Skip to content
CAPIntelligence

Capital Portfolio Development & Challenge

Tier 2 · Annual

The right projects.
At the right depth, in the right year.

An evidence-based challenge of every project on the capital plan, run with your site leadership before the budget locks. Each item takes the same plan-entry test, gets a scored tier rather than an argued one, and is phased against the cash the year can carry. What comes out is a plan your steering team can defend line by line, and a portfolio every surviving project is then assured against at its gates.

Run one on your capital planAll services→

Why it usually fails

Most capital plans are a wish list with a total at the bottom.

A spreadsheet goes up in the budget meeting. Seventy rows, a rough number on each, a color for whether it is going ahead. The person who argues hardest keeps their project, the reliability engineer who is not in the room loses theirs, and the total is trimmed to what finance will sign, one row at a time, from the bottom.

Ask in March why a particular project is funded and nobody can reconstruct the reasoning. Ask in September why the year is overspent and the answer is the three emergent items that were added without anything being displaced. That is the real failure. Not that the wrong calls were made, but that no test was applied, so none of the calls can be defended or learned from.

The fix is not a longer budget meeting. It is giving every item the same test, at the same three points in the cycle, with the reasoning written down as you go.

The plan-entry test

Nothing goes on the plan without four things.

This is the whole discipline in one rule, and it is the rule most sites do not have. An item that cannot produce these four is not ready to be compared with the ones that can, so it does not get compared. It gets sent back.

01

A Terms of Reference

One page that says what the problem is, what the item would do about it, who sponsors it, and when it has to land. Approved by the capital projects lead and the sponsor, not just listed.

02

A risk-register entry

Every sustaining item points at the risk it retires. If the risk cannot be written down and rated on your own matrix, the item is a wish, and the wish list is not the plan.

03

A rough-order cost and a class

A Class 5 number is fine. A blank is not. The estimate class is stated so nobody reads a rough order as a budget.

04

A front-end plan and cash phasing

Which tier, which gates, roughly when, and how the spend falls across the years. An item with no phasing lands in one year and distorts every comparison.

Classified, so like is compared with like

Every item carries a capital class and a funding source. Sustaining capital splits into the non-discretionary (regulatory and compliance), the risk-based (integrity, reliability, end of life), and the turnaround-window scope that is funded through the turnaround plan and executed as a project. Growth capital splits into cost reduction and new or expanded business. Funding is internal, grant or program, insurance, or third party, because a grant-funded project carries a conditions register the others do not. A compliance item and a debottleneck are never argued against each other on the same axis, because they are not on the same axis.

The tier

Scored, not judged.

A client’s binder process usually has one ladder. Big projects tolerate it and small projects evade it, and the evasion is where the surprises come from. The answer is not a shorter ladder, it is four of them, and a calculator that decides which one a project climbs.

Seven factors, each rated one to five: total installed cost, engineering novelty, tie-ins, coupling to a turnaround window, regulatory exposure, contractor interfaces, and schedule criticality. The score assigns the tier, and the tier assigns the gate ladder, the document set, and the assurance the project carries. Two people sizing the same project land in the same place, and a factor that fires on its own, a novelty score of four say, adds its own requirement regardless of tier.

Same process at every tier. Fewer activities as the tier falls. That one mechanism is the whole difference between a major project’s hundred and fifty steps and a replacement-in-kind’s sixteen.

The four tiers, and what each one carries

Major

Above about $20M

G1 to G6 · CAP·AR1 to AR4

Full front-end rigor, every gate, full document set, independent review mandatory at sanction.

Standard

About $5M to $20M

G1 to G6 · CAP·AR1 to AR4

Full gate structure, condensed deliverables.

Small

Under about $5M

G1 to G4, then G6 · AR3 as a desk review

Combined definition, weeks of front end rather than months, startup readiness absorbed into the safety review.

RIK

Any value, like for like

G3 and G6 only · None

A single combined definition-and-funding gate, standard templates, pre-approved basis, expedited change screening.

Bands are indicative and calibrated to each client’s population. The calculator is yours to edit once the standard is adopted.

Illustrative

The annual cycle

Three levels of challenge, ten months to one.

The challenge is not one meeting. It is a ladder timed to the budget year, and each rung asks a narrower question of a shorter list.

M-10

Months before the budget year

Level one: the list

Every candidate held to the plan-entry test. Items that fail are sent back for a Terms of Reference or dropped. Replacement-in-kind candidates are screened onto the light track. Turnaround-coupled work is flagged for its window.

M-6

Months before the budget year

Level two: the challenge

The survivors are tiered by the calculator and challenged as a set: the risk threshold decides which sustaining items proceed this year, growth items are tested against the business case, and the year's cash envelope is drawn.

M-1

Months before the budget year

Level three: the lock

The plan is locked with a firm twenty-four-month window and a budget substitution rule, so an emergent risk mid-year displaces something by a recorded decision rather than by adding to the total.

The risk threshold

A line on your own matrix decides which sustaining items proceed in the year. Above it, funded. Below it, deferred with the risk accepted by name. Nobody argues a rating in the meeting; the rating was set before the meeting.

The substitution rule

An emergent risk mid-year does not add to the year. It displaces something, and the displacement is a recorded decision in the portfolio forum, so September is not a surprise.

The minor-capital pool

Items below a threshold are pooled and drawn down, not listed. The plan stops being seventy rows of which forty are under the threshold, and the challenge spends its time where the money is.

What you get

A plan you can defend, line by line.

01

The challenged plan

Every item with its class, its tier, its funding source, its stage, its year-phased cash, and its disposition, with the reasoning recorded.

02

The tier register

Each project scored on the seven factors and assigned a tier, which sets its gate ladder, its document set, and how much assurance it carries.

03

The firm window

A twenty-four-month itemized plan inside the rolling five-year one, with the monthly refresh your project managers actually run.

04

The Terms of Reference set

One approved page per item on the plan, revalidated annually and expired at twenty-four months if the item has not reached its first gate.

And it does not stop when we leave

The portfolio lives in CAP·PATH, so your team keeps working it after the session: new items arrive through a scoping request that routes through the disciplines before anyone sees a row, the portfolio forum meets monthly on live data, and every surviving project deploys its tier’s work process and carries its gate reviews on the same record. The facilitated challenge is where the discipline is set. The platform is where it holds.

How it runs

Most of the work happens before the room sits down.

Send what you already have

Your capital forecast in whatever shape it is in, your risk matrix, and your approval ladder. The forecast is imported as it stands, section headings and colors read as stages, so the first view is your list, not a blank page.

We calibrate the calculator

The tier bands and the seven factors are set against your population and your thresholds, so the tiers mean something at your site rather than ours.

The three rungs

Level one on the list, level two on the survivors, level three at lock, each a working session with your site leadership and every disposition recorded as it is made.

Lock and carry on

The plan locks with its firm window, the Terms of Reference set is issued, and the portfolio forum takes over on a monthly rhythm.

Method per our capital portfolio and long-range plan standard, built on public frameworks: AACE recommended practice for estimate classes and contingency, CII front-end planning practice cited for definition maturity, and PMI portfolio management practice for the cycle.

Bring order to the list

How much of your capital plan could you defend today?

Bring your forecast and your matrix. We will show you what the plan-entry test does to it, and what the reasoning looks like written down.

Start a conversation