# Accuracy and Limitations

## How accurate are the capital cost estimates? {#how-accurate-are-the-capital-cost-estimates}

!!! Summary
The accuracy of a fixed-capital estimate depends on how proven the technology is and how much information is available about it — from −15% / +25% for an established process with ample information, widening to −40% / +70% for a conceptual one.
!!!

The accuracy of a capital cost estimate in a Commodity Production Costs report is expressed as a range around the reported [fixed-capital](../glossary/d-g.md#fixed-capital) figure, and it is set by two things: how mature the production technology is, and how much information is available about it. The less proven the technology and the less that is known about it, the wider the range. The table below gives the expected accuracy range by technology maturity (rows) and information availability (columns):

{.compact}
| Technology maturity | Low | Average | High |
|---|---|---|---|
| Established (Outdated) | −25% / +40% | −20% / +30% | −15% / +25% |
| Established (In Use) | −25% / +40% | −20% / +30% | −15% / +25% |
| Emerging | −30% / +50% | −25% / +40% | — |
| Embryonic | −35% / +60% | — | — |
| Conceptual | −40% / +70% | — | — |

!!!secondary
A blank cell marks a combination that does not occur. High-information estimates do not exist for the least-mature technologies: while a process is still in development or scale-up, its uncertainties make the amount and reliability of the information about it not comparable to established technologies.
!!!

The five technology-maturity levels are defined in the [analysis methodology](../commodity-production-costs/analysis-methodology.md#how-is-technology-maturity-assessed). Why the ranges reach further above the reported figure than below it is explained in [Why are the accuracy ranges asymmetric?](#why-are-the-accuracy-ranges-asymmetric).

## How accurate are the operating cost estimates? {#how-accurate-are-the-operating-cost-estimates}

The operating cost estimate is accurate to −10% to −20% on the low side and +10% to +20% on the high side, with the exact range depending on the maturity level of the process examined. That accuracy is stated at a 90% confidence level, consistent with the conceptual-evaluation nature of the analysis. As with capital costs, less mature processes carry the wider range.

## Why are the accuracy ranges asymmetric? {#why-are-the-accuracy-ranges-asymmetric}

The accuracy ranges are asymmetric — for example +50% / −30% rather than a symmetric ±40% — because a lack of available information usually leads to underestimating a project's cost rather than overestimating it. The upside allowance is therefore larger than the downside: an estimate is more likely to fall short of the eventual cost than to exceed it, so the range extends further above the reported figure than below it.

## What accuracy class do the estimates follow? {#what-accuracy-class-do-the-estimates-follow}

The estimates follow the [class 4 budgetary estimate](../glossary/a-c.md#class-4-budgetary-estimate) standard — a conceptual-level cost estimate used for investment screening and preliminary feasibility decisions. A Commodity Production Costs report is a class 4 budgetary estimate, not a process design package, a front-end engineering design package, or construction-ready engineering. This class matches the conceptual basis of the analysis, and is the reference point behind the accuracy ranges reported for capital and operating costs.

## What assumptions underpin every estimate? {#what-assumptions-underpin-every-estimate}

!!! Summary
Every estimate assumes a standardized design practice typical of major chemical companies, a plant built on a clear field, and a typical large single-line capacity — unless a report states otherwise.
!!!

Every estimate rests on a common set of design assumptions, applied unless a report explicitly states otherwise:

- a **standardized design practice**, typical of major chemical companies;
- a plant built on a **clear field** — a site with no existing infrastructure to reuse or build around;
- a **typical large single-line capacity**.

These assumptions keep reports comparable across commodities, production routes, and countries, because every analysis reflects the same kind of modern, world-scale plant. The specific design standards a real project adopts can move both capital and operating costs, so where a report departs from this basis it says so.

## Are the reported consumption figures exact? {#are-the-reported-consumption-figures-exact}

!!! Summary
Raw material and utility consumption figures come from a conceptual design and are generally reasonably accurate, but tend to be slightly understated versus a real plant, which loses additional material and energy to vents, unscheduled equipment, inerting, and start-up and shutdown.
!!!

The raw material and utility consumption figures reported are derived from a conceptual process design, and are generally reasonably accurate. They do, however, tend to be somewhat understated relative to a real operating plant, because a conceptual design does not capture every source of loss. In practice, a running plant loses material and energy to vessel vents, unscheduled equipment, inerting systems, inaccuracies in physical-property data, and start-up and shutdown — along with other operations not addressed at the conceptual-design stage. The reported consumptions should therefore be read as sound estimates that a real plant will tend to exceed slightly, rather than fall below.

## Can estimates be compared with actual plant costs? {#can-estimates-be-compared-with-actual-plant-costs}

!!! Summary
An estimated project cost is a projection of the probable cost, not an exact number. Comparing it with an actual plant cost or a contractor's quote calls for care — seven factors can make the two legitimately diverge, and residual error is partly absorbed by the project contingency.
!!!

A reported estimate can be set alongside an actual plant cost or a contractor's quotation, but the comparison has to allow for factors that make the two legitimately differ. Seven are worth keeping in mind:

- **Minor process differences** — small, often unnoticed differences between similar processes can noticeably affect cost.
- **Omitted process areas** — omitting process areas from the design considered can invalidate a comparison with the estimate presented.
- **Overdesign** — industrial plants may be overdesigned for particular objectives or situations.
- **Equipment and construction cost swings** — rapid fluctuation of equipment or construction costs can invalidate a capital estimate.
- **Market price swings** — market price fluctuations can invalidate an operating cost estimate.
- **Below-margin bids** — equipment vendors or engineering companies may provide goods or services below profit margins during economic downturns.
- **Location-specific taxes and fees** — some locations impose higher taxes and fees, which can affect costs considerably.

Errors from these and other factors — cost and labor changes, construction problems, weather, strikes — are partly absorbed by the [project contingency](../commodity-production-costs/capital-investment.md#what-is-project-contingency-and-how-is-it-set). An estimated project cost is not an exact number, but rather a projection of the probable cost.