7.2 AACE 18R-97 Cost Estimate Classification System

7.2 AACE 18R-97 Cost Estimate Classification System

Quick Summary: AACE Recommended Practice 18R-97 establishes the worldwide benchmark for classifying cost estimates in the process, engineering, procurement, and construction industries. It defines five distinct estimate classes—Class 5 (conceptual concept screening) through Class 1 (definitive check estimate/tender). The sole primary characteristic that dictates an estimate's class is the Degree of Project Definition (percentage of completed engineering deliverables). All other attributes—end usage, estimating methodology, expected accuracy range, and preparation effort—are secondary characteristics that correlate directly with definition maturity. Crucially, expected accuracy ranges are asymmetric, showing higher upside risk than downside potential due to scope expansion dynamics.

---

1. Genesis & Purpose of AACE RP 18R-97

Historically, the engineering and construction industries suffered from ambiguous, non-standardized terminology. Practitioners used subjective terms such as "ballpark," "order-of-magnitude," "budget," "definitive," and "feasibility" interchangeably, resulting in severe misunderstandings between project sponsors and contractors regarding estimate reliability.

To resolve this chaos, AACE International published Recommended Practice 18R-97, Cost Estimate Classification System — As Applied in Engineering, Procurement, and Construction for the Process Industries.

RP 18R-97 establishes an objective, technical framework that:

1. Aligns the maturity of project engineering deliverables with cost estimating expectations.
2. Establishes realistic, empirically verified accuracy ranges for each project stage gate.
3. Prevents project sponsors from demanding definitive precision when engineering definition is virtually non-existent.
4. Standardizes commercial risk communication across international joint ventures and financial institutions.

While 18R-97 specifically addresses the process industries (oil, gas, chemicals, pharmaceuticals, power), AACE maintains companion classification standards: 17R-97 (general principles), 56R-08 (building and general construction), and 69R-12 (civil infrastructure).

---

2. Primary vs. Secondary Characteristics

The fundamental architectural rule of AACE 18R-97 is the strict division between primary and secondary characteristics:

+-----------------------------------------------------------------------------------+
|                    AACE 18R-97 CLASSIFICATION ARCHITECTURE                        |
|                                                                                   |
|                      [ PRIMARY CHARACTERISTIC ]                                   |
|                  DEGREE OF PROJECT DEFINITION (% Complete)                        |
|           (The SOLE metric that dictates the estimate class)                      |
|                                    |                                              |
|                                    v                                              |
|                    [ SECONDARY CHARACTERISTICS ]                                  |
|           (Correlate with and flow from project definition)                       |
|                                                                                   |
|   +-----------------------+-----------------------+-----------------------+       |
|   |                       |                       |                       |       |
|   v                       v                       v                       v       |
| [ End Usage ]         [ Methodology ]         [ Accuracy Range ]      [ Effort ]  |
| Screening / Budget    Stochastic to           80% Confidence          Preparation |
| Tender / Check        Deterministic           Asymmetric +/-          Cost Index  |
+-----------------------------------------------------------------------------------+

The Sole Primary Characteristic

• Degree of Project Definition: Measured as the percentage of complete engineering deliverables available at the time of estimate preparation.
• Governance Principle: An estimate cannot be elevated to a higher class simply because management needs it for a critical funding vote. If only 3% of engineering is complete, the estimate is a Class 5, regardless of what title management assigns to it.

The Four Secondary Characteristics

1. End Usage: The typical project phase gate or management decision supported by the estimate.
2. Estimating Methodology: The analytical technique used to compute costs (ranging from stochastic capacity-factoring to fully deterministic bottom-up line-item pricing).
3. Expected Accuracy Range: The statistical variation expected between the estimated cost and final actual cost, evaluated at an 80% confidence interval.
4. Effort and Preparation Time: The relative cost, staffing, and calendar time required to develop the estimate.

---

3. The Five Estimate Classes Detailed

AACE RP 18R-97 outlines five classes numbered 5 down to 1, representing increasing project maturity:

PROJECT DEFINITION CONTINUUM:
Class 5 ------------> Class 4 ------------> Class 3 ------------> Class 2 ------------> Class 1
[0% - 2%]              [1% - 15%]            [10% - 40%]           [30% - 75%]           [65% - 100%]
Concept Screening      Feasibility Study     Budget Sanction       Control Baseline      Bid / Check

Class 5 Estimate (Concept Screening)

• Project Definition: 0% to 2% complete engineering.
• Primary Purpose / End Usage: Strategic screening of concepts, market entry evaluation, initial business case development, and project portfolio prioritization.
• Methodology: Stochastic / conceptual methods. Uses analogous project ratios, capacity-factored power curves, and high-level parametric models.
• Engineering Deliverables: Block Flow Diagrams (BFDs), general project description, preliminary plant capacity, and gross site location.
• Accuracy Range: Low side: -20% to -50%; High side: +30% to +100% (often wider in frontier technology).
• Effort Index: Baseline reference = 1.0 (requires nominal estimating hours, typically completed in days).

Class 4 Estimate (Feasibility / Study)

• Project Definition: 1% to 15% complete engineering.
• Primary Purpose / End Usage: Feasibility analysis, screening of technology alternatives, concept selection, and preliminary project planning.
• Methodology: Primarily stochastic methods with early equipment factoring. Factored estimating using Lang factors, Hand factors, and preliminary Cost Estimating Relationships (CERs).
• Engineering Deliverables: Preliminary Process Flow Diagrams (PFDs), heat and material balances, preliminary major equipment lists with rough sizing, conceptual plot plan, and utility consumption summaries.
• Accuracy Range: Low side: -15% to -30%; High side: +20% to +50%.
• Effort Index: 2 to 4 times the effort of a Class 5 estimate (typically takes several weeks).

Class 3 Estimate (Budget Authorization / Sanction)

• Project Definition: 10% to 40% complete engineering.
• Primary Purpose / End Usage: Budget authorization, project capital sanction, corporate board funding approval, and establishing the initial cost control baseline.
• Methodology: Mixed methodology (stochastic and deterministic). Major mechanical equipment is priced based on vendor budgetary quotes; bulk materials (piping, civil, electrical) are estimated using factored percentages or preliminary unit rates; job indirects are semi-detailed.
• Engineering Deliverables: P&IDs issued for design, frozen PFDs, approved general arrangement (GA) plot plan, preliminary single-line electrical diagrams, equipment datasheets, geotechnical site report, and design basis memorandum.
• Accuracy Range: Low side: -10% to -20%; High side: +10% to +30%.
• Effort Index: 3 to 10 times Class 5 effort (typically requires 1 to 3 months of dedicated multidisciplinary engineering and estimating effort).

Class 2 Estimate (Control Baseline / Tender)

• Project Definition: 30% to 75% complete engineering.
• Primary Purpose / End Usage: Detailed project control baseline, contractor lump-sum or GMP bid submission, subcontractor tender packages, and procurement buyout tracking.
• Methodology: Primarily deterministic bottom-up estimating. Detailed Quantity Takeoffs (QTO) for all major direct disciplines (concrete, steel, piping); firm vendor quotes for all major equipment; productivity-adjusted crew rates; detailed job-site general conditions schedule.
• Engineering Deliverables: P&IDs issued for construction/procurement, completed equipment specifications, structural steel layout drawings, piping isometric drawings initiated, electrical single-line diagrams completed, soil foundation designs finalized.
• Accuracy Range: Low side: -5% to -15%; High side: +5% to +20%.
• Effort Index: 5 to 20 times Class 5 effort (substantial engineering commitment, often costing hundreds of thousands of dollars).

Class 1 Estimate (Check Estimate / Definitive Tender)

• Project Definition: 65% to 100% complete engineering.
• Primary Purpose / End Usage: Check estimate to evaluate incoming bids, final definitive tender submission, change order dispute resolution, and baseline audit validation.
• Methodology: Fully deterministic bottom-up unit rate estimating. Every single direct work item is quantified from Approved for Construction (AFC) drawings; 100% firm equipment and subcontractor purchase commitments; detailed construction logistics and execution plans.
• Engineering Deliverables: AFC drawings, complete structural fabrication details, 100% piping isometrics, certified vendor equipment drawings, complete electrical wiring diagrams, fully drafted subcontracts.
• Accuracy Range: Low side: -3% to -10%; High side: +3% to +15%.
• Effort Index: 10 to 100 times Class 5 effort (represents the culmination of full engineering design).

---

4. Master Comparison Matrix: AACE RP 18R-97

The following master table synthesizes the core classification parameters tested on the AACE CCT examination:

---

5. Understanding Asymmetric Accuracy Ranges

A critical mathematical feature of AACE 18R-97 is the asymmetry of the accuracy ranges:

ACCURACY DISTRIBUTION CURVE (Positive Skewness / Asymmetry):

        Probability Density
               |
               |          * *
               |        *     *
               |       *       *
               |      *         *  <-- Expected / Estimated Cost
               |     *           *
               |    *             * 
               |   *               *  <-- Higher upside tail (Risk / Growth)
               |  *                 * 
               | *                   *             *              *
        -------+-------------------------------------------------------> Actual Cost
             -30%                    0%          +50%           +100%
            (Low)                (Estimate)     (High)        (Extreme)

Why Accuracy Ranges Are Asymmetric

1. Positive Risk Skewness: In capital construction, risks are fundamentally non-linear. The opportunities to decrease cost below zero or achieve dramatic savings are bounded, whereas the threats that increase cost—such as schedule delays, scope creep, labor strikes, severe weather, regulatory intervention, and material price spikes—are open-ended.
2. Thermodynamic Law of Project Complexity: Unmanaged entropy always increases project scope. A Class 5 estimate that is 30% low can easily double (+100%), but it will rarely cut its final cost in half (-50%).
3. Exam Implication: When an exam question displays accuracy as a symmetrical figure (e.g., "Class 5 is ±50%"), that is technically a simplification. AACE 18R-97 mandates asymmetric intervals (e.g., -50% to +100%).

---

6. The Engineering Deliverables Maturity Progression

To determine where an estimate sits along the project definition axis, estimators evaluate specific technical deliverables:

+-----------------------------------------------------------------------------------+
|                    ENGINEERING DELIVERABLES MATURITY PROFILE                      |
|                                                                                   |
|  Deliverable Item           Class 5     Class 4     Class 3     Class 2   Class 1 |
|  ------------------------   ---------   ---------   ---------   -------   ------- |
|  Process Block Flow (BFD)   Defined     Final       Final       Final     Final   |
|  Process Flow Diags (PFD)   None/Prelim Preliminary Final       Final     Final   |
|  P&IDs                      None        None/Prelim Issued Des  Issued Con Approved|
|  Plot Plan                  None        Preliminary Frozen      Final     Final   |
|  Equipment Datasheets       None        Key Items   All Major   Purchased Certified
|  Piping Isometrics          None        None        None/Prelim 30-50%    100% AFC|
|  Electrical Single Lines    None        None        Preliminary Completed Certified
|  Soil Borings / Geotech     None        Regional    Site Prelim Final     Final   |
+-----------------------------------------------------------------------------------+

---

7. Exam Watch: High-Yield Traps & Rules of Thumb

[!WARNING] The "Management Intent" Fallacy: The most common AACE CCT exam trap presents a scenario where executive management demands a "Class 2 estimate" to submit a definitive bid, but the engineering team has only provided preliminary equipment lists and block flow diagrams (under 10% engineering). The estimate is NOT a Class 2! It is a Class 4 or Class 5 estimate. Under 18R-97, the degree of project definition is the sole primary characteristic. Management's desires cannot magically elevate an estimate's technical classification.

[!CAUTION] Class 1 Does Not Mean "Zero Uncertainty": Even a Class 1 estimate, developed with 100% completed engineering and firm vendor quotes, carries an expected accuracy range of -3% to -10% on the low side and +3% to +15% on the high side. No estimate has an accuracy of 0.0%, because field construction always encounters site-specific labor productivity variations and market dynamics.

[!TIP] Memorization Aid for Accuracy Ranges: Remember the progression of the high-side accuracy threshold as definition improves:

• Class 5: up to +100%
• Class 4: up to +50%
• Class 3: up to +30%
• Class 2: up to +20%
• Class 1: up to +15%