9

Click here to load reader

Boiler Retrofit Approach and Result

Embed Size (px)

Citation preview

Page 1: Boiler Retrofit Approach and Result

Babcock & Wilcox 1

W.E. BeckerBabcock & Wilcox

Barberton, Ohio, U.S.A.

G.R. VeerkampPacific Gas & Electric

San Francisco, California

Presented to:American Power ConferenceApril 1-3, 1997Chicago, Illinois, U.S.A.

Partnering on a Utility Boiler Retrofit: Approach andResults

BR-1632

SynopsisIn 1993, Pacific Gas and Electric (PG&E) began the process

of evaluating contracting options and selecting a contractingapproach for use on its NOx retrofit program, which had thegoal of achieving significant reductions in emissions from itsgas and oil fired units. PG&E determined that partnering wasthe preferred method, based on its greater potential for bothimproved quality, risk identification and management, and costcontrol. Ultimately, Babcock & Wilcox (B&W) was selectedas one of two partners and was assigned the retrofit of PG&E’s750 MW Moss Landing Unit 7, a supercritical, opposed wall-fired boiler that is configured with burners in “three per” cellsand with modest flue gas recirculation (FGR) capability.

The approach that B&W and PG&E used to implement ourversion of “partnering” at Moss Landing was founded on: theuse of innovative (for PG&E) contracting terms and conditions;a limited use of integrated design teams; joint scoping, costing,and cost benchmarking efforts; mutual risk sharing; and an in-centive program whose goal was to reward innovation and costmanagement while discouraging the use of change orders as themeans to both maximize B&W’s return and minimize PG&E’sbottom line cost. The scope of work ultimately implemented atMoss Landing using this contract structure consisted of the de-sign, fabrication, installation, and startup/testing of 48 naturalgas S-burners, 8 overfire air “NOx” ports and associated benttube sections, optimized use of FGR, hardware and softwaremodifications to the burner management and combustion con-trol systems, and other ancillary modifications. In addition,PG&E elected to include within this partnering umbrella comple-mentary major refurbishment work on the unit’s two airpreheaters, and waterwall replacement. (Under a separate butsimilar partnering contract, major overhaul work was performed

by the OEM on the HP/IP and LP steam turbine trains, and onthe HP and LP generators.)

The actual NOx and main unit overhaul contractual and sys-tem performance achieved on Unit 7 was very encouraging.During the course of the largest single unit overhaul in PG&E’shistory (based on dollar volume of work):

• No scope-related change orders were issued on any of thethree B&W contracts;

• The final cost was less than the project’s Target Price bynearly 13%;

• Material and design quality/design coordination were aboveaverage;

• Construction efficiency by all prime contractors wasexcellent;

• Construction coordination between PG&E, B&W, and theother outage contractors was outstanding; and

• The bottom line performance of the retrofit was good, withtuned full load NOx emissions at the design operating point(14% FGR to the windbox, 1.8% backend O2, and CO <200 ppmv) running in the range of 65-75 ppmv, versus aguarantee limit of 84 ppmv.

BackgroundIn the first quarter of 1993, PG&E initiated the first of what

would be two significant restructurings/downsizings of its PowerGeneration and Engineering/Construction business units, sig-nificantly reducing the size of the home office power plant en-gineering staff, and eliminating the construction componentwithin Power Generation. At the same time, a major programwas initiated to address new NOx emission rules being devel-oped in the three air quality management districts within whichtheir gas/oil fired power plants operate.

Page 2: Boiler Retrofit Approach and Result

2 Babcock & Wilcox

This program was charged with several tasks:• Participating in rule development and public comment dis-

cussions with the districts;• Determining the anticipated scope and timing of the re-

quired retrofits;• Identifying implementation risks/uncertainties, and meth-

ods to minimize those risks and uncertainties;• Reviewing contracting strategies and selecting a contract-

ing approach; and• Managing the resulting retrofit projects and associated con-

tracts.In assessing contracting strategies, major considerations in-

cluded:• The highly dynamic nature of PG&E’s business environ-

ment;• The recent restructurings, resulting in major reductions in

both manpower and experience levels;• A strong need for cost reductions in the midst of an in-

creasingly demanding and uncertain regulatory climate,centered around expected major restructurings of the Cali-fornia power industry;

• Uncertainties with respect to both the timing and final scopeof the expected retrofits; and

• Mixed experiences with “traditional” contracting ap-proaches (typically Engineer Procure Construct [EPC]/Lump Sum Turn Key [LSTK]).

PG&E’s resulting goal in reviewing contracting methodolo-gies was to identify those that provided the greatest flexibility inaddressing and managing the scope and timing risks, while mini-mizing costs and maintaining an acceptable level of quality.

Because of its familiarity with EPC/LSTK contracting, PG&Einitially directed its efforts along this line. This included thepreparation by an A&E in late 1993 of an EPC “Standard NOx

Specification” that could be used as the base from which unit-specific retrofit specifications could be generated. In parallelwith this effort, PG&E began its detailed assessment of alterna-tive strategies.

Early on in this process it became evident that the “risk trans-ference” that EPC contracting provides would be costly and ofquestionable benefit, especially given the scope and timing un-certainties. In addition, the level of PG&E manpower that wouldbe required to support (for each project) specification prepara-tion, bidding, contract administration (both technical and com-mercial), and engineering reviews, would tax the limited re-sources that were available, and rob them from other work re-quired to support day-to-day operating needs. As a result, theEPC approach lost its appeal in favor of strategies that promotedflexibility. The two alternate methods that surfaced were:

• The use of an A&E firm to act as a combination Owner’sEngineer/Project Manager – an approach used by numer-ous other utilities but not recently used by PG&E. Thisapproach would assign to the A&E the responsibility forall design, procurement, contracting, and construction/con-struction management, with PG&E assuming a generalmanagement and oversight role. The brunt of PG&E’s ef-fort would be in the initial selection of the A&E, determin-ing the timing and general scope of each retrofit, and re-viewing/approving the cost estimates. The A&E would befully charged with detailed scoping, cost estimating andproject execution.

• Partnering – a more novel approach that was at that timejust beginning to surface as a “legitimate” strategy within

the industry, having been employed to a limited degree byOntario Hydro and Florida Power & Light. In this instance,PG&E’s major efforts would include partner selection; jointscoping, cost estimating and benchmarking with the part-ner; and to a limited degree joint design, with lead designresponsibility for a given design area assigned to the partythat could best manage that area. PG&E’s partner would beresponsible for its design areas, overall design coordination,material procurement and/or fabrication, and construction.

In selecting a preferred approach, a combination qualitativeand quantitative decision analysis was performed, comparingand contrasting the strengths and weaknesses of these two ap-proaches along with EPC contracting. The analysis graded theapproaches on a high (“1”) to low (“3”) scale for 10 key at-tributes, the most important being:

• Flexibility/Adaptability. The ability to easily respond toscope and/or schedule changes was expected to be best sup-ported by partnering, based on experiences described byother companies contacted who had used the partneringapproach. The proviso to this was that a strong workingrelationship between the partners needed to be achieved.

• Quality. With a joint focus on risk identification and man-agement, PG&E felt that, during component and materialsizing/selection, partnering would tend to promote a morereasonable balance between the legitimate requirements ofthe application and the desire to maximize return/minimizecost.

• Cost Management. Again, partnering appeared to have theedge. A key assumption was that joint risk assumption andmanagement would focus both parties attention on how bestto achieve the end objective. Further, the success of thiseffort would depend on a level of trust being attained be-tween the partners. To achieve this, the support and com-mitment of management within both companies would becritical to both the development and acceptance of theunique terms and conditions that would be required, and totheir equitable application on a day-to-day basis.

• Risk Management. With the implementation of risk shar-ing, and through the use of appropriate monetary incen-tives, it was felt that partnering offered the most potentialfor risk mitigation.

• PG&E Manpower. Partnering would require a greater up-front manpower commitment than would the use of an A&Eas described previously, but less on average than EPC con-tracting. The major efforts that would be expended werein selecting a partner(s) and setting up the partnering ar-rangement; in the initial scoping, costing, and cost bench-marking efforts; and in the design efforts that would beultimately assigned to PG&E.

The advantages, in PG&E’s opinion, far outweighed the needfor additional internal resources, especially given that the in-crease was expected to be modest. A management review ofthe decision analysis supported the recommendation to partner,and the partner selection process was initiated in late 1993.

Project Setup

Partner SelectionA two step selection process was used. The first was a screen-

ing of thirteen A&E, system supplier, and OEM companies; thesecond was the issuance of a formal partnering “Request forProposal” (RFP) to the short-listed companies.

Page 3: Boiler Retrofit Approach and Result

Babcock & Wilcox 3

A screening “Statement of Qualification and Interest” (SQI)was prepared that solicited several “types” of informationdeemed fundamental to selecting a partner(s). These can be seg-regated into the following areas:

Interest and Experience in Partnering. It was possiblethat one or more of the companies would not feel comfort-able entering into a nontraditional contract. Further, it washighly desirable that a future partner have some experiencewith this approach. Each provided information on their in-terest in, experiences with, and commercial concerns rela-tive to partnering (particularly as they pertained to cost/sched-ule/performance risk sharing, contractual terms, open bookcosting, and dispute resolution processes).

Experience with and Understanding of NOx ReductionTechnologies. Given the nature of the program, experiencein NOx technologies was a must, along with the ability toboth analyze the full range of potential technologies for agiven application, and select the most cost effective technol-ogy/technology combination.

Boiler/Draft System Analysis. In reviewing possible sys-tem retrofit strategies, PG&E determined that it was likelythat certain units could benefit from an analysis of currentcombustion/draft system performance, and some level ofoptimization/tuning. In addition, draft margin was minimalon several others. To minimize the impact of the retrofits onboiler/draft system performance and to take advantage ofpossible cost effective tuning options, it was important thata partner have experience in this area.

Staff Depth and Commitment. In conjunction with theabove items, the depth of the staff that would be assigned tothe work, and the level of commitment each company waswilling to provide regarding the duration of the assignmentof lead personnel to the NOx work was of interest.

Design and Construction Coordination, Site Coordina-tion. The ability to: manage (and preferably perform) bothdesign and construction, and coordinate the two with the goalof identifying and managing scope, schedule, performance,construction, or other risks; and to identify and addressconstructability issues, were critical to project cost minimi-zation. Further, the willingness to coordinate with other sitecontractors was important to PG&E in managing overall out-age costs and risks.

Financial Health. Each company provided an assess-ment of their current financial condition, which PG&E com-pared against its own information sources.To “score” the responses and select the final bidders,

weightings were assigned to each of the above areas consistentwith each one’s relative importance to PG&E. The responses tothe SQI in each of the areas were either quantitatively or quali-tatively graded on a 1-10 scale, with an overall score developedby multiplying the grade by its weight, and then totaling up allof the resulting products. The partnering RFP was then issuedto the five companies with the highest scores.

As might be expected, the RFP was divided into “technical”and “commercial” sections. The technical portion provided forthe following interchange:

Information/Material provided:The intent of the RFP. To select at most two partners,

each of which would be assigned the responsibility to scope,cost/benchmark, design, procure/fabricate, erect, and startup/test the retrofit of at most two units (additional work, if any,would be subject to performance on these two jobs);

The expected unit-by-unit retrofit schedule;The Standard NOx Specification. Included to provide

functional/performance requirements, and the industry codesand standards that would apply to the work;

Conceptual designs for two units. These had been de-veloped by PG&E and various contractors during earlier plan-ning efforts, and consisted of general arrangement drawingsfor the retrofit of SCR on each unit, along with a writtendescription of both the scope and boundary conditions thatshould be assumed for each retrofit. These scopes/arrange-ments, along with the other information included in the RFP,provided a consistent basis for each company to use in esti-mating staffing requirements, associated labor overheadsrequired to support that staff, staffing durations per project,etc. These were in turn needed in setting chargeout ratesand profit margins/fee percentages.Information requested:

Staffing. The bidders were required to identify the leadstaff that would be committed to the Program and the timeperiod of their commitment, and to provide an expected re-source loading curve for the duration of projects similar tothose described by the conceptual designs.

Final scoping/design/construction approach. Each com-pany was requested to describe the approach they would usein preparing and optimizing the final scopes and cost esti-mates for the two conceptual jobs, and in conducting thedesign, procurement, construction, and startup/testing phasesof a project in general.

Benchmarking. The approach proposed for open bookcost estimate preparation/reviews and benchmarking was tobe submitted.

Rates. Labor rates, per diems, multipliers, and proposedprofit/fee structures were required.

Partnering and/or Commercial Issues. Thoughts andproposals were requested on: appropriate goals and objec-tives for the projects; how risk would be shared; how costswould be monitored and change orders managed; and howdisputes would be resolved.The commercial section provided PG&E’s proposed Blan-

ket Terms and Conditions that would apply to the overall con-duct of the work. These included significant modifications fromstandard PG&E terms in the areas of indemnification, liability,warrantee requirements, and dispute resolution. This documentalso allowed for cost and risk sharing, with the specific details ofhow cost and risk would be shared to be determined for eachproject when its specific scope, cost, and risks had been identified.

In evaluating both the resulting proposals and informationreceived during follow-on site interviews, a qualitative/quanti-tative analysis similar in nature and methodology to that per-formed for the initial screening was performed. Of primary sig-nificance was the experience/depth of the staff and the processthey would use in performing technology selection and scoping,cost estimating and benchmarking, and project execution; andthe expected overall cost effectiveness of that entire process.Further, the methods proposed to identify, manage, and sharerisks; establish project-specific goals and objectives; and es-tablish and manage a partnered contract were considered. Theultimate result of this evaluation was that B&W was selected asone of two partners.

Project ScopingA two day partnering workshop was subsequently scheduled,

Page 4: Boiler Retrofit Approach and Result

4 Babcock & Wilcox

directed by a jointly agreed-to facilitator. This “meeting” servedthree primary functions: 1) it provided a means for team mem-bers to meet each other; 2) it had as a primary goal the prepara-tion of a mission statement, and goals and objectives; and 3) itallowed for an expeditious preparation of each company’s phase1 (preliminary scoping) responsibilities, along with a Phase Ischedule.

While any given project will have its own unique issues andobjectives, the following seven primary goals and objectives,selected for the initially expected SCR retrofit project (andmodified only slightly for the ultimate combustion modifica-tions work scope), were identified at the workshop and are listedto provide an indication of the issues driving the Moss LandingUnit 7 retrofit:

Safety. No safety incidents occur that affect people, theenvironment or equipment. The intent was to eliminate bothlost time and recordable accidents; prevent damage to exist-ing equipment; and perform the work in a manner that mini-mizes the chance for environmental incidents.

Cost and Schedule. The project is completed at the low-est total life cycle cost and does not impact the overhaulschedule (e.g. maintain the competitive position of the unitin the generation market).

Project Improvement. PG&E and B&W would jointlydevelop in a timely manner the detailed project scope; per-form benchmarked cost comparisons with comparable proj-ects and/or portions thereof; and then attempt to beat theTarget Price estimate by at least 3%.

Project Direction. The project is done right the first time(e.g. no equipment problems; no curtailments as a result ofproject activities; etc.); and the industry recognizes this as atrend setting project.

Plant Impacts. During the outage, the project would notnegatively impact either other projects; plant operations; oroverall unit performance.

Project Support. The project obtains regulatory, commu-nity, and labor support; positively impacts the local economy;and meets PG&E’s small business/minority (“EOPP”) par-ticipation guidelines.

Partnership Satisfaction. B&W will realize a financialreward commensurate with achieving the above goals (andby definition, although not specifically stated in this goal,PG&E will obtain a cost effective, well-running retrofit).At the conclusion of the main body of the workshop, PG&E

and B&W project managers, lead engineers, and constructionpersonnel met to lay out the project’s Phase I (scoping) sched-ule. The planned outage start date was used to establish theexpected start of pre-outage work. This was in turn used todetermine dates of major Phase I milestones such as: providinginformation to support permit activities; preliminary review ofapplicable technologies/technology combinations; on-site re-views; development of the final project scope and cost; costbenchmarking; etc.

Phase I scoping, which initially culminated in the prepara-tion of a preliminary design (described in a “Project ScopeBook” - PSB) and associated benchmarked Target Price for anSCR retrofit (defined in a “Project Cost Book” - PCB), includeda high level review of a wide range of technologies that couldmeet the project’s goals. The technologies reviewed includedSCR (both conventional, and “in-duct” such that an air preheaterrelocation would not be required ), Selective Non-CatalyticReduction (SNCR), low NOx burners, Overfire Air (OFA) ports,

and increased/optimized flue gas recirculation. SNCR, whichhad been reviewed in prior PG&E assessments, was eliminatedbased partly on its inability to meet the 10 ppmv NOx emissionlimit, but mainly due to expected high levels of ammonia slip(which was likewise regulated to 10 ppmv).

Screening of the technical alternatives was performed usinga matrix of thirteen independent variables shown as Options 1through 10 below, and resulted in a matrix of 384 individual,unique cases. The basic options and subsets (note that the draft-related options were included to address the limited margin ondraft system capacity and associated furnace pressure designlimits) were:

1. Relocate the existing air preheater 2. Change air preheater size 3. Use conventional SCR 4. Forced draft fan upgrades 5. 40 low NOx burners without Over Fire Air (OFA)

ports (using existing 8 top burners as “OFAs”)• Operate with maximum flue gas recirculation• Operate with minimum flue gas recirculation

6. 48 low NOx burners with 8 new OFA ports• Operate with maximum flue gas recirculation• Operate with minimum flue gas recirculation

7. SCR Catalyst Type• Size SCR for plate type catalyst• Size SCR for honeycomb type catalyst

8. Consider economizer modifications to reduce draftloss

9. Remove the steam coil air heater10. Change air preheater surface profileSubsequently, B&W and PG&E analyzed the impact of one

additional variable: relocating the existing air heaters (in a dis-assembled state) as they were being refurbished (new rotors,heating elements, diaphragm sections, and seals) under a sepa-rate plant maintenance program. This analysis was conductedin a second matrix consisting of 96 cases. The operative as-sumption was that the cost associated with relocating a com-pletely disassembled air heater should be less than relocatingan entirely assembled one. This assumption proved correct, andthe ultimate number of alternative arrangements screened wasincreased to 480.

The screening matrices were constructed by developing allof the logical combinations of the independent variables andthen using a spreadsheet to calculate the value of all of the de-pendent variables for each combination. The key dependentvariables included:

1. Inlet NOx to the SCR; 2. Total catalyst volume required (plate or honeycomb); 3. Aqueous ammonia consumption rate; 4. Expected dP of the SCR system (including any

future catalyst additions); 5. Net difference in operating kW, considering both FD

and GR fan systems; 6. Total dP saved by the combined modifications; 7. Net dP increase/decrease; 8. Catalyst cost for type and volume selected; 9. Total erected cost for the related portion of project;10. Hourly full load ammonia cost;11. Hourly full load kW costs;12. Ability to install additional future catalyst; and13. Future catalyst replacement scenario. Scenario A was

a full catalyst replacement every six years, while

Page 5: Boiler Retrofit Approach and Result

Babcock & Wilcox 5

Scenario B was only a partial replacement over thesame period.

The selection process short listed 24 cases for economicevaluation, from which two arrangements were selected for fi-nal evaluation. These two arrangements, while using differentcatalyst types, both included the installation of 48 low NOx burn-ers and 8 OFA ports. The scope for the arrangement selectedwas documented in the PSB, which included all of the detailedassumptions for and bases of the work that would be performedby both companies, along with the guarantee, warrantee, risksharing, and other commercial terms specific to this project.

Risk sharing on this project took two forms: 1) sharing ofany overrun or underrun; and 2) sharing the cost for corrective/warrantee work. For underruns, it was agreed that B&W wouldbe paid a bonus that was a function of both the degree ofunderrun and the project’s performance against its goals. Incalculating performance, to the maximum degree possible quan-titative measures that supported the previously stated goals wereidentified, with numeric ranges of 0 (failure) to 1 (perfection)established. Each specific measure had a weighting assignedbased upon the relative importance of that measure. Total scoreswere then calculated as the sum, for all measures, of the valueof each measure multiplied by its weight.

For both overruns (in which case no bonus would be paid)and warrantee/corrective work, a schedule was prepared thatdefined (depending on the degree of overrun and/or the amountof corrective work), the percent participation for each company.

An associated PCB detailing all of the costing assumptions,unit pricing information, labor rates, material pricing, etc., wasalso prepared. This pricing information and the associated TargetPrice were benchmarked to demonstrate the competitiveness ofthe project. The benchmark study was prepared by comparingthe pricing (and its components) with recent B&W bid informa-tion, commercial data, and industry standard costing factors toverify competitiveness with both industry and recent B&W gas-fired utility boiler low NOx projects. The highly site specificnature of this retrofit and the desire to corroborate B&W’s database made it desirable to incorporate the use of commercial dataand industry standards in the analysis to verify the cost of keymaterial and erection components of individual subsystems.

Out of this effort, several key facts resulted :• The cost of the overall SCR and its subsystems (catalyst,

ammonia storage and injection, and related controls/elec-trical) was in fact competitive, and less than comparablesystems studied. The cost of site specific retrofit compo-nents (flues, ducts, steel, foundation, and demolition) wascomparable to retrofits that had either been recently com-pleted and/or bid for utilities in Southern California.

• The Moss Landing combustion system material costs (in-cluding engineering) were lower than comparable costsfrom three recent commercial burner and NOx port retrofitprojects.

• Installation estimates for the flues, ducts, reactor, combus-tion system, and ammonia system compared favorably withboth other competitively bid commercial projects and in-dustry data.

Project RescopingSeveral key advantages of the contractual arrangement es-

tablished for this project (and its associated joint SCR scoping/costing effort) were demonstrated when PG&E and the local airpollution control district initiated discussions to review poten-

tial amendments to the NOx rule. These discussions were begunat the completion of Phase I/the beginning of final design andprocurement . The amendment that was ultimately enacted al-lowed deferral of Unit 7 SCR in favor of combustion modifica-tions on both Unit 7 and its sister Unit 6:

1. The close scoping coordination facilitated the develop-ment, in a very short two week time frame, of technicaldocumentation needed to substantiate preliminary perfor-mance predictions for the proposed combustion modifica-tions-only retrofit.

2. Understanding the detailed bases and assumptions of theSCR scoping provided the ability to quickly convert thatscope to the new low NOx burner/combustion modifica-tions scope, with its final scope and cost books.

3. Since B&W had been released by PG&E to begin finalSCR design and procurement, the ability to coordinatetogether to control that effort during rule negotiations al-lowed PG&E to minimize associated cancellation costs (agood “for instance” is catalyst procurement, the contractfor which had already been let) while simultaneously de-veloping the new work scope. As a result, the final SCRcloseout costs were minimal.

A new PSB and PCB were developed to reflect the majorchange in project direction allowed under the revised NOx rule.These new documents were created by eliminating the SCRportion in the original books, retaining the combustion portion,adding some new scope (e.g. increased reheat spray water ca-pacity), and adjusting some of the guarantee and warrantee as-pects of the commercial terms.

Concurrent Outage ScopingIn parallel with the above SCR detailed (“Phase II”) design

and its subsequent conversion to the final combustion modifi-cations scope, PG&E was reviewing and finalizing the scope ofthe overall Unit 7 overhaul. Because of historical problems withrotor post cracking on the horizontal shaft air preheaters, it hadbeen decided to perform a full replacement of both wheels. Inaddition, furnace wall fireside thinning necessitated replacementof both of the opposed firing walls from part way up the hopperslope to above the top row of burner cells.

Because of the extreme degree of material-related site con-gestion; the intimacy of the APH and waterwall work with theSCR and combustion work scopes respectively; and because ofthe generally favorable performance that had been achieved todate with the partnering arrangement, it was decided in early1995 to include the APH demolition/construction activities un-der the partnering umbrella but with its own contract. A similardecision on the waterwall effort was made later that year. Thesecontracts both used the partnering terms and conditions estab-lished for the NOx project, and were priced and benchmarked inthe same fashion as for that project.

The other major portion of the overhaul centered around theHP/IP and LP steam turbine/generator trains. Again, because ofthe satisfactory performance that had been realized up to thatpoint under the B&W agreement; to hopefully attain a higherquality work product than had been realized using more tradi-tional ST/G overhaul contracts; and to do this while also en-couraging contractor coordination, PG&E decided to develop apartnering agreement with the OEM. Such an agreement, simi-lar in nature to B&W’s, was executed in the latter half of 1995,and PG&E, B&W, and the OEM thereafter began coordinatedoutage planning.

Page 6: Boiler Retrofit Approach and Result

6 Babcock & Wilcox

Project ExecutionWhile after Phase I scoping the project was, in many respects,

executed in a standard fashion (e.g. completion of final design -with two formal design reviews; material procurement, fabrica-tion, and inspection largely coordinated by B&W; and erectionlargely supervised/performed by B&W), several things weredone differently that, in the authors’ opinions, contributed sig-nificantly to the final results. These included:

Integration of ForcesThe strength of each company’s engineering resources had

previously been reviewed, with the aim of both establishing in-tegrated design teams where appropriate and assigning lead de-sign responsibilities in the project’s various design areas. Thiswas done to both reduce duplication of effort and to assign workto either B&W or PG&E based on each company’s ability tobest perform that work (and thus minimize project risk). Sev-eral teams resulted:

• Electrical/Controls. Led by PG&E, this design group wasresponsible for all electrical and controls design/optimiza-tion, DCS programming and hardware modifications to theburner management and combustion controls systems, pre-paring procedures for and performing/documenting the re-sults of electrical pre-ops, and initial checkout of the soft-ware modifications.

• Combustion System Design. While B&W naturally hadthe lead in this area and largely performed the work with-out direct PG&E involvement (other than for design re-views), a small group of B&W and PG&E personnel wasassigned the responsibility for coordinating and managingthe physical and numerical flow modeling efforts that werecentral to one of the project’s goals of optimizing the per-formance of both the forced draft and flue gas recircula-tion systems (and thus, hopefully, NOx performance).

• QA/QC. A team of B&W and PG&E inspectors was usedto establish both fabrication and construction quality con-trol requirements and acceptance criteria, monitor the con-struction effort, perform tests/inspections of the physicalwork, and document the results of same. This team wasled by PG&E.

• System Testing/Tuning. B&W led this effort with a teamconsisting primarily of individuals from PG&E’s powerplant testing group, along with Service engineers fromB&W. The actual testing/tuning plan was developed byB&W with input from PG&E, while data gathering and re-cording was primarily PG&E’s responsibility.

“EOC”Direct Management support and reviews were provided via

the use of quarterly “Executive Oversight Committee” (EOC)meetings. This forum was used both as a vehicle for keepingmanagement within both companies informed of project status,but also to relatively quickly discuss and resolve the few issuesthat arose during execution.

Baseline TestingBaseline testing was performed to determine necessary de-

sign information. These baseline tests were also used to deter-mine existing unit performance, since several guarantees werebased on not degrading boiler performance. As a result, baselinetesting established pre-retrofit capabilities for guarantees atmaximum and minimum boiler load, boiler ramp rate (both for

raising and lowering load), boiler emissions of NOx and CO,stack opacity, tube metal temperatures, spray water flows, fur-nace pressure, and boiler efficiency.

Baseline data also included burner-to-burner and FGR faninlet traverses at maximum flue gas fan flow. This data wasused both to determine windbox O2 (for calculating the degreeof flue gas/combustion air mixing), and to establish an FGRflow profile. The flow profile was used as input data to both thenumerical and cold flow models.

ModelingThe design process included the use of both numerical and

cold flow modeling. Numerical modeling using B&W propri-etary software was used to locate the OFA ports so as to mini-mize CO, and also to calculate the mixing of flue gas with thecombustion air on a burner-to-burner basis. Cold flow physicalmodeling with a 1/10th scale model was used to optimize theflue gas flow to the SCR and determine the expected increase insystem pressure drop. As a result of SCR scope elimination,this model was modified to include the air preheaters, mixingfoils, windbox, burners, and OFA ports.

The objectives of the final cold flow boiler model were: a)to determine what flow balancing devices (e.g. turning vanes,perforated plates, etc.), if any, would be needed to balance airflows between the front and rear windboxes to within ± 2% ofthe overall bulk average, and between burners on a wall to within± 10% of that wall’s average; and b) what modifications to thewindbox and/or FGR mixing foils were needed to attain the samedegree of FGR distribution (FGR was modeled using CO as atracer gas). Combining the baseline test data with the results ofthe modeling effort, the “as found” condition was determinedto be within our acceptance limits, and addition of flow deviceswas not needed.

Construction/Outage Integrated Planning andScheduling

Several design and constructability reviews were performedat Moss Landing. These were attended by lead B&W construc-tion and design personnel, along with PG&E designers and keyplant personnel, and focused on operating interfaces, mainte-nance access, and constructability issues. These meetings, alongwith coordinated outage planning, helped to adjust the designto meet the needs of the plant and to minimize rework.

Detailed outage planning began six months prior to outagestart. This planning occurred on several levels. At the highest,the plant developed an overall outage plan and “master” sched-ule. This schedule was prepared with not only the input of plantpersonnel responsible for the various overhaul activities, but alsowith the participation of the lead contractors.

In preparing the master schedule, PG&E and each contrac-tor first developed their own schedules using their normal soft-ware tools and processes. Sequences were reviewed and modi-fied as needed to minimize and, where possible eliminate worksite congestion, access, laydown, utilities, or other constructioncoordination issues. When identified, additional resources (e.g.temporary construction power supplies, portable compressors,etc.) were included in the appropriate contractor and/or PG&Ework plan. When this process was completed, the resulting con-tractor key activities and/or milestones were identified, and theseitems were mapped for statusing in the master schedule.

This relatively detailed upfront planning allowed for numer-ous iterations and “what ifs” to be analyzed using the schedul-

Page 7: Boiler Retrofit Approach and Result

Babcock & Wilcox 7

ing software, which in turn led to reductions in both the overallNOx retrofit duration and its cost. An example of a significantidea that resulted from this planning/constructability effort wasthe decision to shop fabricate the burners in a “three per” con-figuration that was essentially ready for installation on site. Thiscontrasts with the original (and normal) process of fabricatingand shipping burners and support hardware separately, and thenindividually installing the burners and supports via windboxaccess holes. This revised approach was practical since water-wall removal allowed the entire burner assemblies to be riggedinto place through the lower corner hopper vestibule/furnaceaccess holes that were cut for waterwall demolition and instal-lation purposes.

During the outage, the schedule activities were electronicallystatused and forwarded daily to PG&E using the plant local areanetwork as the communication vehicle (each contractor had atleast one net ID). Depending on contractor, either an Excelspreadsheet of current activity start and end dates and remain-ing durations, or a milestone status listing was provided. Inaddition to this statusing, daily outage coordination meetings(maximum of one-half hour) were held. These provided a fo-rum to discuss and resolve unanticipated personnel, material,access, or other coordination issues. Those that could not beresolved within the time constraints of these meetings were ad-dressed “off-line” by the affected individuals later that sameday.

Contractor CooperationCertainly during these meetings, but also during the over-

haul in general, a high degree of cooperation was demonstratednot only between PG&E and B&W, but between contractors. Itwas not uncommon, when a site visitor was about to leave, forthat individual to make a remark to the effect of “I‘ve neverseen this level of teamwork before.” An example of this coop-eration is the numerous times that either B&W, PG&E, or themain unit OEM provided the use of either their own personnel,material, or equipment to address and resolve a near term, criti-cal need of another.

In the authors’ opinion, there were two primary motivatingforces at work:

• The relatively strong working relationships that developedbetween both the contractors and PG&E leads during thefinal scoping and integrated schedule development efforts.With these individuals given the authority and responsibil-ity by their respective organizations to “make it happen”,this relationship provided the catalyst needed to promoteon-site cooperation.

• The risk/reward sharing features of the contracts. With thefocus of both contractors directed towards achieving theircontractual performance goals rather than on identifyingchange orders as the means of maximizing profit; with thesegoals tied to quality, safety, and final equipment/systemperformance targets of importance to PG&E; and with eachcontractor’s goals structured such that they did not forceone into an adversarial position with the other; a situationwhere collaboration could be (and was) achieved resulted.

ResultsThe obvious goal of the project was to reduce NOx. This

objective was achieved, with initial startup and tuning of Unit 7resulting in performance 15% below the guaranteed NOx levelat the design point, and up to 30% below guarantees at maxi-

mum FGR. In addition, a significant 0.8% improvement in boilerefficiency at full load was achieved (primarily as a result of theAPH modifications). A summary of key results is listed in Table 1.

Table 1

Baseline Max. ExpectedUnit 7 FGR FGR

Load, MW 750 750 750

Steam Flow, 103 PPH 5200 5100 5100

O2 @ Economizer Outlet, % 1.8 1.1 1.8

NOx @ 3% O2 , ppmv 115 58 73

CO @ 3% O2 , ppmv 174 297 179

Flue Gas to Secondary Air, % 8 17 14

FD Fan Discharge, in. wg. NA NA 35.3

Windbox to Furnace 7.0 5.8 5.3 Differential, in. wg.

Furnace Pressure, in. wg. 21.5 23.5 22.2

Reheat Spray Flow, 103 PPH 17 101 120

Especially for the maximum FGR operating condition notedbelow, reheat spray water flow and boiler furnace pressure wereat the high end of the expected range. Moss Landing personnelwere concerned that, over time, furnace pressure would increase(primarily as a result of the impact of degraded turbine effi-ciency on boiler firing rate), resulting potentially in lower FGR,higher NOx, and/or continuous operation at or in excess of maxi-mum reheat spray water capacity.

Data from previous pre and post outage testing, along withbaseline performance tests and a sensitivity analysis that used acomputer boiler model, indicated that the boiler was perform-ing as expected, although very little operating margin remained.Because of the concern over lack of margin, potential fixes wereidentified (scoped and costed) for implementation should furnacepressure increase over time. These included creating lanes inthe horizontal reheater to reduce draft loss between the furnaceand stack, and increasing reheat spray capacity. Either or bothof these fixes will be implemented at a later date if necessary.

Aside from these technical results, several other boiler-re-lated and general project/contractual performance items shouldbe noted:

• The outage safety record was exceptional. Out of 110,000total work hours, B&W experienced only one OSHA re-cordable incident and no lost times; the OEM’s record waszero and zero.

• The electrical/controls design team was able to significantlyimprove the design from that originally envisioned duringPhase I scoping, reducing the cost and enhancingconstructability. One consequence of this improvement wasthat the preoperational checks and startup testing and tun-ing went very smoothly. As an example, out of approxi-mately 500 circuits pulled and terminated, fewer than 10misterminations were found.

• The amount of corrective “punchlist” work was minimal.• No scope-related change orders were issued against any of

the three B&W contracts, and in fact savings for each wererealized. The savings specific to the NOx project was

Page 8: Boiler Retrofit Approach and Result

8 Babcock & Wilcox

slightly less than 13% of the Target Price. The majority ofthese savings were the result of construction efficienciesidentified either during constructability reviews or on-siteduring final outage planning and execution. An example isthe identification by the construction leads of savings thatcould be attained by shop fabricating the burners in the“three per” cell configuration previously noted.

• All critical schedule milestones were achieved. The Me-chanical Completion date (combustion system ready forboiler cold air testing and then unit startup) was betteredby one day, and could have been pulled up by an additional4-6 days had it been necessary. Instead of pulling up thedate, the work plan was adjusted during the latter half ofthe overhaul, with the workforce size and extent of twoshift construction reduced to minimize costs while meet-ing the required date.

• Similar performance was achieved on the turbine/genera-tor overhaul contract. While in this case there were changeorders (which would be difficult to avoid given the natureof the work), those for extra work essentially balanced outthose for scope reductions. When combined with the effi-ciencies realized by the OEM in its use of internal engi-neering and shop resources, the final cost was within liter-ally a few dollars of that contract’s final Target Price. Thiscost performance was the result of thorough preplanningand a high degree of teamwork between the OEM andPG&E, and contrasts with PG&E’s historical experienceon most turbine/generator overhauls of cost increases ofabout 3-10% over the go-in contract price.

• Steam turbine/generator performance during startup, ini-tial testing, and normal operation was quite good. The unitwas able to achieve in excess of 650 MW prior to placing abalance shot, and only two shots were required (the prioroverhaul on the same unit had required numerous shots -three alone prior to parallel). In addition, HP/IP efficiencyincreased about two points.

• Since releasing the unit for normal operation, no forcedoutages or curtailments have been encountered due eitherto the NOx/air preheater work performed by B&W or thesteam turbine/generator work performed by the OEM.However, some corrective work was required on the boilerduring startup testing and tuning - most on components notaddressed by B&W during the overhaul.

In achieving these results, the importance of Managementsupport and commitment cannot be overemphasized. It is onething for companies to say they are going to “partner”, sharerisk, and in so doing be required to trust each other’s judgment.It is quite another to “put your money where your mouth is”and do it. To a significant (but certainly not perfect) degree,this is what occurred. PG&E management committed to con-ducting this project in a totally new manner at a time of majorchanges and uncertainty for the company, and gave the projectthe requisite support and freedom of action.

During the project’s execution, B&W management was sub-jected to significant organizational changes. These and otherinternal pressures resulted in some differences of opinion withintheir Management on how best to execute a partnering contract(and thus share risk) where the customer is privy to “all” of thedetails, with the largest issues raised by the construction arm.To a reasonable degree these issues were resolved over the du-ration of the contract as both the organizational changes werefinalized and the credibility of the project team grew.

As previously noted, Management support and guidance werealso provided in a “hands on” form via the use of the quarterlyEOC meetings. These meetings - attended by key executives andmanagers from both companies - provided a forum where the projectmanagers could review status; present problems, issues, or con-cerns for discussion; and receive immediate resolution/direction.While the vast majority of project issues were resolved at or belowthe project manager level, this forum was extremely valuable forresolving in a timely manner the few key issues that arose.

Lessons LearnedMuch went right with this project. However, there naturally

are things that things that could have been done better, with themost notable including:

Design/construction coordination: this project was reason-ably successful at involving the construction leads in the scopingand design process (and even more so in scheduling), and in theuse of integrated design teams in the electrical/controls, QA/QC, and testing areas. These efforts can and should be expandedon future projects.

It was very easy for the project to revert to the traditionalmode of “B&W scopes and executes, and PG&E reviews andinspects,” and this indeed happened at various points. How-ever, it should also be noted that the areas of the project withthe fewest problems and the greatest support from both organi-zations tended to be those that made maximum use of both inte-grated design teams (including construction involvement) andplant staff. As a result, future projects should involve key de-sign, construction and plant personnel from both organizationsin all scoping, design, constructability reviews (including oper-ating and maintenance access) and final execution.

Material quality control : a few quality control issues arosethat could have been prevented had the appropriate procurementQA/QC personnel from both organizations been integrated bet-ter into the team during project execution. The relatively minorcost required to provide this support, especially during materialprocurement/fabrication, would have more than paid for itselfin reduced field rework.

Bonus arrangement: for this project, a bonus would onlybe awarded if the Target Price was underrun, regardless of theresults achieved on the performance measures. This stipulationwas based on PG&E’s correct desire to focus the project on costmanagement. In practice, this led to adding some margins tothe Target Price buildup so that the probability of an underrun(and resultant bonus) would be increased. The lesson learnedis to choose a formula that decouples cost management incen-tives from all other incentive mechanisms, while shifting thebonus emphasis to overall job performance.

By contrast, the OEM’s turbine/generator contract includeda two-part bonus: the OEM and PG&E unconditionally splitany underrun; and an additional (and potentially more signifi-cant) bonus component was calculated based solely on theOEM’s performance against their performance measures. Thispromoted a closer alignment between PG&E and the OEM dur-ing the scoping and costing effort, resulted (as perceived byPG&E) in a tighter final estimate, and helped further focusPG&E’s and the OEM’s efforts on results.

Cost Forecasting: as a project moves forward, accurate fore-casting is critical to projecting cash flow, profitability, etc. Ona project such as this one, the utility and the supplier can oftenhave different philosophies and levels of conservatism whenpreparing cost forecasts and subsequent updates.

Page 9: Boiler Retrofit Approach and Result

Babcock & Wilcox 9

During the job, unofficial end-of-job cost forecasting wasperformed on numerous occasions and more rigorous estimateswere prepared for the EOC meetings. Because B&W manage-ment has historically viewed a forecast for a project as a com-mitment from the organization to perform no worse than thatforecast (while PG&E views them as a prediction, and not anabsolute expectation), these estimates tended to be overly con-servative, especially during the construction phase. This con-servatism made it difficult to prepare meaningful forecasts,making cash flow management more challenging than was nec-essary. Future jobs should have the ability to prepare forecaststhat do not have this expectation associated with them.

Risk Sharing: as previously noted, some reservations rela-

tive to risk sharing existed, especially in the early stages of theproject. These reservations persisted until the credibility of thisbusiness arrangement developed. With that credibility estab-lished, future project teams should act more aggressively indefining scope and cost, in setting cost and risk sharing agree-ments, and set even more substantial goals and objectives thanwere established for this first partnering endeavor.

Phase 1 duration: the initial SCR scoping effort extendedover a 6+ month period. While this duration is understandablefor defining and documenting all of the technical and commer-cial details of this initial partnering contract, a three to five monthduration (length dependent on project scope) is practical andshould be the goal for follow-on projects.