In Part 1 of this series, I made the case that most end-of-FEED cost estimates don’t fail because of bad estimating technique. They fail because the engineering input that fed into them was incomplete, out of date, or handed over without the context the cost team needed to use it correctly. The Basis of Estimate documents the foundation of every number in the sanction estimate, and that foundation is built from discipline engineering inputs.
This post is the practical companion. It’s written for lead and senior engineers across piping, civil, structural, mechanical, electrical, instrumentation who are responsible for submitting quantification inputs at FEED closeout. Where a missed input has a cascading effect across other disciplines or into construction cost, I’ve called it out explicitly, because those are the ones that matter most.
I’ve reviewed estimate packages and participated in cost reviews on enough international projects to know that the problems are rarely caused by bad estimating technique. They’re caused by poor engineering input — delivered late, missing critical caveats, or disconnected from the actual state of the design. Let’s fix that.
What a Basis of Estimate Actually Contains
Before going into discipline-specific detail, it’s worth being clear on what the BOE is and what it needs to cover. It’s not a summary of the estimate numbers. It’s the document that explains how every number was arrived at, what it includes, and what it doesn’t. Under AACE RP 34R-05, the BOE is formally identified as the one deliverable that defines the scope of the project and ultimately becomes the basis for change management throughout execution. Every scope variation, every change order, and every cost recovery claim in EPC will be measured against what the BOE says was included at sanction.
A well-structured BOE for an end-of-FEED Class 3 estimate typically covers the following:
The purpose statement establishes the estimate classification, project type, facility capacity, plant location, and overall timing. The design basis lists every source document used, by specific revision number and date e.g. equipment lists, plot plans, PFDs, P&IDs, line lists, piping material specifications, instrument index, single line drawings, structural layouts, foundation plans, and any estimate sketches used where formal drawings don’t yet exist. The methodology section explains how quantities were derived for each discipline account, what percentage is based on vendor quotes versus factored estimates, and what design development allowances have been applied. The planning basis covers contracting strategy, modularisation approach, project execution schedule, and construction productivity assumptions including location and site-specific factors. The cost basis addresses equipment pricing sources, bulk material pricing, labour wage rates, escalation indices, and currency exchange assumptions. Contingency is documented separately, with the derivation method stated and the key risks that drove the contingency quantum listed explicitly. Finally, assumptions and exclusions bound the estimate. It is everything a reviewer might reasonably expect to be included but isn’t must be listed here, from catalyst first fills and spare parts to contaminated material disposal and owner-furnished equipment.
The BOE should also carry the names of the discipline leads who provided the input quantities for each account. This isn’t formality. It’s accountability. When your name is on the document that will be referenced for the next three to five years of project execution, the quality of what you submit tends to reflect that.
Before Anything Else: The Document Register Is Everyone’s Responsibility
Every discipline MTO, equipment list, instrument index, and quantity takeoff must reference the specific revision and issue date of every source document used. P&IDs, plot plans, equipment lists, and line lists can all change in the final weeks of FEED, with multiple revisions in circulation simultaneously. If your takeoff was done against Rev C of a P&ID and Rev D has since been issued with additional instruments or modified line classes, the cost team has no way of knowing unless you tell them. This applies universally, greenfield or brownfield, and it is the most common source of silent estimate error on every project I’ve worked on.
Now let’s take a look at major disciplines and cascading risks on other disciplines if originating discipline misses something
Piping
The piping bulk estimate is typically one of the two or three largest cost categories on any process plant. The first thing to document explicitly is the Pipe size cutoff between model-extracted quantities and P&ID-stamped quantities. PID stamped quantities are all pipe fittings as per PID, except 90deg and 45deg elbows and pipe length which are calculated based on thumb rules or transpositions made on plot plan. Large bore process piping of NPS 4 and above on main systems will generally be modelled at FEED stage. Small bore branch connections, instrument connections, vents, drains, utility stations, and steam tracing manifolds are quantified by stamping P&IDs. Both the cutoff and the stamping methodology need to be stated with reference to the specific P&ID revision used.
The weld quantification basis matters more than most piping engineers realise. Construction labour for piping is substantially driven by weld count, and the line list feeds directly into this because NDE requirements, PWHT, stress relief, and pressure test category all attach to weld cost through line data. An incomplete or outdated line list means construction labour is being priced against assumed line data rather than actual project requirements.
On greenfield projects, confirm with construction and document the hydrotest medium, golden weld philosophy, shop fabrication strategy, and modularisation scope before the MTO is submitted. On brownfield projects the picture is significantly more complex. The tie-in list must be quantified separately and its maturity stated explicitly. Decontamination and blinding of existing lines before tie-in work begins is typically owner’s scope and cost, not contractor’s. If that exclusion isn’t visible in the piping BOQ, the contractor will price it in EPC. Underground piping assumptions also need to be stated on brownfield sites where existing underground services may only be partially surveyed, assuming clean unobstructed soil without flagging the uncertainty needs an explicit allowance, not a silent assumption.
Cascading risk: Electrically traced pipe lengths must be formally communicated from the piping team to the electrical discipline. If those lengths are missing or derived from an outdated line list, the electrical heat tracing material and installation cost will be wrong. This is one of the most consistently missed cross-discipline handoffs at FEED closeout.
Mechanical Equipment
The equipment list is the backbone of the entire estimate, not just the mechanical cost account. Equipment composites, which define inter-discipline interfaces for each tagged item, drive associated structural steel, piping connections, electrical supplies, and instrumentation quantities for every piece of equipment on the list. An equipment list that hasn’t been reconciled against the current P&IDs is therefore not just a mechanical problem.
Cascading risk: Tags absent from the equipment list but present on P&IDs, or tags whose classification has changed from a process vessel to a vendor package unit mid-FEED, will cause composites to generate quantities against wrong assumptions. The impact isn’t just equipment cost. It’s foundations, nozzle piping, instrument connections, electrical supplies, and access steel. On a project with 400 to 600 tagged items, even a 3% error in list completeness is a serious cost implication.
For brownfield projects, the equipment list must explicitly distinguish between existing items being reused, items being modified, and new items. Without that distinction the estimate will either price installation of already-installed equipment, miss modification costs, or both.
Civil and Structural
Civil quantities are directly dependent on two inputs that are often not finalised at FEED closeout: the geotechnical interpretive report and the topographical survey. Pile capacities, bearing pressures, and excavation strategies all derive from geotech data. If the geotechnical study is still preliminary, that assumption must be stated explicitly in the civil input along with the uncertainty range it introduces. On greenfield projects on previously undeveloped sites this dependency is particularly significant. On brownfield sites, existing foundation data may be available but its adequacy for new or modified equipment loads still needs to be confirmed and documented.
Transportation distances to disposal sites or borrow pits, and temporary materials including shoring, dewatering, and construction fencing, are not estimated by civil and structural. Those are construction-provided inputs that must be confirmed and documented as such.
Cascading risk: The stick-built versus modular split on structural quantities must be resolved before structural finalises its MTO. Module yard fabrication and site-installed steel carry different unit rates and construction productivities. If that strategy changes after quantities have been submitted, the structural MTO needs to be re-split entirely. This is a project-level decision that must be locked before the estimate is frozen.
Electrical
The electrical estimate is highly sensitive to the revision status of single line diagrams, the load list, and area classification drawings. Main cable tray quantities come from manual takeoffs based on design layouts. Branch distribution and lighting are typically factored from equipment and instrument counts, which makes the accuracy of those counts critical. If the area classification drawing used for cable tray routing was based on an older equipment arrangement, the run lengths will be wrong.
On brownfield projects, the boundary of electrical scope needs to be precisely defined against the existing power distribution infrastructure. If it’s unresolved whether the existing MCC has capacity to accommodate new loads, an allowance must be carried explicitly. That open question left undocumented becomes a significant cost exposure.
Cascading risk: Instrument cable quantities sit in the electrical account, not the control systems account. Electrical needs the instrument index and junction box location assumptions from control systems before the cable schedule can be finalised. If junction box locations are still being determined, or if the assumed standard field cable length doesn’t reflect the actual plant layout, the quantities will be wrong and neither discipline will own the error clearly.
Control Systems and Instrumentation
The instrument index, derived from the current P&ID revision, is the primary quantification basis for control systems at FEED stage. Items within vendor package boundaries included in vendor pricing must be identified separately from items shipped loose for field installation, as these carry different installation labour profiles.
On brownfield projects, the interface between new and existing control systems is one of the highest-risk areas in the instrumentation estimate. Whether new instruments tie into an existing DCS or require a new system, and the extent of modifications to existing panels and junction boxes, are often unresolved at FEED closeout. These need to be carried as explicit risk allowances with a clearly documented basis.
Cascading risk: Control valve specifications drive the flange rating, metallurgy, and connection class for the valve’s piping connections. If those specifications are still generic at FEED stage, the associated piping bill of materials is based on assumed data. On a plant with 150 to 200 control valves, particularly on high-alloy or exotic material service lines, this is not a negligible exposure.
Demolition and Revamp Scope (Brownfield Only)
This account is consistently one of the most underestimated areas on revamp projects. Every discipline identifies its own demolition scope and codes it to this category. Piping demolition quantities should be based on a formal site survey and existing piping arrangement drawings. If a site survey hasn’t been done, that uncertainty requires a documented allowance.
Cascading risk: Decontamination and blinding of existing lines before demolition or tie-in work begins is typically owner’s scope and cost, not contractor’s. If this is not documented as an exclusion, the contractor will price it in EPC. Disposal of demolished materials must also be stated explicitly. If contaminated material requires specialist disposal and that isn’t documented at FEED stage, it appears from nowhere during construction as one of those cost overruns that nobody can trace cleanly back to a root cause.
The Signed Checklist Is Not a Formality
The formal quantification checklists that exist for each discipline are not administrative paperwork. They are the mechanism by which a discipline lead formally confirms to the estimating team that all items within their scope have been considered and reported. In design, an error caught in a model review costs hours. An error that reaches the sanction estimate costs months of cost recovery in EPC.
Greenfield vs Brownfield: Where the Estimate Risk Sits
On a greenfield project, the primary risks in estimate input relate to engineering maturity and the completeness of design development allowances applied to undefined scope. Good discipline practice, well-documented allowances, and honest flagging of open decisions are the primary controls.
On a brownfield project there is an additional layer of uncertainty that is fundamentally different in character: you don’t always fully know what’s already there. Existing underground services, structural capacities, contaminated soil, aging equipment conditions, and undocumented historical plant modifications all introduce risks that don’t exist on a greenfield site. The estimate inputs on a brownfield project need to be explicit not just about what’s been designed, but about what has and hasn’t been surveyed, verified, or confirmed in the existing facility. Every assumption about existing conditions is a potential cost risk, and every one of them needs to be visible in the Basis of Estimate.
Key Takeaways
The Basis of Estimate is the financial expression of every engineering decision made during FEED. Its core contents which are purpose, design basis, methodology, planning basis, cost basis, contingency, and exclusions, must have direct engineering input behind them. Your estimate input is a professional statement about what you know, what you’ve assumed, and what remains uncertain. Before your FEED estimate submission, sit with your discipline team, walk through major cross-discipline dependencies with the relevant leads, and review the BOQ draft for your discipline to make sure the methodology, allowances, and exclusions stated are accurate. That discipline, applied consistently, is what separates estimates that hold through EPC from those that unravel the moment scope firms up in detail engineering.
Leave a Reply