed for optimum desuperheater spray-nozzle performance (top right)
Gas temperature to SCR tracks saturation temperature in tubes and steam drum. At 1200 psig, saturation temperature is 569F; at 800 psig, 520F. SCR controls may stop ammonia flow when gas temperature is between 500F and 550F Feedwater inlet Deaerator

SCR Gas turbine exhaust gas Duct burner Superheater LP evaporator HP evaporator HP economizer

D. Non-pressure parts also can be damaged by low-load operation. These perforated plates became warped from overheating, caused by changes in exhaust-gas flows and temperatures
n Modify superheaters to reduce

E. Operating below the basementor floorpressure can cause several problems, among them improper functioning of the SCR catalyst supports can solve these problems. In other cases, it may be necessary to replace components with improved designs or better materials. Problems with duct- burner and catalyst performance may require physical or CFD (computational fluid dynamics) modeling of gas-flow profiles before effective modifications can be made.

heat pickup. n Install a steam bypass around the superheater. n Investigate if higher superheater outlet steam temperature is permissible without causing other problems to piping, steam turbine, etc. n Reduce GT exhaust temperature.

Look at non-pressure parts, too

Pressure parts are not the only components that can take a beating

during low-load operation, as Wambekes presentation made clear. Reduction in exhaust-gas flow and changes in exhaust-gas temperature patterns occur, when moving from base- to low-load operation. This can cause non-pressure parts in the exhaust-gas streamsuch as perforated plates, turning vanes, liner systems, and tube supportsto overheat or experience unanticipated vibrations (Fig D). It also can degrade duct-burner and catalyst performance. Sometimes simple fixes such as stiffening components and adding

How low can you go?

Many of todays combined-cycle plants operate in sliding pressure mode. This means that as GT load is reduced, the steam turbines inlet
45

COMBINED CYCLE JOURNAL, Second Quarter 2006

HRSG USERS GROUP

valves are left wide open (for maximum thermal efficiency), resulting in lower HRSG steam pressures. As steam pressure falls, water and steam densities decrease and their velocities increase. Eventually, as Wambeke explained, a point is reached where further increase in steam/water velocities will cause carryover of water into the superheater or water-side erosion of evaporator tubes and risers. Realize too that when evaporator pressure drops, so does evaporator temperatureas the Mollier Diagram illustrates for saturated steam conditions. Because the evaporator temperature sets the exhaust-gas temperature entering the catalyst, and because the catalyst has a minimum operating temperature, there is a minimum evaporator pressure below which the catalyst will not function properly (Fig E). When the first of these lower pressure limits is reached, the boiler is said to be at its basement or floor pressure. Unfortunately, significant operating time at low load often was not anticipated by the OEM, Wambeke said, and many HRSGs were designed with relatively high floorpressure limits. Some possible solutions if too high a floor pressure limits your ability to operate at low load? Increas-

Continued from page 43 The issue stems from thermodynamic fundamentals. While at high loads, steam pressure and mass flow are high, and GT exhaust temperature is conducive to desired steam conditions. At low loads, steam pressure is lower (at least for units operating on sliding-pressure control) as is steam mass flow. However, GT exhaust temperature actually is much higher. This apparent paradox is a design characteristic of advanced gas turbinesparticularly evident on the 7FAthat is required to maintain flame stability in the combustor. Unfortunately, many users report that at low loads they are not able to cool the steam sufficiently using their interstage attemperators without creating an overspray condition where some spray water fails to vaporize. Overspray conditions create very severe thermal stresses and fatigue damage in piping, headers, and tubing. A lengthy discussion ensued about turndown ratios, valve trim, soft-

An experienced OEM ware options, ASME Boiler engineer astutely pointed & Pressure Vessel Code out, This is really not a requirements, etc. It was control problem. This is an clear that this controls probintegration problem, of the lem is serious, and widegas turbine, the HRSG, and spread. One user summed the control elements. Its up the situation quite well: very complicated. There I have to say that, in my are many different poscompany, weve become reluctant experts on this sible solutions, and there topic. Weve had steam lines are many different possible Lant crack and fail because of problems. thermal quenching. Collaborative efforts. As menWhen you run in a regulated, dis- tioned above, several members of patched-to-the-grid environment, the team responsible for designing energy control centers love the flexi- combined-cycle plants offered by GE bility of a combined cycles ramp rate. participated in the meeting. Clearly, So we have some units that load-fol- these engineers came to work. One low pretty dramatically, and spend example: During the discussions on a lot of time running at low loads controls, a prominent GE designer because theyre serving the grid and raised the following question. Faster the spinning reserve needs. [With starting of HRSGs is the big issue these units,] weve had problems on we see for new plants. Specifically, the reheater side, and its thermal- easing or eliminating the limitaquench-driven from having the high tion on GT loading. Uncouple the spray flows. Were working on trying GT and steam cycles with respect to to design this out when we build a startup. Low-load GT holds increase new plant. the startup emissions, which are now controlled in some areassuch as southern California. And of course we need to do this five times a week throughout the year and still achieve a minimum 30-year life. We would like to think out of the box to find ways to address this issue. Talk about a spark ignition! This question set off a benchmark exchange of ideas between users and some of the real decision-makers at GE. If in future plants you see GE implementing some of the following ideas, you will know the brain trust present at the Broadmoor and the healthy exchange of technical information had something to do with it: n Use of terminal attemperation to thermally uncouple the steam turbine startup from the GT loading rate. 4. Chairman Bob Anderson (right) and Bob Gunnerson of Xcel Energy disn Use auxiliary boiler to pull vacucuss the O&M challenges of advanced combined-cycle plants. Most of the um before GT startup. conference agenda is comprised of Open Forum Discussions, in which users n Use auxiliary boiler to warm the present technical problems they currently face at their facilities; other attendHRSG via steam sparging before ees are asked to share their experiences and debate solutions GT startup.
COMBINED CYCLE JOURNAL, Second Quarter 2006

46

HRSG USERS GROUP

ing the catalyst temperature without increasing the evaporator pressure is difficult. Throttling the steam-turbine inlet valves can raise HRSG evaporator pressures at the cost of some ST efficiency. Consider checking with the HRSG OEM or a qualified consultant for evaluation of your units evaporator circulation and possible modifications to mitigate the problem. Wambeke also pointed out that operating in sliding-pressure mode and at low load can exacerbate flow-accelerated corrosion (FAC), because FAC tends to increase at temperatures around 300F. If this pertains to your plant, you should have a program in place to monitor the highest risk areas. (Note that all HRSG users should have an FAC monitoring program covering their entire steam plant, even if they never operate at low load.) Another component that takes a beating at low operating pressure is the feedwater control valves when used with fixed-speed boiler feedwater pumps. In this situation, the control valves experience enormous pressure drop across their control surfaces for extended periods of time. Wambeke suggested that a second low-flow feedwater control valve may need to be installed, to mitigate this problem.

n Use electric heating pads to warm

the HP drum before GT startup. n Full automation of total plant startup. n Careful integration of all plant equipment and systems. n Better access for pressure-part inspection and maintenance. n Keep the difference in size between tubes and headers as small as possible. n Allow sufficient time in the project schedule for the HRSG designer to do a proper job. n Specify a once-through, Bensontype boiler rather than a drumtype unit.

Taking it home
During the two days of Open Forum Discussions, many more in-depth and valuable deliberations transpired, covering the seven remaining categories: piping systems; ductwork, dampers, and stacks; valves; supplementary firing; environmental systems; and balance of plant (BOP). Many questions were asked, and many solutions identified. Still other questions were raised that require either more time than a conference allows, or more study and analysis by both the user and the supplier communities. One thing seemed clear: The hundreds of users and suppliers who come together each year at the HRSG UG conference to share their knowledge, experience, problems, and solutions will continue to drive the industry forward. As Chairman Anderson (Fig 4) reminded the crowd at the events conclusion, This format works because of the people that attend and openly contribute to the discussions. No matter what your level of experience, we hope you leave each conference with more than you brought.

Pre-conf seminar debut

This year the HRSG UG added something new to its annual conference: a

one-day technical seminar held the day prior to the events official start. Damage Mechanisms in CombinedCycle Plants was attended by 80 power professionals and conducted by European Technology Development (ETD). The Surrey, UK-based firm is an engineering advisory, consulting, and R&D company specializing in plant life extension, maintenance, materials, and engineering issues in all types of power generating and process plants. In addition to conducting the preconference seminar in Colorado, ETD has organized numerous workshops, training courses, and conferences in Europe and Asia. The seminar was led by Tony Lant, ETDs plant services manager, who covered such damage mechanisms as: n Creep. n Fatigue. n Thermal fatigue. n Acid dewpoint corrosion (Fig 5). n Flow-accelerated corrosion. n Stress corrosion cracking. n Corrosion fatigue. Lant also discussed drivers in HRSG design, thermodynamic considerations in design, HRSG tubing and panels, welding in HRSG fabrication, and issues for cycling operation. Following are a few excerpts from the seminar: 1. Ever wonder why HRSGs are so much larger than fossil-fired boilers of similar capacity? Seminar attendees now know. As Lant explained: Temperature differencesbetween furnace and flue gas, and between water and steamare much lower in HRSGs. This makes heat-transfer rates in HRSGs low, relative to conventional boilers. Low temperature differences in HRSGsbetween the exhaust gas and the steam/water mixtures require the use of finned tubes, and lots of them. Not all HRSGs use all of the features for maximum heattransfer efficiency. Tradeoffs are often necessary. For example, tight

5. Dewpoint corrosion, clearly evident on these finned tubes, is one of the HRSG damage mechanisms that were explored in detail during a preconference seminar conducted by European Technology Development Ltd
CCGT Fossil

6. Substantially higher stack losses are one of the many characteristics that distinguish HRSGs from conventional fossil boilers. Such differences result in unique operating and maintenance challenges, which both users and suppliers are working to address tube pitch with a staggered arrangement may not be appropriate for the cold end of HRSGs that burn oil or are equipped with an SCR. Gas-turbine exhaust contains 100 to 200% excess air. This is because of the need to air-cool hot parts in the combustion system and hot gas path, to avoid overheating of GT components. This high excess-air content of the exhaust gas results in stack losses that are two to three times higher than in a conventional boiler. In addition, flue-gas flow through an Continues on page 50
47

COMBINED CYCLE JOURNAL, Second Quarter 2006

HRSG USERS GROUP

Safety/relief valves: Installation, maintenance, testing

ike all high-pressure boilers, HRSGs are required by the ASME Boiler & Pressure Vessel Code (Code) to employ a variety of safety valves (covered in Section I) and relief valves (Section VIII) to prevent overpressure operation and eliminate the risk of explosion. In fact, because of their multiple pressure levels and greater complexity, HRSGs typically will have many more safety/relief valves than conventional fossil boilers. Note: While the terms often are thought of as interchangeable, a safety valve is a pressure-relief device characterized by rapid and full opening (or pop action) at its setpoint. By contrast, a relief valve is a device that opens in proportion to the increase in pressure that exceeds its setpoint. Both types of valves have a long track record of reliable performance and often are considered simple devices. But they still require careful attention to installation details, as well as recurring maintenance and testing, if safe, reliable performance is to be attained (Fig A). In his presentation to the 2006 HRSG Users

Group conference, Robert Pabst, a valve design/maintenance consultant for Movaco Inc, Bradenton, Fla ([email protected]), provided a useful summary of those details and reminded users of the ongoing valve work they need to accomplish in their maintenance programs.

Start with drains, vents

Pabst began with a review of the vent and drain subsystems so important to pressure-relief valve operation. For example, rainwater and condensation routinely collect in the valve body, discharge elbow, drip pan, vent piping/stack, and silencer. If the small-bore, carbonsteel drain piping that removes this water is not installed correctly or is allowed to plug with rust Pabst and debris, you can count on expensive, and possibly dangerous, system damage (Fig B). Water left standing in vent piping or silencers also will, if the valve lifts, immediately flash to steam. This, in

A. Maintenance and testing of safety/relief valves should be essential elements of any combined-cycle or cogeneration plants O&M program. These pressure-relief devices protect the vital investment of the facility
48

turn, can cause backpressure fluctuations conducive to shortened blowdown cycles, irregular valve behavior, water hammer, and internal damage to the valve, piping, and silencers. Pabst said, Incidents of shrapnel being ejected from the silencers are not uncommon. Even if the valve is never forced to lift, standing water in the valve body is certain to cause corrosion damage to valve internals, altering the valves operationif not entirely preventing it. Another potential problem arises if the drains from the pressure-relief system are routed to locations that restrict flow or otherwise cause the drains to pressurizesuch as improper tie-ins to other piping systems. Improper tie-ins to pressurized systems have prevented safety valves from lifting at their setpoint, Pabst reported, resulting in catastrophic plant damage. Heres another all-too-common installation mistake that Pabst has seen: Many safety valves have a small-bore vent connection on the body cap. This vent is supposed to be piped to a safe location where escaping steam during valve operation will not cause injury to personnel, but where it also will not restrict vent flow or otherwise create backpressure. Pabst has sometimes seen this vent incorrectly piped into the valves drain system. When this occurs, undesirable backpressure and improper valve operation is sure to follow. Pabst recommended that plant personnel include drain-line inspections as part of each of their regulatory-mandated valve tests and repair tasks. Check to ensure that no alternative source of pressure can be introduced into the drain system, and that vent cap ports are vented to atmosphere using very short runs of pipejust enough, he said, for personnel protection as the short burst of steam vents during a lift. The manufacturers manual is an excellent reference for both drain-line installation and inspection. Where drain pipe is hard-plumbed into the valve, Pabst urged users to install a pipe union at the first avail-

COMBINED CYCLE JOURNAL, Second Quarter 2006

HRSG USERS GROUP

user in the audience brought up just such an example. Only a month after initial commissioning, both of his hot reheat safety valves started leaking. One was chattering noticeably, before the spring and stem broke. Pabst suspected that these valves were incorrectly specified because theyre operating too close to their set pressure. Go back and look at your system pressures at the time the chattering happened, he responded, and see if your safety valve setpoint is within 20% of that margin. If it is, you need to think about going to a different type of valve, or raising the setpoint. You can do that within the Code, but you have to check with the boiler manufacturer and go through some hoops. A final installation tip that Pabst covered deals with initial quality control. This year the National Board of Boiler & Pressure Vessel Inspectors, Columbus, Ohio, will publish some new rules concerning QC, he reported. At issue is that some Section I safety valves are designed with capacities greater than that of the manufacturers boiler capabilities for flow testing and cannot be tested for capacity blowdown. Set pressures are checked at the manufacturer. Manufacturers normally cannot ship a new valve without flow testing and still comply with the Code, but the National Board has given suppliers allowance for this constraint. A typical allowance is a sign-off check sheet from the manufacturer notifying the end user that capacity blowdown cannot be tested and that an alternative means of setting the blowdown will be done. Pabst made HRSG users aware that if they want to test for full blowdown, they must do so after installation at the plant site.

B. A reliable pressure-relief system also requires adequate draining of water that collects from each safety/relief valves vent stack and drip pan. Drain lines, typically constructed of carbon steel, can corrode internally and clog

C. Pilot-operated safety valves (left) used in high-temperature applications Valve maintenance, should incorporate a heat exchanger on the pilot sensing line to protect vital Orings and seals inside the valve. Bellows-type safety valves should be selected testing when backpressures are known to exist that could shorten the blowdown Developing and executing a safety/ relief valve maintenance program compliant with all codes and regulaable opportunity, to facilitate future from the initial design of combined- tions is important for all powerplants. inspections. cycle/cogen plants. Pabst offered the following suggesWhen backpressures are known tions for such a program: to exist that could shorten the blowDetermine the code that applies to Valve installation down, Pabst advised that a bellows- each of your valves. A V stamp on Pabst also had a lot to say about type safety valve should be used (Fig the ASME nameplate indicates Secthe selection and installation of the C). A typical application in a com- tion I applies. A UV stamp on the valves themselves. For example, bined-cycle plant is on the cold reheat plate indicates Section VIII applies. when pilot-operated safety valves are steam system, if the valves discharge Define maintenance intervals for used in high-temperature applica- into a single or combined silencer. each valve based on the most restrictions, a heat exchanger on the pilot He also said that if any of your tive requirements among the National sensing line should be incorporated pressure-relief valves require fre- Board, local jurisdictions, and your to protect vital o-rings and seals quent, repetitive maintenance, then insurance carrier. inside the valve. Such heat exchang- your existing valve probably is not Typically, an annual pressure set ers, Pabst reported, often are omitted suitable for the given application. A test is required for all Section I safety
COMBINED CYCLE JOURNAL, Second Quarter 2006 49

HRSG USERS GROUP

valves. This should be done just prior to your annual outage, Pabst said, so that if any valve that fails to reseat during the test can be repaired during the upcoming outage. Typically, pressure set testing for Section VIII relief valves is required on a three-year interval. This requires that some valves be removed from the system and tested at a qualified test and repair facility. Smaller relief valves should simply be replaced, Pabst noted, because the cost of testing them might exceed their replacement cost. Typically, Section I safety valves must undergo teardown and internal inspection every six years. This must be performed by a technician from a VR stamp-holding repair facility. During such teardown and inspection, certain components must have critical dimensions checked and recorded, and any found to be outof-tolerance must be either replaced or machined. Which components require this scrutiny is determined by the manufacturer for each model of safety valve. After completing inspection and reassembly, the safety valve also must be set-pressure tested either at the qualified shop or on the owners equipment. If set pressure testing is performed using plant steam pressure, plant load must be held steady between 80 and 90% of the full-load rating. Sometimes this is difficult to accomplish on intermediate- and low-pressure (IP and LP) systems in multi-pressure HRSGs. Continued from page 47 HRSG is about 10 times greater than for a conventional plantall adding up to increased stack losses (Fig 6). 2. Ever wonder why longitudinal pipe welds in steam piping and headers are of greater concern than circumferential welds? Seminar attendees now also know this. Lant explained: Design stress equals hoop stress equals twice the axial stress. Hoop stress is the stress trying to expand the pipe because of the internal pressure. Only axial stress operates over a circumferential weld, hence it experiences approximately half the stress that the longitudinal seam weld sees. Exposure to higher stress while at high temperature causes creep damage to occur faster, particularly in the weaker heat-affected zone (HAZ) of a weldment. Therefore, premature creep failure in seam welds is more likely than in girth welds, all other things being
50

Therefore, Pabst suggested and has worked well for conthat electronic valve testventional boilers, the wider ing equipmentsuch as range of operating pressures that supplied by AVK Indusof todays more complex tries Inc, Jacksonvillebe HRSGs can tax the simple used for on-boiler valve setaction of the direct springpoint testing. A certified test operated safety valve. report is required for each Section VIII of the Code, valve. A warning from Pabst: w h i c h a d d re s s e s r e l i e f When performing on-boiler valves, allows the use of setpoint testing, an accurate pilot-operated valves and Gertz reading of the steam presdirect spring-operated with sure at the valve is required. Normal power actuator safety relief valves, DCS pressure read outs, even when in addition to the traditional direct recently calibrated, may not give spring-operated valve. sufficiently accurate pressure at the Gertz and other valve professionsafety valves location. Best practice als are working within the ASME to involves installing a pressure tap near modify the Code to allow the use the safety valve for temporary use dur- of alternative safety valve designs ing testing. on Section I applications. This has already been accomplished to some extent, Gertz reported, via Code Case New designs on the 2446 which permits the use of pilothorizon operated safety relief valves on SecA s a n e x c e l l e n t s u p p l e m e n t tion I economizers. Benefits of this to Pabsts presentation, Jorgen change, according to Gertz, include: Gertz, now marketing manager for n Improved seat tightness. North American Power Products at n System can operate closer to set Valvtechnologies Inc (jgertz@valv. pressure. com), presented information on very n Stable operation in multi-phase recent developments that could flow. change the type of safety valves n Ability to modulate. allowed for use on Section I boilers. n Can handle higher backpressure. The basic design of the direct n In-situ setpoint verification withspring-operated safety valve, Gertz out lifting main valve. pointed out, has been around almost Gertz anticipates that the next as long as boilers themselves. Cur- change will be expansion of the rently, Section I of the Code permits scope of Code Case 2446 to peronly this type of safety valve. While mit the use of pilot-operated safety this has been the case for many years valves for all Section I boilers. equal. Of course, all other things are not always equal. Large unplanned stresses can be placed on girth welds because malfunctions in piping support systems. Thats why pipe hangers and supports need to be checked often, and problems fixed promptly. 3. Ever wonder what causes most HRSG tube failures? Lant explained: n One of various internal corrosion mechanisms in water-wetted sections. n Or one of various mechanisms associated with thermally induced stresses at tube-to-header attachments. These include corrosion fatigue in preheaters, economizers, and evaporators. In superheaters and reheaters, they include thermal fatigue, creep fatigue, and ductile overload. ETDs advice for flexible operation and high HRSG durability: n Select appropriate design features and details for the HRSG that eliminate or minimize its susceptibility to tube and header temperature anomalies. n Adopt combined-cycle unit operating procedures for startups and shutdowns that develop transient thermal gradients and stresses in critical HRSG components compatible with the desired design cyclic life.

Support group for HRSG users

The 14th annual conference was supported by representatives of 74 suppliers of combined-cycle equipment and/or services who participated in more than just the events trade show. In contrast to many user events, which restrict suppliers to the expo hall only, all HRSG UG conference sessions are open to all industry participantsincluding manufacturers, EPC contractors, water-treatment suppliers, engineering consultants, insurance carriers, and so on. CCJ

COMBINED CYCLE JOURNAL, Second Quarter 2006

The Industry's New "Blue Book"

"There is no better single source of information on the topic of HRSGs! Typically a book of this nature is full of theory, with a few practical experiences interjected as examples. However, the HRSG Users Handbook is a compendium of experience-based knowledge, supported by theory. It has become an essential tool in our control room, and is a great reference during outage planning discussions at our site."

Terry Toland, Facility Manager

GE Contractual Services - River Road Generating Plant

"We found the HRSG Users Handbook to be very useful for our plant engineers. The book covers all aspects of O&M issues, and provides excellent guidance to the operating engineers. It's compact, easy to read, and easy to understand."

Lenin Vadlamudi, Planning Engineer

Improve Your New Project Designs Upgrade Your Existing Plant O&M Get Top Value in Plant Acquisitions
TransAlta Energy Corp

"This book is written with practical application in mind. The topics are definitive to daily operations and maintenance issues. There was sufficient interest from operators, maintenance, and management personnel, after reviewing the book, that we purchased multiple copies for each of our sites."

Charles Dameron, Resource Manager

Duke Energy - Combined Cycle Non-Regulated Fleet

"HRSG Users Handbook should be required reading for all who manage or operate a GT-based combined-cycle or cogeneration plant. It's a foundation upon which all users can build more reliable and efficient generating facilities."

Bob Schwieger, Editor and Publisher

Combined Cycle Journal

Others have tried to imitate it, but this wide-ranging, up-to-date, and affordable handbook could have been assembled only by the HRSG User's Group. We tapped our 14 years of educational events and 1400-plus members to create the book's 40 insightful chapters on steam-plant design, operation and maintenance. Contributing authors include HRSG users, OEM engineers, service contractors, steam-equipment designers, chemistry specialists, and licensed engineering consultants.
COMBINED CYCLE JOURNAL, Second Quarter 2006

Available exclusively through the HRSG User's Group! This 2006, 2nd Edition of HRSG Users Handbook comprises nearly three times more content than the 1st Edition, with far greater depth and clarity. Individual copies cost $345, with discounts offered on orders of five and up. For more details, as well as on-line ordering, visit www.HRSGusers.org or call Rob Swanekamp at 1-406-582-8655.
51