Commonwealth Cost Estimation Guidance (DITRDCA) P50/P90

The Federal Funding Gate

What the Guidance Is — and Why It Matters

The Cost Estimation Guidance is published by the Australian Government Department of Infrastructure, Transport, Regional Development, Communications and the Arts (DITRDCA). It sets out the principles a proponent must follow when preparing the cost estimate that accompanies a Project Proposal Report (PPR) submitted under the Notes on Administration for funding under the land transport Infrastructure Investment Program (IIP) — generally road and rail.

It is aimed at jurisdictional public-sector agencies (including local government) and their estimators, covers capital cost only (not operations and maintenance), and applies across successive project phases from the business case forward, scaled up for major or high-risk projects. If a project wants Commonwealth money, this is the framework its estimate is judged against.

National Land Transport Act 2014

The statutory basis (Parts 3 & 7) under which the Commonwealth invests in land transport infrastructure.

Notes on Administration

The administrative rules that govern IIP funding, including how and when a cost estimate is submitted and reviewed.

Project Proposal Report

The submission that carries the cost estimate. The department reviews and assesses that estimate before any recommendation to the Minister.

An estimate that does not present a defensible P50 and P90 with probabilistic contingency will not pass departmental review for any project of material size. Distinguish these Commonwealth documents from state manuals that implement the same P50/P90 logic — such as Queensland TMR's risk and contingency manual. Same probabilistic philosophy, separate document families.

The Document Set

A Numbered Series of Guidance Notes

The guidance is not a single document. It is a set of Guidance Notes that fit together under an Overview, supported by a Project Cost Breakdown template that must accompany PPR submissions.

Note	Title	Version / date
Overview	Framework, definitions, how the notes fit together	v2.0, Nov 2023
1	Project Scope	v2.0, Nov 2023
2	Base Cost Estimation	v2.0, Nov 2023
3A	Probabilistic Contingency Estimation	v2.0, Nov 2023
Supp. 3A	Probabilistic Cost Estimation (detailed mechanics)	v2.0, Nov 2023
3B	Deterministic Contingency Estimation	v2.0
4	Escalation	v2.0, Nov 2023

Guidance Note 4 is Version 2.0

A common error reads Note 4 (Escalation) as "v3.0" — the "-v3" appears only as a filename revision tag in the document URL. The document cover and footers read Version 2.0. We cite it as v2.0.

Mandatory Submission Artefact

The Project Cost Breakdown (PCB) Template

A Locked, Macro-Enabled Workbook

The PCB is a macro-enabled, locked Excel workbook with nine worksheets and separate road and rail versions. The Overview states PCB templates must accompany each PPR submission. We keep the threshold general — the framework does not state a firm dollar trigger for the PCB itself.

Built-in escalation: it auto-calculates escalation from the department's seven-year escalation forecast, pre-loaded with jurisdiction-specific road indices and the national rail index, reissued annually.

Reference-Class Forecasting Built In

The PCB carries a "Reference Class" dropdown, signalling that reference-class forecasting is built into the framework as a benchmark and governance check on the bottom-up estimate.

Lineage: the 2011 "Best Practice Cost Estimation Standard" (Evans & Peck, now Advisian) was a never-formally-published proof/training version. The numbered Guidance Notes are the department's replacement, carrying forward its P50/P90 probabilistic principles.

Estimate Anatomy

The Three Estimate Components

A project cost estimate under the guidance comprises three distinct, separately maintained parts. Keeping them separate is a discipline the framework insists on.

1 · Base Estimate

The cost of the defined scope, excluding contingency and escalation. Its top-level split is client costs + construction costs; construction cost in turn comprises the contractor's direct costs (plant, equipment, materials, labour), indirect costs (preliminaries, supervision, site establishment) and margin (overhead + profit).

2 · Contingency Allowance

The amount in excess of the base estimate for the residual risk the funder is prepared to bear — covering inherent and contingent risk. Contingency = P-value − base estimate.

3 · Escalation Allowance

Applied to the project cash flow to convert real cost to out-turn (nominal) cost. Escalation must be kept strictly separate from contingency.

Base + Contingency = (non-escalated) P-value estimate → + Escalation = out-turn cost

First-Principles Is the Preferred Method

The guidance states "the first principles method is the preferred method… recommended… wherever possible", and expects it at the development and delivery phases for projects over $25M out-turn. First-principles build-ups are structured as Plant, Labour, Material, Subcontract (PLMS) with productivity assumptions, on a 4-level WBS (Element / Heading / Item / Rate build-up) obeying the "100 percent rule". Alternatives — global/order-of-magnitude, composite-rate, unit-rate and hybrid — are used at earlier phases.

The Worked Example (Overview, Table 2)

From Base Estimate to Out-Turn Cost

A single example from the Overview shows the whole chain: a base estimate tags up to P50/P90 project estimates through contingency, then to out-turn cost through escalation. The figures below are taken directly from the guidance.

Base estimate → Project estimate

Contingency is added on top of the base estimate to reach the P-value project estimate. The P50 contingency here is $15M; the P90 contingency is $40M.

Project estimate → Out-turn cost

Escalation (2.5% in this example) is applied to the project cash flow to convert real cost to nominal out-turn cost — $124.33M at P50 and $151.36M at P90.

Both P50 and P90 are reported

The out-turn P50 is the normal approval basis; the out-turn P90 represents the funder's total exposure and is notionally held against the project — see the next section.

The Centre of the Framework

P50 Approval Basis vs P90 Notionally Held

The funding mechanism is more precise than the common shorthand "P90 is the commitment". Approval decisions are made on the P50; the P90 is held against the project and released only on demonstrated need.

P50 — The Approval Basis

The estimate with enough contingency for a 50% likelihood the final cost is not exceeded — the expected/most-likely cost. "Funding approval decisions are normally made at each phase on the basis of validated estimates of P50 out-turn costs", re-escalated each phase with current forecasts.

P90 — Notionally Held

The estimate with enough contingency for a 90% likelihood the final cost is not exceeded. "The P90 funding is notionally held against each project, but only released to jurisdictions as required for individual projects on a demonstrated needs basis." Both P50 and P90 are reported.

Risk & Contingency Methodology

The $25 Million Decision Gate

The method you may use to derive contingency depends on the project's total anticipated out-turn cost, including contingency. Above $25 million, probabilistic Monte Carlo estimation is mandatory.

Probabilistic estimation is quantitative risk analysis using Monte Carlo simulation — generating a large sample of possible outcomes and a cumulative distribution (S-curve) from which P50 and P90 are read. 5,000 to 10,000 iterations are recommended (margin of error around 1.4% at 5,000 iterations, 1.0% at 10,000).

Model Discipline

The Risk-Factor Method Is Preferred

Risk-Factor / Risk-Driver (Preferred)

The department's stated preference. It explicitly warns that "the structural difficulties inherent in line-item ranging make it difficult to arrive at realistic contingency assessments". We present risk-factor as the preferred approach.

Distributions & Three-Point Inputs

Inputs are elicited as three-point estimates (optimistic / most-likely / pessimistic → P10 / mode / P90) and fitted to Beta PERT, triangular or Trigen distributions — the choice of shape "usually has negligible impact".

Correlation Must Be Defined

"Valid Monte Carlo assessment specifically requires that the dependencies, or correlation, between the model inputs… be defined." The cardinal rule: no iteration may produce a combination of values that could not possibly occur.

Practical Modelling Discipline

Right-size the model: keep models to roughly 20–40 inputs — avoid over-disaggregation.
Model rare high-impact events separately rather than burying them in general contingency.
Treat narrow P10–P90 spreads with thin tails as a red flag of an unrealistic model.
Discrete risk events come from the risk register at a strategic, not technical, level — documented and quantified in risk workshops with stakeholders.

Guidance Note 4

Escalation to Out-Turn Cost

Note 4 exists so escalation rates used to out-turn IIP projects are derived by a robust, replicable method across jurisdictions. Escalation development is an economist's task and must be kept strictly separate from contingency.

Escalation accounts for cost change from the base date of the estimate (the date the rates used were current — not necessarily when the estimate was prepared) to a future period, generally construction completion.
Annual escalation = average of indices in the current financial year ÷ average for the previous financial year, expressed as a percentage change.
Escalation is applied to the project cash flow (base + contingency spread across the financial years of expenditure).
A cost including escalation is the out-turn (nominal) cost; excluding escalation it is a real cost.
Exemption: escalation need not be included if the project completes within 12 months of the base date.

The Escalation Engine: RCOCI / Rail-COCI & the Zero Floor

Road escalation uses jurisdiction-specific RCOCI indices; rail uses the national Rail-COCI (with separate above-/below-ground weightings). A "zero floor" policy means departmental escalation rates are never set below 0% in any financial year. The PCB template auto-applies these indices from the department's annually reissued seven-year forecast.

Estimate Review & Assurance

The Documentation Bar a Reviewer Can Demand

The cost estimate is reviewed and assessed by the department — using internal resources or independent specialists — before any recommendation to the Minister. This is exactly where Cenex's offering fits.

After a review meeting, the reviewer submits a draft project cost estimate review report, and the department provides feedback for the proponent to address. Governance sits within the Notes on Administration framework, with validation at successive project phases. The Overview requires proponents to be able to provide:

Basis of Estimate & BOQ

The basis of estimate, bill of quantities and detailed estimate — the documented logic behind every number.

Risk-Tool Outputs

S-curve (CDF), histograms, tornado / sensitivity charts and the P50/P90 values — the confirmed required model outputs from Notes 3A / Supplementary 3A.

Electronic Native Files

"At least full access, under supervision, to all underpinning data… such as native Expert Estimation, Primavera P6, and @Risk files."

Estimate Quality Framing

The Overview frames quality as Accuracy, Bias, Calibration and Reliability. Note the term "optimism bias" is not used in the Overview or Note 4 — that belongs to broader policy, not these notes.

How Cenex Delivers

Commonwealth-Ready Estimates That Survive Review

Cenex prepares estimates built to anticipate and survive the department's review — or acts as the independent reviewer the framework anticipates, with Chartered Engineer sign-off.

First-Principles Base Estimate

Bottom-up PLMS build-ups split into client and construction costs, direct / indirect / margin

Monte Carlo with @Risk

Risk-factor models, dollarised risk register, correlation defined, P50 and P90 both reported

Note 4 Escalation

RCOCI / Rail-COCI escalation to out-turn cost, contingency and escalation kept separate

Independent Estimate Review

Cenex can act as the independent reviewer, traceable and auditable to the workshop record

Related frameworks & services

The Commonwealth guidance is one of three named frameworks Cenex builds contingency under. Explore the rest of the hub:

TMR Risk & Contingency

Queensland's Project Risk Management & Contingency Development Process Manual — the state implementation of P50/P90.

RES Contingency Guideline

The cross-jurisdiction best-practice authority on probabilistic contingency development.

Frameworks Compared

How TMR, RES and the Commonwealth guidance line up on thresholds, P-values and method.

See also our Risk Modelling & Management service and the Cost Risk & Contingency hub.

Seeking Commonwealth Funding?

Cenex prepares DITRDCA-aligned cost estimates — first-principles base, Monte Carlo P50/P90 contingency and Note 4 escalation — built to survive departmental review. We can also act as your independent estimate reviewer.

Talk to Cenex

Commonwealth Cost Estimation Guidance (DITRDCA)