Upload
nilanga123
View
213
Download
0
Embed Size (px)
Citation preview
8/11/2019 22 Cost Estimating for Turnarounds
1/8
Cost estimating for turnarounds
Turnarounds are major eventsfor re neries and other petro -chemical facilities. Theytypically cost signi cant sums ofoperational expense (opex) and capi -tal expense (capex) money to
execute. They cause lost opportunitycost through lost production whilethe facility is shut down. If poorlymanaged, turnarounds offer signi -cant risk of accidents. If poorlyexecuted, they can also be the causeof signi cant production disruptionafter startup, due to leaks and otherproduction trips. Hence, there is apotential to save signi cant sums ofmoney, by ensuring turnarounds arerun correctly.
However, despite the large poten -tial for saving money, there has,until recently, been very little focusin the turnaround world on makingsure turnarounds are run cost ef -ciently or on nding ways toimprove cost estimating and execu -tion skills. In the past seven to eightyears, this attitude has begun tochange and turnaround teams are beginning to look at how they canimprove. One potentially rich sourceof ideas on how to improve is forthe turnaround teams to examinehow their capital project brethrenestimate and execute projects. In thecapital project world, the potentialfor leaving money on the tabledue to poor project estimating andexecution has meant that decades oftime and effort have been spent, andcontinue to be spent, on developingtechniques to ensure that estimationand execution is carried outef ciently.
This article focuses on cost estima -tion. It does not presume to provideall the answers on how to improve
C st stim t s f y tu u s c l ck ccu cy, but l ss s lf m c pit l p j ct stim ti g c ul imp v m tt s
Gordon LawrenCe Asset Performance Networks
turnaround cost estimates. Rather, itattempts to lay the groundwork fora discussion on how to improve costestimating in process facility turna -rounds. It highlights the point thatcost estimating in the turnaround
world is currently not very accurate.It then goes on to look at ideas onhow to improve cost estimate accu -racy, by looking at ideas that might be adapted from the capital projectworld. In particular, it looks at thestage gate approval system, thedevelopment of scope and estimate bases for different levels of estimateaccuracy, and at the development ofallowances for known unknownsand contingency for unknown
unknowns. Finally, it makes somesuggestions for the next steps indeveloping better turnaround costestimates.
wh t is p c ss i ust ytu u ?A turnaround (in the context of theprocess industries) is de ned by theAmerican Petroleum Institute as aplanned, periodic shutdown (total orpartial) of a... process unit or plantto perform maintenance, overhauland repair operations and to inspect,test and replace process materialsand equipment. 1 Re neries andother petrochemical facilities thatrun on a continuous rather than a batch production cycle must, everyfew years, shut down operations toprovide access to the productionunits in order that essential mainte -nance, modi cation and inspectionwork can be carried out that couldnot be done while the units are in
operation.Turnarounds are events that areplanned well in advance and
typically take place on a four-to-sixyear cycle. The length of a typicalturnaround execution phase (ie, theperiod when the facility is shutdown and hydrocarbon free) isusually around three-to- ve weeks.
The scope of a turnaround typi -cally includes: Inspection of equipment tocompany regulations or governmen -tal rules Inspection of pipework for corro -sion and erosion damage, bothinternal (process weak points) andexternal (corrosion under insulation,or CUI) Cleaning, repair and maintenanceof equipment, pipework and instru -
mentation (pulling and cleaning heatexchanger tube bundles, repairingleaks in pipework or checking ofpressure relief valves) Minor upgrades and modi cationsto the facilities (items controlledunder the management of change[MoC] procedures) Tie-ins for capital projects.
Hist ic lly p stim ti x cuti f ci cy
Historically, there was a tendencyamong operating companies to viewturnarounds as an inevitable andnecessary evil, and to accept thatthey would cost what they costand take as long as they take.However, in recent years, there has been a growing recognition that thisattitude leaves money on the tablein the form of the opportunity costof lost production while the facilityis shut down for longer than neces -sary; and unnecessarily excessive
expenditure of money during theturnaround, through running theturnaround inef ciently.
www.eptq.com PTQ Q1 2012 33
Reprinted fromPetroleum Technology QuarterlyQ1 2012www.eptq.com
8/11/2019 22 Cost Estimating for Turnarounds
2/8
www.eptq.com PTQ Q1 2012 35
results in a 50% accuracy estimate,the following conceptual designstage results in a 30% estimate, andthe nal basic design stage results ina 10% estimate. Hence, the focus ofthe stage is in developing a scope basis to meet the estimate basisrequirements for the level of accu -racy intended The project cannot proceed to thenext stage without rst goingthrough a stage gate review to checkwhether it has achieved the requiredscope and estimate basis Full funding of the project isusually only received at the end ofthe third, basic design phase.
The focus, therefore, in the capitalproject world is on calculating Howmuch money do we need in order tocarry out this scope?.
Tu u sIn the turnaround world, there has been a movement in recent yearstowards a form of stage gated
approach, in an attempt to improvefront-end de nition. However, thesystem is not yet as clearly de nedas in the capital project world.
There are some key differences inthe way the approach is applied.The length of each stage is (more orless) xed, based on the amount oftime remaining until the turnaround
In order to improve turnaroundef ciency, there is a need to examinethe estimation of costs for turna -rounds, the estimation of scheduletime, and the ef ciency of planningand execution of the turnaround.This article focuses on the rst ofthose three areas, the estimation ofcosts.
accu cy f tu u c ststim t s
A cost estimate needs to be accuratein order to provide managementwith the information needed todecide how to proceed, to allowcash ow planning and to aid in
rm control of expenditure.In a survey of conference partici -
pants at the Turnaround IndustryNetworking Conference (TINC)
Europe, held in March 2011 inAmsterdam, The Netherlands, 2 83%of respondents said that their turna -round control budget was intendedto be a 10% estimate (see Figure 1). 3 The remaining 17% said that theirestimate was supposed to includesuf cient contingency/reserve thatit was a not to exceed number.None said that their budget wasintended to be a 30% estimate.
We then took a sample of 93 recent
turnarounds and in each casecompared the total actual cost forthe turnaround (including both capi -tal and maintenance work) with thecontrol budget. Figure 2 plots theover- or under-run of the actualcosts, expressed as a percentage ofthe control budget. It shows themean (the horizontal bar) and 80%con dence range (the vertical bar) ofthe results in our sample set andcompares it to the results we wouldhave expected if the cost estimatestruly were 10%, or if they were30%.
Based on the results shown inFigure 2, we can say that turnaroundcost budgets tend, on average, tounder-estimate the actual cost byaround 16% (the distance of themean bar from the 100% point)and the budgets themselves show avariability of closer to 30% accu -racy than their purported 10%accuracy (the vertical line shows the
80% con dence range around themean).Clearly, there is an opportunity to
improve cost estimating accuracy inturnarounds.
C mp i g c pit l p j cts ithtu u sCost estimating of capital projects inthe process industries has gonethrough a number of improvementsover the years. It may well be thatsome of the lessons and devicesused in capital projects can providea basis from which to develop asystem for better turnaroundestimates.
St g g t sC pit l p j ctsIn the capital project world, it is nowgenerally accepted that front-endde nition is the key to a successfulproject. In order to achieve goodfront-end de nition, capital projectteams have largely moved toward astage gated approach, whereby thedesign proceeds through threedistinct stages of ever-improving
scope de nition.4
Key points are: The length of each stage is a func -tion of the work required to becompleted in that stage The culmination of each stage is acost estimate, with the estimatesgetting more accurate with eachsucceeding stage. Generally speak -ing, the feasibility stage of a project
120
160
170
150140
130
110
100
90
80
70
Theoretical range
10% estimate
Theoretical range
30% estimate
Turnaround
actual costs
P90
P10
Mean
N=93
A c t u a l
t u r n a r o u n
d c o s t s a s a
p r o p o r
t i o n o f
t h e
b u
d g e
t , %
60
Figu 2Accuracy of turnaround control estimates: expectation and reality
Figu 1 Expectation of accuracy in turnaround control estimates: vox populi
8/11/2019 22 Cost Estimating for Turnarounds
3/8
starts. Generally, by about 1412months before the turnaround, the
rst review of status will be held. Asecond review may be held aroundnine-to-six months before and thelast review about three-to-twomonths before.
The stages are not focused onscope de nition for cost estimatesof a given accuracy. Instead of theissuance of an estimate marking theend of a stage, the ends of thesestages are delineated by their prox -imity to the start date of theturnaround. There are guidelineson what level of scope de nitionshould be expected at each of theseturnaround stage gates, but theseguidelines are not as universallyadopted as their counterparts areon capital projects.
By the nature of a turnaround,refusing to let a turnaround teamproceed to the next stage because itis insuf ciently prepared at the timeof the stage gate review is notusually a feasible option. The budgettends to be xed back at the 30%or even 50% estimate stage.Consequently, with turnarounds, thefocus tends to be on How muchscope can we carry out for thisamount of money? rather than the
capital project paradigm of Howmuch money do we need in order tocarry out this scope?.
Sc p b sis stim t b sisIn this section, we make a distinc -tion between the scope basis (thelevel of de nition of the designdeliverables and documents) andthe estimate basis (the methodologyused to calculate costs, based onthose deliverables).
C pit l p j ctsIn the world of process industrycapital projects, the scope basisand the estimate basis are bothrelatively well de ned for variouslevels of estimate accuracy, from50% down to 10% or better. Onewell-known example of where thisis documented is AACE InternationalRecommended Practice No. 18R-97. 5 Lawrence also lists typical require -ments. 4 (Table 1 and Table 2 are
adapted from Lawrence, 2008.) Inaddition, it is accepted practice todocument the cost estimate in a
www.eptq.com PTQ Q1 2012 37
written basis of estimatedocument.
e ly stim t : 50% ugh fm g ituIn capital projects, the rst cost
estimates are usually highly deter -ministic, using high-level factoringmethods rather than the stochastic
approach of calculating from the bottom up. The cost of a facility istypically factored off such details as,for example, the production capacityof the new facility, with adjustmentsmade for in ation and location. Such
estimates generally have little interms of scope basis, often merelyan outline of the facility capacity
T bl 1
50% ROMestimate
General project data
Engineering deliverables
Project scopedescription
Facility capacity
Facility location
Ground surveys
Projectexecution plan
Contract strategy
Project schedule
Cost estimating plan
Plot plans
Process flow diagrams
Utility flow diagrams
Project execution plan
P&IDs
Heat and materialbalances
Processequipment list
Utility equipment list
Electrical single linediagram
Process engineersequipment datasheets
Mechanical engineers
equipment datasheetsEquipment generalarrangement
Block flow diagrams
30% ROMestimate
10%control
estimate
General
General
None
None
None
Outline
None
None
None
NoneNone
None
None
None
None
None
None
None
None
Assumed
Assumed
Level Imilestones
Defined
Defined
Defined
Defined
Defined Defined
Defined
Level III detailedresourceloaded
schedule
Defined
Defined
Defined
Defined
Specific
Complete
Complete
Complete
Complete
Complete
Complete
Complete
Complete
Complete
Complete
Complete
Complete
Complete
Complete
Complete
Specific
Preliminary
Preliminary
Preliminary
Preliminary
Preliminary
Preliminary
Preliminary
Preliminary
Preliminary
Preliminary
Preliminary
Preliminary
Level IIpreliminary
8/11/2019 22 Cost Estimating for Turnarounds
4/8
the previous turnaround scope, andthe estimate basis uses that historicaltotal cost.
C c ptu l stim t : 30%On turnarounds, teams appear toaim to have the overall scope frozenand challenged by this point.However, this has not always beenachieved. In terms of deliverabledocuments for the scope basis, themechanical workpacks are usuallydeveloped. The estimate basis is amix of deterministic and stochastic,as the mechanical work packs arecosted (in-house) in some detail, butthe other work (electrical, instru -mentation scaffolding, insulation,and so on) is usually factored off themechanical costs, using historicalfactors.
auth is ti /c t l stim t : 10%Some turnaround teams do nowhave all discipline scopes developedand externally costed. However, inour experience, the number of teamsthat take their estimates to this levelof detail are relatively few, despitethe fact that most teams claim theircontrol estimates are 10% or better.
w k b k st uctu
C pit l p j ctsAll companies have their own varia -tions on how they prefer to breakdown costs within the estimate.Indeed, the work breakdown struc -ture (WBS) may differ slightly, fromproject to project, depending on thescope. Nevertheless, at a high level,there is a measure of agreementacross the industry on how to breakdown capital project costs. The breakdown tends to be as shown inTable 3.
Tu u sAt present, there is no clear, commonagreement across the process indus -tries on a high-level WBS forturnaround costs. Most turnaroundteams split costs into direct costs costs that can be speci cally assignedto a cost object, such as an item ofexisting plant and indirect costs costs that cannot easily be directlyattributed to a cost object. However,
the de nition of what should beconsidered direct or indirect costvaries from site to site (presumably
38 PTQ Q1 2012 www.eptq.com
required. The estimate basis relieson historical data of the total cost ofthe entire facilities rather thandetailed cost data.
C c ptu l stim t : 30%In capital projects, a considerableamount of work would be completedon the scope basis, coming close tofreezing the overall scope and devel -
oping many of the design drawings by the time this estimate wasprepared.
The estimate basis would havemoved from a high-level, determin -istic estimate to a mix of thestochastic and deterministic, and theteam moves towards Lang factors orother, similar methods to calculatethe project cost (factoring off oneelement of the project scope, mosttypically the equipment cost).
auth is ti /c t l stim t : 10%By this stage, the scope basis on acapital project is generally verydetailed, with completed P&IDs,order speci cations for major equip -ment and so forth. The estimate basis is now stochastic, built bottomup from the material take-offs andusing market cost data rather thanhistorical database information.
Tu u s
There appears to be no documentequivalent to AACE InternationalRecommended Practice No. 18R-97
for process industry turnarounds. Inthe absence of such a document, wehave informally surveyed a numberof turnarounds to look at the vari -ous scope and estimate bases used by the turnaround teams to developestimates of nominal accuracies of50%, 30% and 10%.
The rst problem we have encoun -tered is that there is some variation
in the methods used. Second, theconcept of writing down a detailedbasis of estimate document issomething that still has to take rootin the turnaround world; hence,understanding how each estimatewas developed is dif cult. (We wereoccasionally given documentsdescribed by the team as the basisof estimate, but these were gener -ally extremely simple, high-levelnotes on the estimate spreadsheetitself rather than a formal anddetailed document.) Nevertheless,we have attempted to provide ageneral overview of the methods being used by turnaround teams forthe various estimates.
e ly stim t : 50% ugh f m g ituThe early estimate is frequentlymerely the cost of the last turna -round, adjusted for in ation, withperhaps an educated guess as to
how the scope may have grown orshrunk since the last turnaround.The scope basis is therefore merely
T bl 2
50% ROMestimate
30% estimate 10% controlestimate
8/11/2019 22 Cost Estimating for Turnarounds
5/8
because of different internal account -ing practices).
The costs of owner supervision aregenerally agreed to be indirect, sinceat any hour of the day a team could be supervising work on a range ofexisting plant areas. Contractedlabour on tasks that can be assignedto speci c existing plant items isgenerally categorised as direct cost.Furthermore, tracking this cost byplant unit/area and by equipmentclass/type and by equipmentnumber is relatively common.
However, costs such as rentedequipment (cranes, heat exchanger bundle pullers, and so on) are lessconsistently assigned to either director indirect categories, presumably because such items can only some -times be attributed to a speci c plant
item (for instance, if a crane is onlyneeded to work on a speci c item).
Theoretically, it should be possibleto assign bulk material (pipe, valves,for instance) to speci c plant itemsand hence categorise the material asdirect. However, this is not alwaysdone. One would logically expectconsumable materials such as weld -ing rods to be indirect costs becauseof the dif culty of assigning them tospeci c plant items. However, the
situation can be clouded by the fact
40 PTQ Q1 2012 www.eptq.com
T bl 3
Home Office
Field
Capital
Expense
Allowances forunknown unknowns
Allowances forknown unknowns
Direct
Design
Engineering
Equipment
Contingency
Commissioning
Operations staff
Bulk materials
Construction labour
Constructionmanagement/supervision
Temporary facilities,craneage, scaffold, etc
Designallowances
on eachline item
Escalationallowances
on eachline item
Projectmanagement/supervision
Indirect
that consumables and bulk materialsoccasionally are not separated.
all c s f u c t i tyAny cost estimate, by nature of it being an estimate, requires funds forthe uncertain elements of the scope.These uncertain elements can bedivided into known unknownsand unknown unknowns. Theknown unknowns are allowancesfor items or quantities that the esti -mator knows, from historicalevidence, will be needed in a speci cline item of the estimate. Theunknown unknowns are the itemsor quantities that were under-esti -mated or completely overlooked inthe estimate. The estimator does notknow exactly in which line item thefunds will be needed, but history
tells the estimator that some fundswill be needed to bring the estimateto the P50 point. These unknownunknowns are contingency funds.The need for contingency is a hotlydebated subject, but has beendiscussed elsewhere 6 and we do notpropose to explore this topic here.
C pit l p j ctsK u k sThe known unknowns in capital
projects take the form of design
allowances allocated to each lineitem for the known unknown ofthat line item (for instance, wastagein the piping material take-off,design development in equipment).
The amount of design allowanceto include is usually determinedthrough expert judgment based onhistorical data.
U k u k sContingency is for the unknownunknowns in the estimate (forinstance, it is discovered duringconstruction that the need for anelectrical transformer was over -looked in the design).
As discussed by Burroughs & Juntima, 7 the amount of contingencyto include is typically calculated inone of four ways: As a predetermined percentage(of the base estimate) Through expert judgment (gutfeeling/experience) based on histori -cal data Through a Monte Carlo-typereview of the cost effect of expectedrisks Through use of a regressionmodel, based on historical data.
Tu u s
In the case of funds for uncertainty,there is a little more general agree -ment than perhaps in other areas ofestimating for turnarounds, in asmuch as it is generally accepted thatfunds are needed for unknownitems.
K u k sIn turnarounds, the knownunknowns are (a) emerging work(work to repair items that have broken/failed between the comple -tion of the estimate and the start ofthe turnaround); and (b) discoverywork (repair work that is discov -ered during the turnaround onceequipment or pipework is openedup for detailed inspection).
In common with the design allow -ances on capital projects, estimatesof likely emerging work and discov -ery work should (in theory) becalculable and assigned to speci cestimate line items. (For instance, we
can expect other X valves to fail between now and the shutdown, or based on past experience we should
8/11/2019 22 Cost Estimating for Turnarounds
6/8
expect that the trays in column Zwill be discovered to be damagedwhen we open it up.)
However, what we have observedis that very often such items are notreported in the estimate to this levelof detail. Instead, they tend to belisted as single line items on theirown. In addition, anecdotally, weobserve that very often the discov -ery work in particular appears to beunderestimated an optimismabout what is likely to be foundseeming to prevail.
The amount of emerging workallowance and discovery workallowance appears to be calculatedon the basis of experience fromprevious turnarounds or from inti -mate knowledge of the state of thefacility. But when it is based on
experience of previous turnarounds,this experience is rarely backed up by hard numbers, showing whatwas spent in previous turnaroundsas a percentage of the base scopecost.
U k u k sIn turnarounds, just as in capitalprojects, a contingency for the itemsthat were overlooked or completelyunforeseen at the time of the esti -
mate issue is required. (For instance,money was included in the discov -ery work allowance to spend threedays repairing trays in column X, because the column was old andwas operating at low throughput.However, it took two days longerthan expected to repair the trays incolumn X. That extra two days costcomes out of contingency.)
Whereas in capital projects contin -gency is calculated on the basis ofone of four methods, each of whichhas at least some logical foundationto it, in turnarounds we observe thatcontingency often appears to becalculated simply as the difference between the current base estimateplus discovery and emerging workallowances and the original, roughestimate quoted to management(assuming the rough estimate to bea higher number!).
Qu ti c ti f tu u
ll c sWe then looked further into howmuch allowance for uncertainty was
included by turnaround estimators
and how much was actually used.
H much is i clu ?The rst problem that we encoun -tered in this study was that whilemost turnaround teams recognisethe three main areas of uncon rmedscope that need funding as being (1)emerging work, (2) discovery workand (3) contingency for the unfore -seen, few teams clearly differentiate between them in their estimates.
Indeed, in a survey that we carriedout, out of 216 turnarounds, only 34(or 16%) could give a clear differen -tiation in their cost estimate betweenthese three areas and the mainestimate.
Nevertheless, from the data gath -ered we are able to show (see Figure3) that teams typically include a total
of just under 13% of the base esti -
mate (where base estimate = thetotal cost estimate before the threeuncertainty allowances are included)as a total of all three allowanceamounts.
That funding breaks down in turninto an average of 16% for emergingwork, 38% for discovery and 46%for contingency (see Figure 4).
H much is sp t?We then attempted to look at how
much in each category is typicallyconsumed. From this information, wehoped to begin developing guidelineson calculating how much should beincluded in an estimate. However,we found that although some teamshave begun to track the consumptionof the three elements in their actualcosts this practice is still immaturewithin the turnaround community.
Without historical data on actualexpenditure in these three catego -ries, it is impossible to close theloop and begin developing guide -lines on how much should beincluded in each category in anestimate.
C clusi sTh p bl mCurrently, turnaround cost estimatesare inaccurate in that, on average,turnaround costs overrun their esti -mates. They are also highlyunpredictable in the level of their
inaccuracy (that is, the overrundoes not remain within the claimed10% accuracy of most estimates).
16
22
20
18
14
12
10
8
6
T o t a l a l l o w a n c e s a s p r o p o r
t i o n
o f t h e
b a s e e s t
i m a t e ,
%
Emerging work + discovery work+ contingency
4
+1SD
1SD
Mean
N=38
Figu 3Uncertainty allowance included in estimate
Contingency38%
Emergingwork 16%
Discoverywork 38%
N=34
Figu 4 Breakdown of uncertaintyallowance into emerging, discovery andcontingency
www.eptq.com PTQ Q1 2012 41
8/11/2019 22 Cost Estimating for Turnarounds
7/8
Furthermore, there is no clearagreement on the de nition ofwhat information a cost estimateshould be based upon in order toallow it to be de ned as a 10%estimate.
There are no guidelines on how tocalculate uncertainty allowances foremerging work, discovery work andcontingency. Nor are many teamsgathering the actual cost expendi -ture data that would allow suchguidelines to be developed in thefuture.
L i g f m c pit l p j ctsThe capital project world historicallyhad similar problems with cost esti -mating. Over the years, it hasimproved through: Tying estimates to stage gates Clearly de ning the scope basisand estimate basis required for acertain level of estimate accuracy Using similar work breakdownstructures, allowing comparison benchmarking of data Gathering data on allowances foruncertainty and developing logicalmethodologies for estimating thoseallowances.
The turnaround world has already begun to learn from capital projects
by copying the stage gate approach.There seems to be little reason tosuppose that turnaround teamscannot continue to learn from capitalprojects in order to develop turna -round-speci c standards for scope basis, estimate basis and calculationof allowances.
r c mm ti s f imm i tcti
If the estimation of turnarounds isto become more accurate andpredictable, we recommend that, asa rst step, a series of actions should be adopted within the turnaroundcommunity.
d th sc p stim t b sisThere should be wide agreement onwhat the scope and estimate basisis for calling an estimate 50%, 30%or 10% accurate. The capital projectscommunity has such agreement indocuments referred to earlier.
Suggested outline versions for turn -arounds are given in Tables 4and 5, respectively.T bl 4
42 PTQ Q1 2012 www.eptq.com
50% ROMestimate
General turnaround scope data
Engineering deliverables
30% ROMestimate
10%control
estimate
Equipment count
Turnaround premises
Risk Based Inspection(RBI) data
Maintenance turnaroundbacklog
Listing littledefinition
Baseline
no markup
None
None
None
NoneNone
None
Originalbaseline
None
Listing highlevel, based
on assetmanagement
strategy
95%complete
99%complete
60-70%complete
99%complete
Hazard and Operabilitystudy (HAZOP)
Listing highlevel
60-70%complete
100%complete
Plant engineering designchanges (expense)
Listing notall inclusive
60-80%complete
100%complete
Leak Detection AndRepair (LDAR)
Listing preliminary
Listing preliminary
Partial
markup
Partial
turnaroundmarkups
Partial fornew capital
Completelisting
Previous turnaroundscope
Plot plans
logistics markups
Plant engineeringdesign packages
Major capital projectengineering design
Flare minimisation plan
Heat transfer data
Utility plan
Oils plan
P&ID markups
Process engineeringdesign basis for plant
engineering capital
Turnaround boundaries
Listing 90%Listing 90%Listing 90%
Process Hazard Analysis(PHA)
Listing highlevel
60-70%complete
100%complete
Management Of Change(MOC) design changes
40-60%complete
100%complete
Complete
Complete
Complete
Complete
Complete
Complete
Complete
Complete
Preliminary
100%complete
100%complete
80-90%complete
100%complete
Preliminary
80-90%complete
100%complete
Capital projects 80-90%complete
100%complete
Listing
Listing
Listing variousstages
Partial
Partial
Partial
None CompletePartial
None CompletePartial
Preliminary
Partial
NoneStartup procedures
maintenance(new equipment)
CompletePartial
Baseline none for new
capital
Original EquipmentManufacturer (OEM)
specifications & drawingsComplete
Preliminary
Plant projects
8/11/2019 22 Cost Estimating for Turnarounds
8/8
St is th wBSMove towards a common standardlayout for cost estimates. The capitalprojects community has such astandard (see Table 3). A suggestedoutline version for turnarounds isshown in Table 6.
T ck th u c t i ty ll c sIf the turnaround community isever to develop even rudimentaryrules of thumb for estimatingemerging work, discovery workand contingency for the unforeseen,it must begin to track expenditurein these areas. This will require tworesponses from the community: awillingness to differentiate clearlythe three categories in cost esti -mates; and a willingness toaccurately track actual expenditure
from these three categories duringa turnaround.
Ti th stim t s t th st g g t sIn the world of capital projects, thestage gate culminates in a cost esti -mate. Currently, the typical reviewstages for preparation and planningof a turnaround are not tied to a costestimate. Tying the estimate andstage gate together could improvereview of planning and preparation
progress.r c mm ti s f l g -t m
xt st psThis article covers only the rst stepsin improving the accuracy andpredictability of turnaround costestimates. Further steps need to betaken. Once there is common agree -ment on the minimum standard ofscope basis and estimate basis forturnaround estimates, and onceturnaround teams begin gatheringactual uncertainty allowance data,the community will then be in aposition to begin developing logicalmethodologies for calculating moreaccurately the required level ofallowance for uncertainty in emerg -ing work, discovery work andcontingency for each turnaround.
ack l g m tsThe author would like to acknowledge theencouragement offered during the preparation
of this article by Bobby Vichich, Vice President,Turnarounds, at Asset Performance Networks,as well as the invaluable advice (particularly,
but not only, relating to Table 4) of John Casper,Senior Consultant and Subject Matter Expertfor Turnarounds and Capital Projects, at AssetPerformance Networks.
r f c s1 http://www.api.org/aboutoilgas/sectors/re ning/re nery-turnaround.cfm.2 http://ap-networks.com/latest-news/tinc-europe-2011-summary.html.3 We are very aware of the fact that in recent years the preference within project costestimating circles is not to talk of a typical10% or 30% estimate, but rather to talkof classes of estimates. This is because ofthe need to recognise that for the same levelof scope de nition, two projects could havedifferent levels of estimate accuracy becauseof the inherent risks in their characteristics. Forexample, a new technology project probablycarries more risk than a similar-sized projectexpanding an existing plant. However, the
turnaround world has not yet reached thatlevel of sophistication. Hence, for the survey,we used terms that are readily recognised, suchas 10% estimate.
T bl 6
4 Lawrence G R, Stage gated approval processes a practical way to develop and lter capitalinvestment ideas, Pharmaceutical Engineering,vol 20, No 2, March/April 2008.5 AACE International Recommended PracticeNo. 18R-97; cost estimate classi cation system,as applied in engineering, procurement, andconstruction for the process industries.6 Lawrence G R, The use and misuse ofcontingency or why cutting contingencymakes projects more expensive, not less.Pharmaceutical Engineering , vol 27, No 5, Sept/Oct 2007.7 Burroughs S E, Juntima G,Exploring Techniques for Contingency Setting , AACE InternationalTransactions, Washington DC, 2004.
G L c is a Senior Consultant withAsset Performance Networks, Amsterdam.He holds a bachelors degree in chemicalengineering from Heriot-Watt University,Edinburgh, a masters in biochemical
engineering from Birmingham University, andan MBA from Strathclyde University BusinessSchool, Glasgow.Email: [email protected]
www.eptq.com PTQ Q1 2012 43
Mechanicalwork
Other disciplines
Cost of the lastTA, adjusted for
inflation
Quantities formechanical workfrom workpacks,but then priced
in-house
Other disciplinesfactored frommechanical,
based on history
Quantities for alldisciplines
calculated andpriced by
contractors,based on
workpacks
50% ROMestimate
30% ROMestimate
10% controlestimate
Home Office
Field(during
shutdownperiod)
Allowances forunknown unknowns
Allowances forknown unknowns
Direct(broken down by
unit/area, equipmenttype and equipment
number
Field (before theshutdown period)
Emergingwork
allowanceson each line
item
Discoverywork
allowanceson each line
itemIndirect
Contingency
Consumables
Equipment
Bulk materials
Turnaround planningand preparation
Turnaround fieldlabour
Pre-turnaround fieldwork
Equipment rental cranes, scaffold, etc
Logistics, temporaryfacilities, etc.
Turnaroundmanagement/ supervision
Direct
Indirect
Maintenance
Capital projects
T bl 5