Capital Cost Estimation for Process Industries

Where Early Numbers Go Wrong

1. Why Capital Cost Estimates Matter Beyond the Budget

Capital cost estimates are risk management documents as much as financial ones. They determine whether a project proceeds, at what scale, in what configuration, and with what financing structure. An estimate used for a decision it was not designed to support (sanctioning a project on an order-of-magnitude number, or committing construction financing against a preliminary estimate) creates downstream consequences that compound through the project lifecycle.

The cost structure of the facility being built is a direct consequence of the capital estimate process. Fixed cost allocation, depreciation schedules, maintenance assumptions, and unit economics are all derived from the capital cost of the asset. An estimate that understates capital cost by 30% does not merely produce a budget overrun: it produces a facility whose real fixed cost structure was never properly modeled in the investment case, with effects on product economics that can persist for the life of the asset.

Build-vs-outsource analysis is among the contexts most exposed to estimation error. A capital estimate that understates owned-facility construction cost against a CDMO or contract manufacturing alternative distorts the analysis in ways that may not surface until the project is well into detailed engineering, by which point the decision has been made and the cost of reversal typically exceeds the cost of proceeding.

Estimation accuracy is also a governance issue. Boards and investment committees that approve capital projects on estimates presented without their class, basis, or applicable accuracy range are not in a position to assess the risk they are accepting. An estimate is not a number: it is a number with an uncertainty range, a defined scope, and a set of assumptions that, if wrong, change the number.

2. The Estimation Hierarchy: Classes, Purpose, and When to Use Each

The AACE International classification system (Class 5 through Class 1) is the most widely used framework for defining estimate maturity in process industries. Equivalent systems are used in pharmaceutical and life sciences project management (FEL 1/2/3) and in infrastructure and construction (RIBA stages); many organizations also maintain internally calibrated formats mapped to these frameworks. The underlying logic is consistent: as engineering progresses and scope is defined, uncertainty narrows and the estimate becomes a more reliable basis for financial commitment.

The table below compares the standard estimate classes: engineering basis, accuracy range, primary purpose, the decisions each is designed to support, and the information required to complete an estimate of that class.

Internally developed estimate formats (built from an organization’s own project cost history and calibrated to its facility types, geographies, and regulatory environments) often outperform published industry factors within their calibrated domain. Such a format is, in effect, a validated Class 3 or Class 2 estimating tool. Organizations with sufficient project history should maintain and update these formats as standard practice.

The principle governing all estimate classes is that each should be used only for the decision it was designed to support. Most projects are sanctioned at Class 3. The −20% to +30% accuracy range of a Class 3 estimate is not a flaw in the methodology: it is an accurate representation of what can be known at that stage of engineering. The flaw is in treating a Class 3 estimate as though it carries Class 1 precision.

3. Industry Cost Drivers and the Safety/Regulatory Layer

Accuracy ranges vary by industry because different facility types carry different dominant cost categories, and those categories carry different levels of uncertainty at early estimate stages, including, in most industries, a significant safety and regulatory cost layer that is systematically underestimated at concept stage. Understanding which cost drivers dominate in a given context, and how the safety and regulatory layer is structured, is as important as understanding the estimate class itself.

3.1 Specialty and Fine Chemicals

Specialty chemical facility estimates are dominated by equipment and process-related civil and structural costs, with a significant overlay from safety, environmental, and hazardous area classification requirements. Accuracy at Class 3 improves once process flow diagrams and preliminary P&IDs are available, but remains sensitive to safety classification decisions that may not be finalized until hazard reviews are complete. Estimates that do not reflect a preliminary hazardous area classification study carry material uncertainty in their electrical and instrumentation cost elements.

Hazardous area (electrical area) classification, determined under IEC 60079 (international) or NEC Article 500/505 (US) based on the flammability and explosive properties of materials handled and the likelihood of their release, determines the specification of electrical equipment and instrumentation throughout the facility, not only in process areas. Equipment in classified zones must meet specific protection standards (Ex d, Ex e, Ex p, or equivalent), and the cost premium over standard electrical equipment is significant across every motor, junction box, luminaire, and instrument in the classified area.

Relief system design adds further civil and structural scope: pressure relief valves and rupture discs sized for credible overpressure scenarios, and emergency vent systems, knock-out drums, and connections to flare or scrubber systems that scale with the inventory of hazardous material and the consequence of its release. Dust handling, from baghouse filters for nuisance dust to explosion-rated systems with suppression or venting for combustible dust, adds capital cost that depends on the dust hazard classification. Flammable material storage, secondary containment, and loading/unloading systems with grounding, overflow protection, and vapor recovery are further items commonly underspecified at Class 3.

Environmental controls for air and water add a parallel cost layer: fabric filters or electrostatic precipitators, carbon adsorption for VOCs, scrubbers for acid gases, flares, wastewater treatment, and hazardous waste handling. This layer cannot be estimated accurately until the emissions characterization is complete enough to scope the required controls. Labor cost differences between countries are a further source of estimate variance: engineering labor, skilled trades, and fabrication costs vary substantially between the US, EU, India, and China, and estimates built on one geography’s benchmarks cannot be applied to another without adjustment.

3.2 Pharmaceutical and Life Sciences

Pharmaceutical and life sciences estimates are distinctive because regulatory compliance, rather than process optimization, determines a large proportion of the design and equipment specification. Cleanroom classification (ISO 14644 and EU GMP Annex 1), HVAC redundancy, airflow patterns, pressure cascades, and segregation requirements are all substantially determined by the regulatory framework applicable to the product and process. Validation (IQ/OQ/PQ protocols, computer system validation, and commissioning documentation) typically represents 10–15% of total project cost in GMP facilities and is absent from most concept-stage estimates.

The clearest example of compliance-driven equipment cost is single-use (disposable) bioreactor systems for antibody and biologics manufacturing. A single-use facility carries a fundamentally different capital cost profile from a stainless-steel facility of equivalent throughput: lower initial capital for pressure vessels and CIP/SIP systems, but higher operating cost from consumables and different utility requirements for bag integrity testing and waste management. The choice between single-use and conventional equipment is a regulatory strategy decision as much as a capital cost one, and it must be resolved before the Class 3 estimate can be meaningfully prepared.

3.3 Food and Nutraceutical Manufacturing

Food and nutraceutical facilities carry capital cost implications from sanitary design and sterilization requirements that, while less stringent than pharmaceutical GMP, are more specific and more costly than general industrial standards. The primary cost driver is the specification of process contact surfaces (materials, surface finish, weld quality, and geometry designed to eliminate contamination accumulation points) and the design of the sanitation or sterilization system. Prevention-based design concentrates cost in materials of construction (typically 304 or 316 stainless steel), sloped drainage, and elimination of dead legs; active sterilization design adds CIP systems, steam sterilization circuits, or UV treatment. Both approaches carry civil and utility requirements that differ from general industrial standards and should be resolved at the concept stage, before the estimate is prepared.

Jurisdiction-specific GMP requirements add a further layer of variability: FDA 21 CFR Part 111 requirements for dietary supplement facilities differ from those under 21 CFR Parts 110/117 for food, and EU food law and supplement regulation impose requirements that differ again for facilities targeting those markets.

3.4 Oil and Gas / Petrochemical

Oil and gas and petrochemical projects benefit from the largest analogous project databases and the most established unit cost benchmarks of any process industry sector. Early-stage accuracy is typically better than generic ranges for equivalent estimate classes, because the cost structure (dominated by equipment and piping, with well-characterized installation factors) is well represented in published and proprietary cost indices. Labor cost differences between geographies remain a significant source of variance for international projects; offshore and remote onshore locations carry cost premiums not reflected in standard onshore benchmarks.

3.5 General Industrial and Infrastructure

General industrial and infrastructure projects have lower process complexity than most process industry contexts but higher site variability. Geotechnical conditions, existing utility infrastructure, environmental baseline conditions, and existing structures carry more uncertainty relative to total project cost, and are systematically underinvestigated at early stages because the investigation cost is visible and the risk cost is not.

3.6 The Cost of Late Integration

Across all industries, safety and environmental requirements identified during detailed engineering rather than front-end engineering produce cost corrections that exceed what early identification would have cost. The design response to a late finding (rerouting piping for containment, upgrading electrical specification, adding structural elements for suppression systems) is constrained by decisions already made, and the correction cost is therefore higher than if the requirement had been incorporated into the original design basis.

4. The Greatest Sources of Estimation Error

The following cost drivers account for the majority of estimate-to-actual variance across project types and industries, in approximate order of frequency and magnitude of impact.

4.1 Scope Definition Gaps

Scope definition gaps are the single largest driver of cost overrun across all project types. Items not in scope at estimate time are not zero-cost: they are deferred costs that surface during execution, typically after the design decisions affecting their cost have already been made. The most common gaps at Class 3 are utility interconnections and offsites, tie-ins to existing systems, and indirect costs applied as percentage factors rather than estimated from defined scope.

4.2 Equipment Specification Changes and Budgetary Pricing

Equipment cost is typically the most reliable element of a Class 3 estimate, but three systematic effects affect its accuracy. First, design changes between estimate stages frequently alter equipment specification with significant cost consequences: a pressure vessel rated for 150 psig carries a materially different cost from an atmospheric tank of equivalent volume, and a change in operating conditions or a safety review that identifies the need for pressure containment can produce a step change not reflected in the Class 3 basis. Second, budgetary pricing from manufacturers is typically issued on limited specification data and tends toward the low end of the actual cost range; when final specifications (materials of construction, surface finish, instrumentation, safety interlocks, delivery schedule) are issued, the cost routinely increases. Third, the availability of used and refurbished equipment on the secondary market compresses the cost variance attributable to the base equipment design itself, which shifts the dominant source of cost variability toward the ancillary items (controls, safeguards, and instrumentation packages) required to bring that equipment into compliance with the project’s specification. Controls, safeguards, and instrumentation not in the budgetary scope remain among the most common sources of cost growth between Class 3 and Class 1.

4.3 Lack of Data for Novel or Scaled Technology

Estimates for facilities incorporating novel technology, processes without commercial-scale analogues, or equipment beyond the established database are structurally more uncertain than estimates for conventional processes. Published factor sets and unit cost databases are calibrated to historical projects; where the technology is new or the scale is unprecedented, the estimation basis is weaker regardless of estimate class. Very large projects also encounter diseconomies of scale in construction logistics, labor management, and supply chain that standard scaling factors do not capture.

4.4 Site-Specific Conditions

Geotechnical conditions, existing utility capacity and routing, environmental baseline conditions, and the condition of existing structures are rarely fully characterized at Class 3 for projects on existing industrial sites. A geotechnical investigation costing a fraction of a percent of total project cost at pre-FEED may prevent a foundation cost increase of several million dollars when soil conditions differ from assumptions: the decision not to investigate early is a cost decision with asymmetric consequences.

4.5 Safety, Environmental, and Regulatory Engineering

Safety, environmental, and regulatory engineering costs are systematically underestimated when review is decoupled from front-end engineering. Section 3 details the industry-specific drivers. The cost consequence of identifying a requirement during detailed engineering propagates through civil, structural, electrical, and instrumentation scope in ways that are difficult to estimate until the detailed design response is defined.

4.6 Cost Uncertainty Over Time: Inflation, Exchange Rates, and Supply

The gap between an estimate’s reference date and the date of actual expenditure introduces cost changes independent of project scope or execution performance. Inflation in labor and materials markets, commodity price movements, exchange rate fluctuations for internationally sourced equipment, and supply shortages that elevate prices or extend lead times all affect actual cost relative to the estimate basis. These factors are distinct from contingency: they are market-driven changes in the cost of executing defined scope over time, not risks within it, and must be estimated and tracked separately, using current cost indices calibrated to the relevant markets and expenditure timeline.

4.7 Project Size and Estimating-Factor Vintage

The sources of error described above do not scale proportionally with project size: the same absolute cost increase represents a larger percentage impact on a small project than on a large one. A cost increase on one vessel in a two-vessel plant moves the total estimate by a materially greater percentage than the same increase would on a ten-vessel plant. Smaller and mid-size specialty chemical, pharmaceutical, and nutraceutical projects are therefore more exposed, on a percentage basis, to the same absolute estimating errors that a large petrochemical project would absorb without materially affecting its accuracy range.

This exposure is compounded by the vintage of the factor sets themselves. Many published installation and indirect-cost factors were originally compiled from large, standardized hydrocarbon-processing projects (plants that were common through the 1960s and 1970s) and share relatively little in equipment configuration, automation level, or degree of customization with the smaller, more specialized batch and continuous-process facilities typical of today’s specialty chemical, pharmaceutical, and nutraceutical sectors. Applying a factor set calibrated to that earlier generation of large, standardized plants to a smaller, highly customized facility, without adjustment for scale and configuration, understates the estimate’s true uncertainty.

4.8 Contingency Miscalculation

Contingency is most commonly set as a percentage of total direct cost by convention rather than by analysis of estimate class, scope completeness, and project-specific risk profile. A conventional percentage applied to a Class 3 estimate with known scope gaps is structurally inadequate, and is the most common proximate cause of apparent budget overrun on projects where scope has not materially changed from sanction.

5. Contingency and Escalation: Definition, Calculation, and Appropriate Use

5.1 Contingency

Contingency is a measure of uncertainty, not an expectation of upside or downside. It quantifies the cost of known unknowns within the defined project scope (items identified as risks but not yet fully characterized) and should be distinguished from scope changes, which require formal change control, and from management reserve, held at a higher organizational level against unknown unknowns.

Contingency should be calculated from three inputs: the estimate class and its accuracy range; the completeness of scope definition relative to a typical project at that class; and the project-specific risk profile, including the probability and cost impact of identified risks. Monte Carlo simulation applied to the estimate and schedule risk model provides the most defensible basis at Class 2 and above; for Class 3, a structured qualitative risk assessment mapped to cost impact ranges is a practical alternative. Contingency set as a fixed percentage of direct cost by convention (typically 10%) is not a risk-based calculation, and provides false assurance on projects with significant site uncertainty, novel technology, or regulatory complexity.

5.2 Escalation

Escalation covers the change in cost of labor and materials between the estimate date and the date of actual expenditure. It is not optional: it is a real cost that accrues regardless of project performance. Inflation, commodity price movements, exchange rate fluctuations, and supply market conditions all contribute and must be modeled separately from contingency; conflating the two produces budgets that appear adequately funded at sanction but are structurally underfunded before construction begins.

5.3 Modeling and Internal Historic Data

The most reliable basis for both contingency and escalation is internal project cost history, applied through structured cost modeling. Organizations with a history of capital projects should maintain a database of estimate-vs-actual data organized by project type, estimate class, facility type, cost category, and geography: this produces contingency factors grounded in actual organizational experience and escalation indices calibrated to the relevant markets.

External cost indices provide market-level benchmarks for escalation and for calibrating unit costs, but their applicability varies by industry and cost component and should be matched to the facility type being estimated. The Chemical Engineering Plant Cost Index (CEPCI), published monthly, is compiled specifically from process-industry equipment, labor, and construction cost data and is the escalation index most widely applied to specialty chemical, pharmaceutical, and food/nutraceutical facilities. The S&P Global Commodity Insights Upstream and Downstream Capital Costs Indices (UCCI/DCCI) serve the equivalent function for oil and gas, refining, and petrochemical projects. The ENR Construction Cost Index, by contrast, is a general building-construction index: its materials basket (structural steel, portland cement, and lumber) and labor basis (common construction labor) track civil and building-construction cost movement rather than the process equipment, piping, and instrumentation that dominate a chemical or pharmaceutical facility’s capital cost; it is appropriately applied to the civil and site-development scope of a process industry project, not to process equipment or installation. Compass International’s cost data is organized by facility type, and its Front-End/Conceptual Yearbook (covering oil, gas, and process/industrial/chemical construction) is the applicable product for process industry estimates; its general Global Construction Costs Yearbook is calibrated primarily to general and infrastructure construction. None of these external indices provide the project-type-specific calibration that internal data supplies. The combination of an industry-appropriate external index for macro-level escalation and internal project history for project-type-specific cost factors produces more accurate estimates than either source alone.

Monte Carlo simulation converts a point estimate into a probability distribution of total project cost, producing the P50, P70, and P90 outcomes that frame contingency correctly as a measure of uncertainty and give sponsors the information needed for a risk-informed sanction decision.

6. Best Practices, Tools, and Techniques

Estimation quality is determined by process discipline and tool selection as much as by the underlying engineering data.

6.1 Process Discipline

Basis of Estimate Documentation

The basis of estimate (BOE) should document, at minimum: the estimate class and accuracy range; scope inclusions and exclusions stated explicitly; data sources and methodologies for each cost element; assumptions that, if wrong, would most materially affect the estimate; and identified risks covered by contingency with their assumed probability and cost impact. An estimate without a fully developed BOE cannot be independently reviewed, updated systematically as scope evolves, or used as the basis for meaningful variance analysis.

Scope Freeze

Scope freeze (formal closure of project scope prior to estimate completion) is a prerequisite for a reliable estimate above Class 5. Where scope cannot be fully frozen, open items should be identified explicitly and carried as separate allowances with stated uncertainty ranges, not absorbed into contingency. The cost of a given change increases the further the project has progressed when the change is made, since each subsequent phase of engineering and procurement is built on the assumption that earlier decisions are fixed: which is the practical reason scope freeze functions as a cost-control discipline and not merely a scheduling one.

Independent Estimate Review

An independent estimate review, conducted by a team without a stake in the project’s approval, is the most reliable mechanism for identifying scope gaps, methodology weaknesses, and unsupported assumptions. It should be commissioned at Class 3 for any project above a defined capital threshold, with findings presented to the project sponsor alongside the estimate, not resolved before the sponsor sees them.

Successive Estimating

Producing and formally comparing estimates at each project phase allows scope growth and cost escalation to be tracked as engineering matures. Divergence between successive estimates at equivalent scope is an early warning signal requiring explanation: scope growth, methodology difference, market movement, or correction of a prior error.

6.2 Tools and Techniques

Factored Estimating

Factored estimating derives total installed cost from a primary cost driver (typically major purchased equipment) by applying installation factors. It is the standard method for Class 4 and Class 3 estimates and produces reliable results when the equipment basis is current, scope is well defined, and factors are calibrated to the relevant industry and geography.

Parametric Estimating

Parametric estimating uses statistical relationships between project cost and measurable parameters (capacity, throughput, floor area, equipment count) derived from historical data. Cost-capacity scaling relates the cost of a facility of one capacity to another through an exponent reflecting economies of scale; the exponent varies by process type, and applying a generic factor to a different scale profile produces inaccurate results. Parametric estimating is most reliable with a large, recent, project-type-specific dataset.

Detailed Bottom-Up Estimating

Detailed bottom-up estimating builds the estimate from quantities and unit costs for each discrete cost element. It requires substantially complete engineering deliverables and is the only methodology that reliably achieves Class 1 accuracy. Unit cost databases require regular calibration; data more than 12–18 months old in an active construction market carries material price risk.

Monte Carlo Simulation

Monte Carlo simulation converts point estimates and contingency factors into probability distributions characterizing the range and likelihood of project outcomes. It is the standard tool for quantitative risk analysis at Class 2 and above and provides the most defensible basis for contingency setting.

Internal Historic Data and Cost Modeling

Internal project cost history, maintained in a structured database and applied through calibrated cost models, is the single most valuable estimation input for organizations with a capital project history. A cost model incorporating estimate-vs-actual variance by project type, estimate class, cost category, and geography produces estimates that systematically narrow over successive projects as the model is recalibrated against completed data.

7. Commissioning and Using an Estimate That Is Fit for Purpose

7.1 Before Commissioning the Estimate

• Define the decision the estimate is required to support before specifying the estimate class.
• Specify the basis of estimate document as a required deliverable alongside the estimate number.
• Confirm that scope has been formally closed, or that open scope items are identified and carried as explicit allowances with stated uncertainty ranges.

7.2 When Reviewing the Estimate

• Ask for the estimate class and its applicable accuracy range before evaluating the number.
• Ask what is excluded from scope. The exclusion list identifies items that will surface as cost additions during execution.
• Ask how contingency was calculated. A percentage applied by convention is not a risk-based contingency.
• Ask whether an independent estimate review has been conducted.

7.3 During Execution

• Use successive estimates as a continuous tracking tool. Divergence between the current estimate and the sanction estimate at equivalent scope is an early warning signal.
• Manage scope changes through formal change control. Contingency is not an appropriate vehicle for absorbing scope growth.
• Track contingency drawdown against specific risk events. When contingency is consumed without a documented risk event, the project’s actual cost position is obscured.

8. Conclusion

The gap between the capital cost estimate that secures project approval and the cost at project completion does not originate in construction. It originates in front-end engineering: scope that was not defined, safety and regulatory requirements that were not characterized, site conditions that were not investigated, equipment costs based on budgetary pricing before controls and safeguards were specified, and contingency set by convention rather than by analysis. The estimate class framework exists to make these limitations explicit, to state the uncertainty range applicable at a given stage of engineering maturity, and to define the decisions for which each estimate is, and is not, an appropriate basis.

The capital cost of a facility determines its fixed cost structure, its unit economics, and the financial return on the investment decision. Getting it wrong at concept stage does not merely produce a budget overrun: it produces an asset whose economics were never properly modeled.

Pathway 2 Product provides capital cost estimation support, front-end engineering assessment, and independent estimate review for specialty chemical, pharmaceutical, and food and nutraceutical manufacturing projects. A companion cost estimating template covering AACE Class 5 through Class 1 is available free in the Tools section. Inquiries may be directed to info@pathway2product.com.

---

References

1. AACE International. (2020). Cost Estimate Classification System: As Applied in Engineering, Procurement, and Construction for the Process Industries. Recommended Practice No. 18R-97. AACE International, Morgantown, WV. web.aacei.org
2. AACE International. (2021). Basis of Estimate. Recommended Practice No. 34R-05. AACE International, Morgantown, WV. web.aacei.org
3. AACE International. (2022). Risk Analysis and Contingency Determination Using Expected Value. Recommended Practice No. 41R-08. AACE International, Morgantown, WV. web.aacei.org
4. Merrow, E.W. (2011). Industrial Megaprojects: Concepts, Strategies, and Practices for Success. Wiley, Hoboken, NJ.
5. Peters, M.S., Timmerhaus, K.D., and West, R.E. (2003). Plant Design and Economics for Chemical Engineers. 5th ed. McGraw-Hill, New York.
6. Hand, W.E. (1958). From flow sheet to cost estimate. Petroleum Refiner, 37(9), 331–334.
7. Lang, H.J. (1948). Simplified approach to preliminary cost estimates. Chemical Engineering, 55(6), 112–113.
8. ISPE. (2018). ISPE Baseline Guide: Volume 3: Sterile Product Manufacturing Facilities. 3rd ed. International Society for Pharmaceutical Engineering, Tampa, FL. ispe.org
9. U.S. Food and Drug Administration. (2016). 21 CFR Part 111: Current Good Manufacturing Practice in Manufacturing, Packaging, Labeling, or Holding Operations for Dietary Supplements. U.S. Government Publishing Office, Washington, DC. ecfr.gov
10. Chemical Engineering. Plant Cost Index (CEPCI). Access Intelligence, LLC, New York, NY. Published monthly; process-industry-specific escalation index. chemengonline.com/pci-home/
11. Engineering News-Record (ENR). Construction Cost Index. BNP Media, Troy, MI. Published weekly; general building-construction index. Applicable to civil/site scope only. See Section 5.3. enr.com/economics
12. S&P Global Commodity Insights. (2023). Upstream Capital Costs Index (UCCI) and Downstream Capital Costs Index (DCCI). S&P Global Commodity Insights, London. spglobal.com/commodityinsights
13. Compass International. (2023). Front-End/Conceptual Yearbook: Oil, Gas & Process/Industrial/Chemical Construction. Compass International, Inc., Blue Bell, PA. compassinternational.net

Back to Technical Insights