The cheapest place to fix a project is before it exists

Somewhere between the first idea and the final investment decision (FID), someone will ask you to fund a FEED study, the front-end engineering design work that produces no steel, no wells and no revenue, only documents. It is tempting to treat that line in the budget as the easiest saving. In practice it is the most expensive one. The months before FID are the only stretch of a project's life when changing your mind is cheap, and what you demand from the FEED study decides how much of that window you actually use.

Why is a change before FID cheap and after FID expensive?

Before FID, a decision lives on paper. Moving a facility, switching the development concept, adding a well or removing one means revising drawings, models and spreadsheets. The cost is engineering time. After FID, that same decision has been turned into contracts, purchase orders, fabrication slots, rig bookings and permit applications. Changing it means unwinding commitments that contractors and suppliers have already priced and planned around, each of whom will send a bill, and going back to the regulator with a revised application.

Early on, your ability to shape the outcome is at its highest and the cost of a change at its lowest. As the project advances, the two trade places. Once equipment is being built, almost anything can still be changed, but only at a price that eats into the return the project was approved on.

That is why we say the cheapest place to fix a project is before it exists. Many flaws are present from the first week, sitting in an assumption nobody tested, and they only surface once the money is committed.

What is a FEED study, in plain business terms?

FEED stands for front-end engineering design. In oil and gas, the FEED study is normally the last major piece of engineering before FID, and deep geothermal and CO2 storage projects borrow the same discipline. Its job is to define one chosen development well enough to make three things possible:

• a cost and schedule estimate that a board can approve with open eyes;
• a scope of work that contractors can bid on without guessing;
• a design package, the drawings and specifications that fix what is being built and why, which detailed engineering can pick up without starting over.

Put simply, a FEED study turns a good idea into something you can buy. Before it, you have a direction. After it, you have a defined project, an estimated price, a timetable and a clear list of the risks that remain.

Concept or FEED: which one are you paying for?

The two are often blurred, and the blur is costly. Concept work answers the question of what to build. It compares different ways of developing the resource (connecting to facilities already in place or building new ones, one well design or another, phasing the development or doing it in one go) and selects the option that best fits the reservoir, the market and your appetite for risk. FEED answers the next question: now that the choice is made, what exactly is it, and what will it cost?

The common trap is to rush the first question and pour effort into the second. If the options were never seriously compared, the FEED is not refining a decision. It is decorating one that was made by default. Our concept engineering work settles that choice properly, so that the FEED starts from a concept that has earned its place.

A detailed FEED of the wrong concept is still the wrong concept, just with more confidence attached.

A simple test before you approve a FEED budget: ask why the chosen option beat the others. If nobody can answer clearly, with the reasons written down, you are not ready for FEED.

What should an owner demand from a FEED study?

A FEED study is paid for by the owner, but it is not always written from the owner's point of view. It is often written by an engineering contractor hoping to win the execution work that follows. That is normal practice, but it means the people writing the study and the people paying for it do not always want the same thing. These are the questions we would put to any front-end package before it reaches a board:

• What was compared, and what was ruled out? The study should show the alternatives considered and why they lost. A single option presented as obvious is a warning sign.
• Which assumptions carry the economics? Every project rests on a handful, such as reservoir performance, well count, equipment delivery times and approval dates. You should see them listed, and see what happens to the case if each one is wrong.
• How mature is the estimate, really? Ask what the cost and schedule figures rest on: supplier quotes, benchmarks from comparable projects, or judgement. A precise number built on a loosely defined scope is more dangerous than a wide range that is honest about itself.
• What is still open? A good FEED ends with a short, explicit list of unresolved risks, each with a named owner. A study that claims to have closed everything has usually stopped looking.
• Can a contractor price it without guessing? Scope left vague at FEED comes back as contingency in the bids or as claims during execution, and you pay for it either way.
• Does the design match what the regulator will accept? The study should show that the authority's expectations were checked, not assumed.

None of these questions needs an engineering degree to ask. Judging the answers is harder: telling whether a well count is credible, whether a delivery date is realistic, or whether the regulator will accept the design takes someone on the owner's side who reads the study for what it leaves out as much as for what it says. That is the job of an owner's engineer. We have done it since 2015, across oil and gas, geothermal and CO2 geological storage.

When should you bring in independent eyes?

Earlier than most owners think. Second opinions usually arrive once execution goes wrong, when the expensive decisions are locked in. The moments that matter come before that:

• When you inherit a concept, through an acquisition, a new partner or a change of team, and nobody can explain why it was chosen. If you are buying the asset, that question belongs in your technical due diligence, before the price is agreed.
• Before the FEED scope is written, because the scope decides what the study will and will not examine.
• Before the execution contracts go out to tender, because bid documents are built on the FEED and its gaps travel straight into them. Found now, they are questions; found later, they are contract changes you pay for. Good tendering and contracting starts with a front-end package that is ready to be priced.
• Before the board sees the FID package, while an independent review is still cheap and can still change the outcome.

The same logic holds beyond oil and gas. A deep geothermal developer or an industrial emitter planning CO2 storage faces the same kind of sequence, with the same cheap window before money is committed, often with less in-house experience of large capital projects. Two of our published cases come from those industries: "Deep geothermal wells, delivered to standard" and "CO2 storage site, characterised and licensed".

If your project is heading for a big decision, such as approving a FEED budget or taking FID, or a FEED study is waiting for your signature, tell us where it stands. We answer every enquiry within 24 hours, which is usually well before the decision date.