Why Estimate Classification Exists
Every capital project carries a number — the CAPEX estimate — that drives sanction, financing, scope decisions, and contractor selection. That number is meaningless without a stated accuracy and a stated basis. The AACE International classification system, codified in Recommended Practice 18R-97 for the process industries, defines five classes of estimate from Class 5 (concept) to Class 1 (definitive) based on the maturity of the project definition, the methodology used to build the estimate, and the expected accuracy range.
Classes are not a measure of competence. A Class 5 estimate prepared by a world expert from a half-page concept will still be a Class 5 estimate — accuracy −30% to +50% or wider — because there is not enough definition to support a tighter number. Pretending otherwise is the single most expensive mistake in project finance.
The Five Classes in Plain Terms
| Class | Typical use | Definition maturity | Range |
|---|---|---|---|
| Class 5 | Concept screening | 0–2% | −30% to +50% |
| Class 4 | Pre-FEED, study | 1–15% | −20% to +30% |
| Class 3 | Budget authorisation, FEED | 10–40% | −15% to +20% |
| Class 2 | Detailed engineering | 30–75% | −10% to +15% |
| Class 1 | Bid, definitive | 65–100% | −5% to +10% |
Each class implies a methodology. Class 5 is capacity-factored and stochastic; Class 3 is semi-detailed with major equipment quoted and bulk quantities allowed; Class 1 is detailed, with firm vendor quotes, take-off quantities, and contractor pricing. The classes are not interchangeable with each other or with confidence intervals from probabilistic ranging.
Where the Process Diverges for Brownfield
The classification system was developed primarily on greenfield experience and translates imperfectly to brownfield work. The differences matter.
Greenfield estimates improve linearly with definition. As the project progresses through pre-FEED, FEED, and detailed engineering, the unknowns retire in order — equipment list firms up, P&IDs mature, plot plan settles, vendor quotes harden, contractor pricing becomes available. The class progresses cleanly from 5 to 1.
Brownfield estimates improve unevenly because the dominant unknowns are not in the new scope — they are in the host facility. The questions that drive brownfield cost are:
- What is the condition of the existing pipework, structure, and equipment that we plan to reuse?
- What is the host facility's outage calendar and what does it cost to move?
- Where are the tie-in points and what is hot-tappable versus shutdown-only?
- What constraints does operations impose on construction access, working hours, and lay-down space?
- What is in the ground that the original drawings do not show?
None of these are resolved by the standard engineering progression. They require survey work, operator workshops, archival research, and sometimes intrusive investigation. A brownfield project that has completed FEED on the new scope but not yet resolved its tie-in strategy and outage plan is not a Class 3 — it is still a Class 4 with a thin veneer of Class 3 on the bolt-on portion.
The Contingency Question
Contingency is the named line item that covers identified-but-unquantified risk. It is not slush, it is not optimism padding, and it is not a substitute for definition. The contingency at each class should be sized to absorb the expected variance at that class — typically 25–40% of base cost at Class 5, falling to 5–10% at Class 1.
The trap is that contingency is often the negotiation lever in project sanction. A sponsor sees a Class 3 estimate with 20% contingency and asks for the contingency to be cut to 10% — not because risk has been retired, but because the headline number is uncomfortable. Acquiescing produces an estimate that is Class 3 in methodology and Class 5 in residual exposure. The estimator's job is to defend the contingency or, if it must be cut, to refuse to call the estimate Class 3.
A separate line — escalation, or general allowance for undefined scope — is sometimes added on top of contingency. The two are different and should not be merged. Contingency covers known unknowns within the defined scope; escalation covers scope creep and price inflation outside the defined scope.
What Drives Cost Differently Brownfield vs Greenfield
The cost composition is not the same.
Greenfield projects are dominated by major equipment, bulk piping, structural steel, and electrical bulks. Labour productivity is high because the site is unconstrained. Risk is concentrated in long-lead items, currency exposure, and the schedule itself.
Brownfield projects shift cost towards:
- Labour productivity — working alongside live operations cuts productivity by 30–60% relative to a greenfield equivalent
- Tie-ins — hot taps, shutdown tie-ins, and the engineering and execution discipline they require
- Demolition and modification — removing the existing before installing the new, often within tight constraints
- Surveys and as-built reconciliation — a real and frequently under-budgeted line for any brownfield modification
- Operator interface costs — permits-to-work, shutdown attendance, simultaneous operations management, contractor escort
A brownfield estimate that does not have separate lines for these categories is not classified correctly, regardless of what label it carries.
How to Progress Classes Cleanly
Moving from one class to the next is an engineering activity, not a calendar event. The checklist for moving from Class 4 to Class 3, for example, includes:
- Process flow diagrams complete, with major equipment sized and listed
- P&IDs complete to a defined revision level (usually IFA or IFD)
- Major equipment datasheets and budget quotes obtained
- Bulk material take-offs at the level supported by P&ID maturity
- Plot plan and 3D model maturity sufficient for steel quantities and piping bulks
- Construction methodology and contracting strategy defined
- For brownfield: tie-in strategy fixed, host outage agreed, surveys complete
If any item is incomplete, the estimate is Class 4 — even if every other box is ticked. Calling it Class 3 anyway misleads the sponsor and erodes the estimator's credibility when the variance lands.
Conclusion
CAPEX estimating is one of the highest-leverage activities in any project: the number sets the sanction, the financing, the scope envelope, and the contractor strategy. The AACE classes are a discipline for being honest about what the number can and cannot support. A correctly classified Class 4 estimate that costs the sponsor a slower sanction decision is far more valuable than an aspirationally labelled Class 3 estimate that costs the project tens of millions when the variance materialises. In brownfield work, the discipline is more important and the failure modes are richer — but the principle is the same.
