- Spoke article

Automation feasibility study. What you get, what it costs, how it shapes the capex decision.

A feasibility study turns a vague automation ambition into a defensible scope, an options analysis, an ROI framing, and a recommendation the capex committee can act on. Here is what the deliverable actually contains and what it should cost.

01 / what-it-is

What a feasibility study actually is.

Have you ever sat in a capex meeting where the automation proposal was just an integrator's quote attached to a covering page?

A feasibility study is the decision-stage document that sits between the operations team's automation ambition and the capex committee's approval. The study examines the option space, costs each credible option, frames the return on investment, identifies the risks, and recommends a path forward. The output is calibrated to inform a decision rather than to commit to one.

The reason it exists as a separate deliverable from an integrator quote is that the quote is for what the integrator can supply. The study is for the customer's decision. Different question, different answer.

02 / deliverable

The actual deliverable.

What a working feasibility study contains, section by section.

Scope statement.

What the project is, what it is not, what the boundaries are. A scope statement that future readers can use to decide whether a new sub-request belongs to the project or to a different project. The single most-revisited section in the document.

Options analysis.

Three to five credible options, ranked by cost, risk, time-to-deliver, and operational impact. Not "do nothing versus the plan" — that is not an analysis. Real options: full automation versus semi-automation, in-house versus integrator-delivered, brownfield modification versus greenfield section, platform A versus platform B where the platform decision has material consequences.

ROI framing.

The cost side: capex (equipment, integration, commissioning) plus the operating-cost delta (energy, maintenance, training). The benefit side: throughput change, quality improvement, labour reduction or redeployment, risk reduction. The framing matters more than the precise number; a capex committee can debate assumptions if the framing is clear, but cannot debate a number presented without its workings.

Risk register.

A ranked list of project risks with proposed mitigations. The register includes integration risks (the cutover that goes wrong), operational risks (the line that runs slower in production than in FAT), procurement risks (the supplier whose lead time blew out), and human risks (the operator change that needs more training than budgeted). Each risk has an owner, a likelihood, a consequence, and a mitigation. The register is the document that gets reviewed at every project gate.

Recommendation.

The study has to recommend something. "Further investigation required" is not a recommendation. A working recommendation reads: "Proceed with Option B (parallel-run brownfield migration on the filler line), subject to the three mitigations identified for the silo-interface risk, with a target capex commitment by [date]." Specific, actionable, gate-aligned.

Appendices.

The site walkdown notes, the equipment lists, the spreadsheet behind the ROI calculation, the references to standards and prior project benchmarks. Appendices support the document; they are not the document.

03 / scoping

How to scope the engagement.

The two questions that decide whether the study delivers value.

What decision will this inform?

"Whether to automate the packing line" is a different scope from "which automation platform to choose for the packing line" which is different again from "whether to add an additional packing line." Each requires a different study scope. The most common failure mode is a study commissioned to answer one question that gets handed back answering a different question.

What is the timeline of the decision?

A study informing a capex decision next quarter has different depth requirements from a study informing a strategic review for the corporate parent in twelve months. Both can be useful. They are not the same study and should not be priced the same.

The right size of a feasibility study.

For most single-line automation decisions, a feasibility study is one to four weeks of consultancy effort, depending on the complexity of the plant and the maturity of the existing controls layer. Studies that run longer are usually expanding their scope; studies that run shorter are usually skipping the walkdown.

The walkdown is the test of whether the study has done the work. A consultant who has spent two days on site, met the operations team, the maintenance lead, and the controls engineer, and read the existing PLC code, produces a different study from a consultant who reviewed a P&ID at a desk. The site time is not optional.

04 / cost

What it should cost.

The cost band is narrower than the variation in quotes suggests.

For a single-line feasibility study on a defined scope, the typical Pac Technologies engagement is fixed-fee in the low-five-figure range. A multi-line, plant-wide, or strategic-review study scales upward. The key word is fixed-fee: a consultancy that cannot fix-fee a feasibility study at scope-out has not done enough of them to know the shape. Time-and-materials feasibility engagements are usually a sign that the consultancy is going to learn the work on the customer's bill.

What the fixed fee should cover:

Site visit (typically two days on site for a single-line study).
Stakeholder interviews (operations, maintenance, engineering, capex committee chair).
Document review (existing PLC code, drawings, prior project records, capex budget context).
Options development and costing.
ROI framing.
Draft, review with the customer, revision, and final delivery.
A presentation to the capex committee if the customer wants one.

What it should not cover: ongoing engineering work, detailed design, or vendor selection. Those are FEED-stage activities.

05 / feed

The relationship with FEED.

Front-End Engineering Design is the stage after the feasibility study, after the capex committee has approved further investment, and before the project is fully committed.

What FEED adds.

Detailed equipment specifications, P&IDs, control philosophy, safety strategy, project schedule, and a Class 3 cost estimate (typically within 10-15% of the final number on a well-run project). FEED is where the chosen option from the feasibility study gets developed to the level of detail needed to commit capex with confidence.

The split that works.

Feasibility informs the go/no-go decision. FEED commits the project to a specific design. Some consultancies blur the line by including light FEED elements in a feasibility study; the result is usually a heavier study with less decision clarity. The cleaner pattern is a defined feasibility study followed by a defined FEED engagement, each at fixed scope and fixed fee.

The exception.

For small projects (capex under several hundred thousand dollars), a combined feasibility-plus-light-FEED engagement can be the right pattern, particularly where the customer already has the technical depth to validate the work in-house. The combined engagement should still produce a stop-point between the feasibility and the FEED conclusions, so the capex committee has a real decision moment rather than a sliding-scale commitment.

06 / dovetail

Where it dovetails with engineering.

The best feasibility studies are not divorced from the engineering that follows. They are written with the engineering team in mind.

Pac Technologies' consultancy practice runs feasibility studies and FEED engagements directly. Where the customer will go to market for the engineering after FEED, we write the deliverables to be vendor-neutral. Where the customer wants Pac to deliver the project, we write the deliverables to flow cleanly into our programming and commissioning work. Either path is supported. The customer gets to pick.

The engineering signals that the consultancy team is grounded:

The ROI assumptions are bounded by what the controls layer can actually measure. Not a theoretical 12% throughput lift; an upper bound the existing instrumentation can verify.
The risk register includes the integration risks the controls engineer will face, not just the project-management risks.
The handover pack from feasibility into FEED includes the PLC code review notes, not just the strategic narrative.
The recommended option survives a sanity check by an engineer who would build it. If no engineer signed off the recommendation, the document is consultancy without engineering.

07 / faq

Common questions.

What does a feasibility study actually deliver?

A defensible scope (what the project is and is not), a small set of options analysed for cost and risk, an ROI framing the finance team can take to the capex committee, a risk register with mitigation, and a recommendation. The deliverable is a document the capex committee can read in an hour and a recommendation the operations team can defend. Studies that hand back more than fifty pages are usually padding the bill.

How much does a feasibility study cost in Australia?

For a single-line feasibility study on a defined scope, expect a fixed-fee engagement in the low-five-figure range. A multi-line or plant-wide study scales from there. The number that should appear on the quote is fixed, not time-and-materials. A consultancy that cannot fix-fee a feasibility study has not done enough of them to know the shape.

When does FEED start?

Front-End Engineering Design starts after the feasibility study has produced a recommendation and the capex committee has approved further investment. FEED takes the chosen option from the feasibility study and develops it to the level of detail needed to commit to capex spend: detailed equipment specifications, P&IDs, control philosophy, safety strategy, project schedule, and a Class 3 cost estimate. Feasibility informs the decision; FEED commits to the decision.

Why not just go straight to a quote from an integrator?

Two reasons. First, an integrator's quote is for what the integrator can supply, framed in the platform and methodology that integrator prefers. A feasibility study looks at the option space without that bias. Second, going straight to an integrator quote skips the scoping work that decides whether the project should happen at all. Feasibility studies routinely conclude that the right answer is a smaller scope, a different scope, or no project — none of which an integrator quote will tell you.

- sources