501G Users Group

Close collaboration between OEM, users facilitates issue resolution

Owner/operators of W501G (SGT6-6000) gas turbines manufactured by Siemens Energy meet face-to-face twice annually to share experiences with one another and with the OEM’s engineers. The G fleet totals 24 units at a dozen sites in the US a dozen sites in the US and one in Mexico.

What’s particularly interesting about this user organization is that its small size, capable leadership (Sidebar 2), and collaborative nature allow it to think and act as a unit, facilitating problem-solving at the deck-plates level and with the OEM. Interesting, too, is that while other gas-turbine user organizations attract perhaps one attendee for every two to three machines in their respective fleets, the G users get about one and a half attendees for every engine.

This is a proud group of engineers and technicians who have “grown up” together—so to speak—and understand each other’s perspective. The first G, installed by Lakeland Electric, began commi s s ioning operations in April 1999, but COD wasn’t until March 2001—only one month before the second machine began commercial  operation at Millennium. These people know each other, and each other’s plants, well.

Most user-group meetings host roughly one-third to onehalf first-timers, so many discussions are similar from year to year because newcomers have to be brought up to speed. There’s not much turnover in the top positions at G facilities which means each meeting pretty much picks up where the last one left off, especially regarding the OEM’s presentations. This certainly contributes to presentation efficiency because there’s a minimum amount of repetition.

To illustrate: At the 2010 annual meeting in Orlando last February, Siemens presentations covered R4 vane deflection, R1 ring segments, R4 blades, exhaust systems,  and rotor through bolt among the dozen or so issues of interest. Updates on each of those areas were featured at the abbreviated mid-year meeting at Siemens’ Orlando offices in September.

Approximately half of the presentation/discussion time at 501G meetings is earmarked for users only. At the annual meetings, where exclusive G content spans three and a half days, most plants in the fleet make at least one presentation. Last February, for example, there were several presentations on the commissioning of the Hillabee Energy Center, which had been mothballed before its purchase by Constellation Energy Inc.

Other owner/operators presented on low-load CO, retubing and repairs of rotor air coolers, impact of corrosive insulating oil on transformers, water hammer, outage experiences, fleet quality findings, dehumidifier performance, etc.

Two presentations from the 2010 annual meeting that stand out in terms of value for owner/operators of all large frames, regardless of the OEM, concerned heat-recovery steam generators. One, by Nooter/Eriksen Inc, Fenton, Mo, focused on HRSG design/operational issues and recommended modifications, the other by Millennium Power Partners LP, Charlton, Mass, concerned an LP economizer modification to reduce gas-path backpressure. More on HRSGs later.

User presentations typically are posted on the group’s website, which can be accessed through www.usersgroups.com by registered members. If you qualify for membership in the organization but have not yet joined, this also can be accomplished on the website. Siemens presentations are accessible by owner/operators of the OEM’s frames who apply for and are registered to access the manufacturer’s Customer Extranet Portal.

A joint effort

How the OEM and its G customers work together serves as an example for others. To better understand the synergy, the editors talked to Siemens Frame Owner Mark Carter and Mark Winne, plant manager at Millennium and the 501G User Group’s vice chair. Winne was a founding member of the user organization.

W501Gs have accumulated over 738,000 operating hours (as of October 2010) since their commercial debut in 2001 and they consistently receive the highest dispatch ratings of all gas-turbine models in Siemens’ North American fleet. However, the engine has had its share of issues—including compressor wear between the stator vanes and compressor casing. When  this issue surfaced,  Siemens worked with users to develop a cross-platform solution, which included the merging of Siemens and Westinghouse technologies in the field. Owner/operators closely followed the development and testing of the modification, enabling its timely acceptance by the user community and integration into the fleet.

Background. In mid 2003, the Siemens W501F fleet began receiving reports of wear on some compressor stator vanes; later, such wear was reported on some Gs. Observations of wear on the W501Gs may have been delayed because of attention to a different issue, prior to 2004, that required parts replacement on the first three rows of compressor stator vanes. Inspections in 2004 revealed diaphragm hook wear in rows 4, 5, and/or 6 on some G machines. Siemens initiated a program to define the  root  cause and  to develop an improved compressor.   The maintenance interval targeted for the compressor modifications, and the length of the outages required to implement the mod, made it imperative that Siemens and its customers work together.

During the program, members of the 501G Users Group steering committee, representing owners of 23 of the 24 engines in operation, were invited to Siemens’ Orlando engineering offices to review its findings in the matter and to verify that those findings were representative of what the users themselves were seeing on their units. Winne told the editors that the opportunity to sit down with the OEM and discuss openly and frankly both the problem and possible solutions was a “unique experience” for a user.Both Siemens and the steering committee worked together to investigate the various aspects of the root cause to arrive at a common understanding of the issue.

Siemens maintained close contact with customers as its engineers arrived at conclusions regarding root cause and as the company worked through the development phase of the modification effort. The OEM determined that certain features of its successful V-style compressor for the SGT5-2000 and SGT5-4000 gas turbines would address the wear observed on the W501Gs. Siemens once again invited the steering committee to Orlando to review its progress. After reviewing data, the users agreed with the OEM on the V-style compressor approach. Lead engineers on the compressor effort, Tom Gordon and Dave Wasdell, facilitated collaboration, enabling the setting of design targets and goals to satisfy a wide spectrum of stakeholders.


Installation of a V-style compressor in a gas turbine designed by Westinghouse was the first of several opportunities for merging Siemens and Westinghouse technologies in the W501G operating fleet. Siemens began procurement of hardware based on broad confidence in the proposed improvement by both the OEM’s engineering team and the users.  It’s important to note that collaboration of users and Siemens did not end with completion of solution development. The steering committee also was invited to Siemens’ Hamilton (Ontario, Canada) facility to witness the first parts being manufactured for validation testing. Next step was to identify the first validation site.

Winne realized the potential benefit  of moving ahead early with the redesigned compressor and the Millennium staff worked closely with Siemens to develop an 18-month program to install, test, and validate the design. Once the parties agreed in principle, a collaborative effort of detailed planning and logistical coordination was pursued to manage the significant potential risks associated with possible program and schedule upsets. Siemens invested in a full rotor (Fig 1) and conducted an extensive testing program—one involving 952 instruments on both stationary and rotating parts in rows 1 through 13. The rotor was fitted with a slip ring to route instrument readings on rotating parts to monitors outside the machine. Winne recalled that when testing began, the plant site looked more like a NASA mission control center (Fig 2) than a powerplant, with teams of engineers monitoring banks of computers and monitors arranged to gauge and verify the performance and integrity of each component in the new compressor.

The benefit of the comprehensive evaluation program was that Siemens could verify that the new compressor would operate in a manner consistent with design parameters and performance goals when it returned to commercial operation. A milestone  in every validation program conducted by Siemens for its gas-turbine upgrades is a physical inspection of parts, with the first opportunity planned to inspect the compressor after approximately 4000 hours of service. Once again, the OEM invited participation of the steering committee, this time to witness the inspections and to see the service-run hardware. The compressor cover was removed when the owner/operators arrived at Millennium. Under their watchful eyes, Siemens field and design engineers combed through the compressor. No measurable wear or issues with the new components were in evidence.

The excellent report card, coupled with the verification testing, convinced Millennium’s owner, MACH Gen LLC, to move forward with the new compressor configuration at its New Harquahala Generating Co LLC. The Millennium test center was demobilized and the gas turbine with the new compressor was returned to commercial operation.Final phase of the test program involved a unit inspection in in September 2010 at about 9500 equivalent base-load hours of operation. Results were within Siemen’s design expectations and the new compressor test program was concluded. Familiarity with the redesigned compressor undoubtedly convinced other owners to change out the compressors on their units: By yearend 2010, seven units had replaced their compressors 2.

Testing program for the new compressor required nearly 1000 instruments to verify the performance and integrity of each component 1. redesigned W501G compressor incorporates features of Siemens’ V-style machine to address wear issues COMBINED CYCLE JOURNAL, Third Quarter 2010  149with the new design. The success of the collaborative development program is likely to usher in a new paradigm in problem-solving: OEM and customers working together on issue resolution for mutual benefit. For the W501G units, which all have long-term service agreements with Siemens, this certainly seems to make good sense.

G-class HrsGsMost gas-turbine user groups were formed by owner/operators to create an open forum on issues specific to a given engine model. The benefits included faster problem resolution, avoid mistakes made by others, facilitate collaboration with the OEM, etc. In the early years of these volunteer organizations, virtually all discussion focused on the engine.As GT punch-list items were resolved, program content for annual meetings expanded to address other plant issues, which oftentimes were major contributors to unit unavailability. Presentations on generators, steam turbines, HRSGs, high-energy piping systems, and plant control systems were added to the collective discussion. With the oldest W501G-powered combined cycles at a nominal 10 years of age, the wear and tear of cycling service was beginning to show on the fleet’s HRSGs, which had been designed for base-load service.

The steering committee invited Nooter/Eriksen to update the group on boiler issues and modifications owner/operators might consider to help assure top reliability/performance over the long term. Joe Schroeder, senior VP engineering, and Paul Gremaud opened their 90-min presentation with a review of N/E’s experience. The company’s HRSGs sit behind two-thirds of the W501Gs and 42% of the W501Fs in combined- cyc le  servi ce.  The  comprehensive review of HRSG issues encompassed more than five-dozen slides.Duct-liner damage was the first subject addressed (Fig 3). It is found in a large percentage of the transition ducts connecting GTs to HRSGs made by virtually all manufacturers—particularly where generating units designed for base-load service have been cycled extensively. Solutions offered by N/E included addition of (1) stiffeners on the casing sidewall and/or floor, (2) intermediate liner pins, and (3) extra batten channels and backup angles to stiffen the liner. An increase in the thickness of the liner plate also should be considered. Much has been written on this subject, and other deterioration mechanisms as well, in the COMBINED CYCLE Journal and elsewhere.

Two examples

Access  www.combinedcyclejourna l.com/archives.html , cl i ck 3Q/2008, click “Orlando CoGen” on the issue cover; also “Learn the basics of HRSG inspection.”n HRSG Users Handbook. If a copy is not available at your plant, order online at www.hrsgusers.org. Meetings of the HRSG User’s Group are another good information resource. Its conference programs dig down into problem areas if you’re looking for detail. Next meeting is in April (see ad, p 79). The inlet-duct distribution grid, like the duct liner, is an uncooled component that “sees” GT exhaust temperature. It is used in supplementary-fired units to assure proper flow distribution in the HRSG. The stiffened flat plate generally is made from a 300-series stainless steel and has from 40% to 60% open area.

Problems typically include grid deformation and weld failures caused in large part by working of the material from expansion/contraction during startup/shutdown and from vibration associated with the high gas flows and turbulence (Fig 4). Modifications to minimize the probability of grid issues include (1) addition of or modification to the floor restraint system and to sidewall restraints, (2) addition of a gusset at the sidewall restraint plate, and (3) the addition of stiffener bars on the grid fabric.

Removal of condensate  during startup, operation, and shutdown was the next topic and one that generates considerable discussion at every HRSG User Group meeting. Gremaud called to the group’s attention the 2006 addenda to Section I of the 2004 edition of the  ASME Boiler and Pressure Vessel Code which assure better reheater and HP superheater condensate management.The addenda include Part PHRSG, “Requirements for heat-recovery steam generators,” which contains mandatory requirements for both desuperheater drain pots and reheater and superheater drain systems. Gremaud stressed, “If condensate is not removed, bad things happen.” N/E explained to attendees how condensate issues materialize and how to prevent them. It also showed its solution for preventing water accumulation during desuperheater operation, which meets the requirements of Part PHRSG. Plus, Schroeder explained N/E’s reheater bypass system, which eliminates the need for a desuperheater while improving reliability and efficiency (energy in steam is not quenched with water).

To get a better idea of what happens when condensate is not removed qu i ck ly,  ac ce s s   www . combinedcyclejournal.com/archives.html, click Visit booth 213 HRSG 20083. Duct-liner damage is relatively common in HRSGs required to cycle daily4. Inlet-duct distribution grids, like duct liners, typically take a considerable beating in cycling service 150 COMBINED CYCLE JOURNAL, Third Quarter 2010501G users GroupFall 2004, click “Avoid desuperheater problems” on cover. For more on desuperheater best practices, click 3Q/2006, click “Monitoring and maintaining desuperheaters.” The subject also is discussed in depth in the HRSG Users Handbook.

Additional improvements to accommodate cycling offered by the N/E team included welding of tube stubs to headers in the shop, springsupported tube coils, internal coil flexibility, and more forgiving piping layouts. Regarding tube stub-toheader welds, careful consideration should be given to the type of joint selected (access CCJ archives, click Summer 2004, click “Review basics of tube-to-header joints”). The optimal joint for a cycling unit is one that minimizes header wall thickness.Welding tube stubs to headers in the shop offers a controlled environment conducive to better joints and simplifies the nondestructive examination critical for assuring quality welds. Spring support of headers and internal coil flexibility allow heattransfer sections to move freely and reduce by orders of magnitude stresses that otherwise would be experienced. Investigation of tube leaks  was another import a n t   s e gme n t   o f   t h e   presentation. It offered O&M personnel a valuable checklist for gathering information and for evaluating that information.

Mapping Failure Location

Critical to every tube-failure investigation is a map of the HRSG that pinpoints each failure location, Schroeder and Gremaud told the group. When you find a leak, they recommended inspecting adjacent tubes, checking for tube buckling, and taking plenty of pictures. One of your goals is to work towards a root-cause evaluation. Can you ID pitting, corrosion, swelling, or blistering? Other questions you should ask yourself, include the following:  What component has the problem? What doesn’t?n What is wrong? What could be wrong and isn’t? n Where is the problem? Where else could the problem be but isn’t?n When did the problem first occur? When could the problem have been observed but wasn’t? Any trend or pattern? If the problem has not been noticed since the first occurrence, why not? Don’t forget to extract failure samples where appropriate. Samples are very important. Before removal, mark the sample top, bottom, orientation, tube number, etc. Mechanically remove the sample; burning can destroy it. Protect sample surfaces. Keep in mind that if you don’t determine the cause of a failure, you are likely to experience it, or something similar to it, in the future. How to minimize back-end fouling was the last segment of the presentation. Solutions discussed included improved design of the ammonia injection grid, better AIG tuning, improved control of ammonia injection, additional catalyst to minimize slip, and delay ammonia feed during unit startup. The N/E presenters pointed to sulfur as a culprit in many back-end fouling situations. The plant owner/operator should be vigilant in limiting fuel sulfur content, they told the group. One suggestion was to limit mercaptan addition for gas-leak detection, or to eliminate it altogether by switching to an odorant that contains no sulfur. Methods for removing ammonia salts deposited on heat-transfer surfaces in the cooler sections of the HRSG included both traditional methods—dry-ice blasting, high-volume water flush, and high-pressure air—and nontraditional methods—such as coil vibration, sonic vibration, steam cleaning, sootblowers, heating of coils, etc. Millennium Lpeconomizer mod.

Managing an unregulated electric generating plant is not for the faint of heart. Meeting the budgeted income target generally is top priority, as it is for most free-market enterprises, making everything  that   inf luences income a priority—which is just about everything. In   such  a  demanding  environment, knotty problems can be particularly wearing on the management team. A small onsite O&M staff, generally limited engineering capability at headquarters, parsimonious budget, and days only 24 hours long mean possible solutions may require a long-term team effort before they can be thoroughly researched and presented for budgetary approval. The gestation period may be years. Fouling of the Millennium HRSG illustrates the point well. Steve Snopkowski, O&M manager, explained to the group that because of accelerated tube-side fouling, HRSG backpressure had progressed to the point where the unit could not achieve its base-load rating at ambient temperatures in the low 30s. Left unchecked, fouling would cause operating restrictions at temperatures as high as 40F. Not a concern in Florida,  perhaps ,  but  Mi l lennium is in Massachusetts. He summarized the plant’s concerns with the following bullet points:n Derates were as much as 50 MW below the declared capability.n Operating limitations were imposed because of high HRSG backpressure and high pressure at the steam turbine HP inlet. Backpressure was increasing over time.

If fouling was not addressed, the frequency of derates, and capability lost, would continue to increase.n Plant value would benegatively impacted by declining production.Plant Manager Mark Winne, who started at Millennium during plant construction, said that back in the 1996-1997 timeframe a decision was made to “design-in” HRSG performance goals. Nooter/Eriksen proposed a HRSG with tight tube spacing both to achieve desired thermal performance and to accommodate an economic footprint. Westinghouse Power Generation, which had project responsibility, accepted the design.Winne said that the tube rows in the LP economizer are so close to each other that the fins on tubes in adjacent rows almost touch. Also, there is relatively little space between fins. He mentioned that, based on a Millennium staff evaluation, locating so much heat-transfer surface in such a small space offered little in terms of performance gain. In fact, Winne con-16 17 18 19 20 Compression ratio2008200330O2 25HRSG inlet pressure, in H 2015AB5.

Historic data show a nominal increase in HRSG inlet pressure of 5 in. H2O (point A to point B) over a five-year period because of fouling COMBINED CYCLE JOURNAL, Third Quarter 2010 151501G users Grouptinued, after only about 1000 hours of operation, fouling, exacerbated by the tight spacing, had negated the efficiency improvement designers intended to provide. Aggr e s s i ve he at – t r ans f e r design was only part of the problem. The unit, built for base-load service, was cycling and often dispatched at less than rated output. The resulting “cool” back-end gas temperature increased the potential for fouling of LP heat-transfer sections. Also, as part of the original design of the emissions control system, the HRSG has a CO catalyst bed upstream of the SCR.  While the first catalyst bed provides environmental benefits by converting CO to CO2, it also is effective in converting SO2 to SO3. At the acid dewpoint temperature, SO3 wreaks havoc by combining both with unreacted ammonia from the SCR and moisture to produce ammonia bisulfate, a cementitious foulant.

Performance Monitoring

Millennium’s rigorous performance monitoring program revealed increased backpressure and bothersome heat-rate degradation in the first couple of years of service. Winne recalled attempts at tube cleaning in 2003-2004 using CO2 and compressed air. While generally effective in improving heat-transfer characteristics, tube cleaning was not successful in restoring the backpressure to original conditions. Each year the backpressure crept a little higher, such that Snopkowski had to work with Nooter/Eriksen to design and install additional stiffeners and reinforcement to increase the HRSG’s backpressure rating to 30 in. H2O from the original 25 in. By 2007, the increasing backpressure trends caused plant personnel to believe that conventional HRSG cleaning techniques might never achieve the plant’s performance goals. Snopkowski illustrated that point early in his presentation, which he prepared in collaboration with Bill Lovejoy and Joe Michienzi. Hourly average data for 2003 and 2008, compared in Fig 5, show that over a period of five years, fouling added about 5 in. H2O to HRSG backpressure.

Note that Lovejoy works out of the NAES Corp regional engineering office in New England; Michienzi is an engineer with Competitive Power Ventures Inc, which provides asset management services to Millennium and its owner MACH Gen. The O&M manager explained that compressor ratio (horizontal axis) provides a good indication of gas-side mass flow. The basic premise, he said, is that HRSG components in the gas path are flow restrictions and the change in pressure (vertical axis) over time indicates the increasing restriction caused by fouling of those components. No surprise, Snopkowski added; there’s plenty of visual evidence inside the HRSG.He commented on the outliers in the 2003 data, saying the green points at the lower right of the graph were for data taken before addition of a second half layer of SCR catalyst (for additional NOx reduction capability), which alone contributed about  3  in.  H2O to backpressure.

The mass of red at the bottom of the 2008 data reflects a “clean” HRSG; the upper band, a dirty boiler. The chart allowed Snopkowski and his colleagues to develop a correction to HRSG backpressure to normalize the reading to a fixed compressor ratio/implied flow. Fig 6 illustrates the seasonal variations in backpressure—peaks in winter, valleys in summer. Operators maintain pressure drop below the maximum allowable 30 in. H2O by manually limiting flow through the HRSG using the compressor inlet guide vanes.Snopkowski and colleagues appl ied  thei r   cor rec t ion and normalized the backpressure to a compression ratio of 20 or about 30F to remove the effects of seasonal variations.

This view of the data showed that periodic cleaning of the tube bundles and catalyst—particularly the latter—had a positive impact on performance. However, the new compressor discussed in the first section of this report (refer back to Fig 1) increased mass flow through the HRSG (by restoring the compressor to its new and clean condition) and, therefore, backpressure. Next goal was to chart the pressure drop through the unit, from the reheater inlet to stack outlet, to see what section was having largest adverse impact on performance. Investigators found that the pressure drop  through  the LP economizer, with its significant buildup of ammonium bisulfate, was 10 in. H2O. Thus, a heat-transfer section with only 11 of the HRSG’s 78 rows of tubes accounted for close to half the pressure drop. What to do?

Winne and the Millennium team, together with Lovejoy and his colleague Jim Koch, an independent plant performance consultant based in Hilltown, Pa, believed a few rows of economizer tubes could be removed to reduce pressure drop and create a lane to facilitate cleaning of the remaining tube rows in the LP section with minimal impact on heat rate. There was no precedent for such a solution that anyone could recall.Numbers generated by Lovejoy and Koch suggested that by removing four rows of tubes, backpressure would decrease about 3.7 in. H2O and there would be a commensurate increase in plant output (Fig 7). The penalty associated with doing this was virtually negligible: An increase in heat rate of less than 10 Btu/kWh, assuming the worst case of no thermal benefit from improved cleaning. N o o t e r / E r i k s e n e n g i n e e r s reviewed the manufacturer’s design for the Millennium HRSG and reengineered the unit with the first four rows of LP economizer tubes removed. Their numbers jibed with calculations made by Lovejoy and Koch, giving plant staff greater confidence in the solution proposed. The cost estimate to alleviate the plant’s winter constraint on power production was close to the expected annual revenue benefit from making the modification and the project was funded.

Field Work

Field work  required to implement the modification was challenging because of a tight schedule requirement—nine days from start to finish. There were three bidders, O302 Backpressure, in. H 2520 Jan Jun Jan Jun Jan Jun Jan Jun Jan Jun Jan Jun Jan 2003 2003 2004 2004 2005 2005 2006 2006 2007 2007 2008 2008 2009 6. HrsG backpressure over time illustrates seasonal variations—valleys in summer, peaks in winter. Boiler limit is 30 in. H2O, forcing operators to manually limit gas flow through the HRSG on occasion by changing position of compressor inlet guide vanes152 COMBINED CYCLE JOURNAL, Third Quarter 2010501G users Groupwith Bremco Inc, from nearby Newport, NH, prevailing. Its project manager, Bob Morse, recalled the job’s first challenge: No opportunity for an internal inspection prior to project start.

All planning and est imat ing was done by M o r s e ,   L e a d   W e l d e r   G a r y   M a rtin, and Estimating Manager Dick Grace based on boiler drawings and  in format ion  suppl ied by M i llennium and gathered during an external walk-down. Morse told the editors that critical to Bremco’s decision to bid to the demanding schedule was the company’s ability to provide jumpers requiring half the number of field welds estimated by plant personnel. When the Bremco team gained access to the LP economizer section on the first day of the project, Morse recognized that the severe fouling would make the job more difficult than anticipated. It was tough to find tube ties, drains, etc, he remembered. First step was to cut off the lower headers by torch and drag them outside through the stack door. Next, workers cut across all four rows of tubes near the tube ties located about 10.5 ft above the headers just removed. The opening created was less than 2 ft across (Fig 8).

Note that the Millennium HRSG is three panels wide with 20 tubes per panel. That adds up to 240 tubes in the four rows, each more than 60 ft high. Perspective: Laid end to end the tubes would extend nearly three miles. As each nominal 10.5-ft tube section was cut free, it was lowered to the boiler floor manually, using appropriate rigging (Fig 9). Each segment we ighed near ly 100  lb,   tube  plu s deposits. After the lower 10.5-ft section was removed from each of the tubes, work began at the first tube-tie level. The next section of tube extended from one tube-tie elevation to the next. Planking was placed on the tube ties to form a platform for workers as they cut into the tube rows. Header connection.

The Bremco team cut off the top section of each tube in the first row with a torch, leaving a stub of about 2 in. still welded into the header. Those stubs were trimmed off with a saw and then prepped for welding. Leaving an adequate stub eliminated the need to weld on the header, which would have required stress relief to comply with the  ASME Boiler & Pressure Vessel Code. Fig 10 shows that installation of the new jumpers from the headers to tubes in the fifth row of the LP economizer required removal of fins from the top 6 in. of those tubes to allow for prep work and welding. Bremco used an Esco Tool (Holliston, Mass) definner for this operation and an Esco external-clamping Warthog to make the 120 weld preps.

The jumpers were made by Potts Welding and Boiler Repair Co, Newark, Del. Before Bremco arrived at Millennium, Potts had done half the work involved in jumper preparation. Specifically, it made one of the 90-deg bends shown in Fig 10 and left the remainder of the pipe segment straight. As Bremco personnel called Potts with exact field measurements for each jumper, the second bend was made and the jumpers were shipped overnight to the site.   T h e   j umpers were welded to the tubes first, to minimize the number of mirror welds. The jumper-to-header (tube stub) welds completed installation.Note that quality control is critical to the success of projects such Millennium’s, both for the owner (schedule did not allow time for rework) and for the contractor (rework reduces profit). Bremco’s qualified inspectors verified fit-up before each weld was started, checked each root pass visually, and relied on dye penetrant to verify integrity of the cap welds. The requisite hydro revealed no leaks.

In Summary

Winne said that all-in-all the project was viewed as very successful by owner MACH Gen. It is performing as well or better than planned; and the improved access for cleaning should allow adequate cleaning to arrest the steadily increasing backpressure that forced the Millennium team into action. The success of the project has caught the attention of several other 501G Users who are wrestling with similar backpressure issues. c c j