Ed Avella

Outage Management Special SectionTHE NUCLEAR NEWS INTERVIEWEd Avella: Turkey Point’s awardwinningvessel head replacementEd Avella is director of projects forFlorida Power & Light Company(FPL). A project recently completedby FPL was a big one—the replacementof the reactor vessel headsat Units 3 and 4 of the Turkey Pointnuclear power plant. FPL’s decision toreplace the heads was a proactive onebecause the existing heads were ingood condition and not yet physicallydegraded. But the company wanted toavoid the extensive examinations andrepairs of the heads that were likely tocome down the road, based on industryexperience with primary waterstress corrosion cracking (PWSCC) inAlloy 600 reactor head penetrations.The replacement project involvedthree main parts. First, because the existingequipment hatches in the containmentbuildings were not largeenough to accept the new reactorheads, temporary openings had to be cut through the steeltendons and concrete that make up the buildings. Second,because of the lack of qualified suppliers of nuclear-gradeconcrete in the area, FPL had to set up a temporary facilityto produce replacement concrete of specification standards.Third, with a lack of work space around Unit 3, atower crane had to be used instead of a more traditionalcrawler crane, and the tower crane had to be anchored tofoundation pilings that reached down 65 feet into the earth.The project was completed last June, when the installationjob at Unit 4 was finished, 34 months after the initialformation of a project team to handle the job for thetwo units. One of the project’s main requirements wassafety compliance, and during Unit 3’s head replacement(completed in December 2004—NN, April 2005, p. 42),Faced with hurricanes and a lack of work space,Turkey Point was still able to complete a pair ofreactor vessel head installations in 46.5 days each.Avella: “What we wanted was thereliability of having the units come on linewhen planned and being able to makeelectricity for our customers.”there were only two OSHA recordableinjuries and 18 minor injuries. By theend of Unit 4’s head replacement,through the use of job hazard analysisand lessons learned, no OSHA recordableinjuries were experienced andonly 10 minor injuries were recorded.By the project’s end, records hadbeen set in the industry for lowest radiationdose to workers for both vesselhead replacements. So impressivewas the performance that Power Engineeringmagazine, in its January2006 edition, named the TurkeyPoint-4 work the “Best Nuclear Project”of 2005.Avella said the outstanding performanceof the whole effort was due tothe teamwork of company and contracted workers. Nosingle person carried the project, but all were responsiblefor completing it safely and under budget.The Turkey Point site, in Florida City, Fla., has fourgenerating units. Units 1 and 2 are fossil powered generators,while Units 3 and 4 are Westinghouse pressurizedwater reactors. The nuclear units are each rated at 720MWe. Unit 3 started commercial operation in December1972, and Unit 4 in September 1973.During his talk with Nuclear News, Avella was joinedby FPL’s Mike Moran, engineering manager; Alan Katz,construction manager; and Dick Sipos, project manager.For purposes of this interview, Avella is the only onequoted. The interview was conducted by Rick Michal,NN senior editor.32 N U C L E A R N E W S April 2006

Interview: AvellaWhat was the project’s timeline, from initialplanning to final completion?We started in August 2002, when theproject team was formed, and ended in June2005. We scheduled 67 days for completionof each installation job, but we were able tocomplete each in 46 and a half days, savingconsiderably on costs. Our initial activitywas to order replacement reactor head forgingsfor the two units. That was followedimmediately by extensive engineering andconstruction planning.What were the conditions of the vesselheads when they were replaced?They were the original Westinghouseheads from when the units first came online in the early 1970s, and they had no historyof repairable indications. The projectwas developed based on what was happeningin the industry in 2002, with data indicatingthat there was a trend towardPWSCC in Alloy 600 reactor head penetrations.For FPL, it boiled down to whatthe cost would be to inspect the headsevery outage and then do necessary repairs.The fact is, we didn’t want that hangingover us. What we wanted was the reliabilityof having the units come on line whenplanned and being able to make electricityfor our customers. The new heads gave usthe confidence that we could do that.The process of doing head inspectionsand repairs and then being able to comeback on line is subjective in that there is nopredictability to it. Other nuclear plants, forinstance, have spent months off line doingrepairs or having to do emergency head replacements.It’s a situation we did not wantto get into, because the “risk profile” forthe Turkey Point units put them in the highriskcategory due to their age and the temperatureat which the reactors operate.That’s what prompted us to move quickly.We did not want to have the risk of unplannedactivities.A service platform was constructed next to Turkey Point-3’s containment building. Faintorange markings on the concrete wall indicate the “cut line” where an opening would laterbe created in order to move reactor heads in and out of containment. (Photos: FPL)Which vendor made the replacement heads?The heads were made by Areva, in France,specifically for our Turkey Point units. Weplaced the first order in October 2002, thefirst of what were a few different submittalsof purchase orders. That first order was forthe raw materials that had to be forged intomono-block—or one-piece—heads. The intentwas to get one forging going as the firststep of a horizontal process. By that I meanthat first we ordered the forgings, then thenozzles, then the drives, and finally the integratedhead assemblies. The heads were thendelivered to Turkey Point one at a time,based on a production schedule that got themhere at least 30 days prior to the outages.What were some challenges of the project?One challenge was that the heads had tobe assembled at Turkey Point. All the partswere delivered from France, where theywere made. We like to joke that there was alabel attached that translated from French toEnglish to say, “Some assembly required.”Another challenge was that we had tomake openings in the containment buildingsthat the heads could fit through. We decidedto make the openings a little bigger thanwhat has been the norm in the industry—approximately20 ft by 32 ft—so that we couldbring the heads in completely assembledand in the upright position. That way wedidn’t have to assemble the replacementheads in radiation areas. Each head was puttogether in a temporary building that we callthe monster garage, which is 40 ft by 40 ftby 85 ft high and cost $1.6 million to build.If you can imagine, before assembly eachhead looked like a giant Frisbee, and we putit in the garage, welded the drives on, andthen built the integrated head assemblies ontop of that. Once the head was assembled,we slid it into the containment building andthen repaired the opening in the building.How were the openings created?For each one, we had to de-tension 108tendons, cut and remove 62 of those tendons,and remove 110 yards of concrete.These steel tendons had been tensioned to748 000 pounds. We worked with SGT[The Steam Generator Team, LLC, a companyformed by Areva and WashingtonGroup International] on cutting the openingsand, later, on the restoration process,and also for the rigging and hauling of allthe reactor heads, old and new. To createthe openings, we had to hydro-blast awaythe concrete, which required twelve 500-horsepower diesel engines that poweredfour nozzles that shot out 90 gallons of waterper minute at 25 000 psi per nozzle. Inaddition, the wastewater had to be takenApril 2006 N U C L E A R N E W S 33

Outage Management Special Sectionaway, and the aggregate had to be removedand remediated. Our work on Unit 4 wasthe briefest duration for such a job in the industry.It took only 29 hours for the tendonremoval, and 23 hours for head removal.These are industry records.We also had to set up a temporary plantto make nuclear-grade concrete to patch upthe openings because there are no nucleargradeconcrete suppliers near Turkey Point.The concrete had to match the original designspecification and achieve a designstrength of 7200 psi in less than 72 hours.And it did. SGT handled the concrete mixand material qualifications.Why was a tower crane used for the Unit 3job?In one word, space. Because of the waythe plant site is configured, we couldn’tbring a large crawler crane onto such asmall footprint. We just didn’t have thespace available because Unit 3 is right nextto one of Turkey Point’s fossil units. On theother hand, a crawler crane was used for theUnit 4 job because space was availablethere, but for Unit 3 there was no room foranything but a tower crane.The tower crane was huge—more than250 ft tall—and it had to be capable of liftinga reactor head weighing 143 tons. Forthat job, we had to set up a 100-ton workplatform that had deep “cast-in-place” pilingsfor support. The pilings were necessarybecause of the poor soil conditions ofthe Florida coastline. Each of the 12 pilingswas 24 inches in diameter and 65 feet deep.The crane also had to be able to withstandhurricane force winds of up to 160 mph,which came in handy because soon after installation,three hurricanes hit us. The cranehad been demobilized before each hurricane,of course, but it never suffered anydamage.What were the radiation dose figures foreach replacement job?The Unit 3 job had an industry-low radiologicaldose rate of 5.8 rem, compared withthe industry average of 23.67 rem for headreplacement projects. Unit 4 came in evenlower than Unit 3, for a new industry record,with a radiological dose rate of 5.685 rem.How many workers were on the job, andhow many person-hours were used?Combined, there were 450 FPL and contractedlabor workers on this job. Most ofthem were union craft. All combined, wespent 750 000 person-hours on direct construction,head fabrication at the plant, andinstallation of the heads. And it was wellin excess of 1 million person-hours whenengineering, planning, and support workwere added in. For all that work, there wereonly two OSHA recordable injuries and nolost-time accidents. It was an incrediblysuccessful project. While good planningcounts, it is teamwork that makes the difference.SGT was our main contractor, andthey used four or five different subcontractors.We also got great cooperation fromAreva. And, of course, FPL Project Managementand Engineering managed theproject. No one team did the job. It was allthree of us—FPL, SGT, and Areva—thatdid it. We all became one team, and therewas no difference between utility and contractedpersonnel when we got up to speed.That’s the one thing that brought us successwhere other places have struggled.The unit’s old reactor head was wrapped in a blue cover before leaving containmentthrough the construction opening.Why did you use union craft?It’s our commitment to the buildingtrades. We have a strong belief in teamwork,and it has us in alignment with the internationalbuilding trades—we’re proud ofthat. They respond to us because they knowwe go to them when we need people. Forthis job, SGT and Areva knew that we hadbuilt a relationship based on mutual respectwith the union craft, and that we trust thecraft to do the job right. Our own FPLworkers are unionized, too, but for this jobwe went almost exclusively with unioncraft brought in by our contractors.And if I can blow our horn a little forthe project’s team leaders, I will. This includesmyself, Mike Moran, Alan Katz,and Dick Sipos, who were all in FPL managementpositions on the project and whowere all instrumental in its successfulcompletion. Some credit also has to go toFPL’s senior management, because theyput together a team of people who knowthe plant. We’re not just guys who haveconstruction and engineering backgrounds,but we’re people who haveworked in these plants. We were able tosay, for example, “This is a good featureto add to the new reactors.” It worked outwell getting people from the utility tomanage the project.34 N U C L E A R N E W S April 2006

Interview: AvellaThe new reactor head is suspended from the tower crane before being placed on theservice platform.What lessons learned came out of the project?The Unit 3 head replacement was donefirst and was successful. It got us to changea lot of the process before we went on to theUnit 4 job. The other thing regarding lessonslearned is that the process didn’t really startwith the Unit 3 job. Before that, we did asignificant amount of benchmarkingthroughout the industry. We went to everyutility possible that had a similar project onthe books or was planning one so that wecould take their lessons learned back toTurkey Point. It’s great being in the nuclearbusiness—everyone else calls it industrialespionage, but we call it benchmarking. Wedid more than 10 000 person-hours ofbenchmarking time for the Turkey Pointproject.Did your team handle the pressurizer projectat FPL’s St. Lucie-1 nuclear plant inlate 2005?Yes, it did. The same project team gotback together at St. Lucie and successfullyreplaced the reactor head and pressurizer atUnit 1—and made many more benchmarksfor the nuclear industry. This was performedwithin four months of the Turkey Point replacements,so we accomplished a total ofthree replacement outages within 15 months.Turkey Point’s head replacements recentlywon you a “Best Project” award fromPower Engineering magazine. What’s upnext for your project team?The team is now honing continuous improvementskills for the “Big One.” Thiswill be a steam generator replacement, reactorhead replacement, and a pressurizerAlloy 600 mitigation project at St. Lucie-2,scheduled for fall 2007. The focus on “firsttimequality of welding” will ensure scheduleswill be met. There is no doubt that thisseasoned team will continue setting recordsand industry benchmarks.April 2006 N U C L E A R N E W S 35