UK lays CBAM emissions rules Sept. 9, carbon tax on imports from Jan. 1, 2027

So let us talk about what this actually means if you are the person deciding whether to take that order for twenty tonnes of aluminium billet into Felixstowe next spring. The first thing to absorb is that the UK is not inventing a new idea here — it is joining a club. The EU CBAM is already running its definitive period, the United States and Canada have their own plans on the table, and now Britain slots in alongside them. The practical effect is that embedded carbon is becoming a standard cost line on the same footing as freight or insurance. You used to quote a price, add shipping, add a margin, and you were done. Now there is a fourth line that nobody in your sales team has learned to quote yet, and it is the one line your customer will argue about hardest, because it is invisible in the product itself.

The tax point is deceptively simple to state and annoying to live with: it bites when the CBAM good is imported and a customs debt arises. In plain English, the moment your container clears and duty becomes due, the carbon charge attaches. There is no grace period on the charge itself — only on the admin. You do not have to be registered on day one of 2027, because registration does not even open until 1 January 2028. But you do have to keep the records from now, which is the part most firms will quietly fumble. HMRC will expect you to show, when the time comes, that you were tracking the embedded emissions of what you brought in. If you have been winging it with a spreadsheet that gets rebuilt every quarter, you are going to have a bad time when the verification window opens, and verification is exactly where this regime gets teeth.

Here is the worked example, and I am going to be explicit about every assumption because that is the only honest way to do this. Picture a single shipment: twenty tonnes of primary aluminium ingot landing at a UK port in, say, February 2027. Aluminium is brutal on carbon — primary smelting is one of the most energy-hungry processes in industry, and a tonne of primary aluminium carries a great deal of embedded CO2. Let us assume, purely for illustration, a default carbon value of one point five tonnes of CO2 equivalent per tonne of aluminium. That is not a published UK figure — HMRC has not released the rates or defaults yet — so treat it as a placeholder to show the mechanics. So our twenty tonnes imply thirty tonnes of CO2e embedded. Now assume an illustrative carbon price of, say, fifty pounds per tonne of CO2e. Again, illustrative — the real UK price will be tied to its own scheme and published later. Multiply thirty tonnes CO2e by fifty pounds and you get fifteen hundred pounds of carbon charge on that one shipment. On a product whose value might be, say, forty thousand pounds, that is roughly a three point seven five percent surcharge that did not exist in your 2026 quotes. Stretch that across a year of regular aluminium imports and the number stops being rounding error and starts being a line your finance director asks about every month.

The reason I keep saying assume is the central trap of this whole regime: you are pricing 2027 contracts today, but the rates that decide your actual cost will not be published until later, and the default values that kick in if your data is weak will be set by HMRC, not by you. Default values are almost always harsher than your real footprint, because a regulator has no incentive to credit you with a cleaner process than the industry average — and the average includes the dirtiest smelters on the planet. So the firm that shows up with verified, supplier-attested emissions data will pay less than the firm that shrugs and lets the default apply. That gap, over a year of steel or aluminium volume, can be the difference between a contract that is profitable and one you wish you had never signed. This is not a marginal effect; for high-carbon products it is the headline number.

For exporters the picture is messier, because you are not dealing with one regime, you are dealing with three or four at once. If you make steel in a third country and sell into both the EU and the UK, you have got two CBAM-shaped holes in your margin, and they will not have identical rules, identical reporting calendars, or identical default values. The EU system and Britain's will drift apart in the details even if they share the same philosophy. The smart move is to build one emissions dataset per product that can be sliced for each jurisdiction, rather than maintaining four separate makeshift records. The data is the same physical reality — how much carbon went into making this coil — so the discipline is to measure once and report many. Firms that treat each border as a fresh paperwork exercise will drown in it; firms that centralise the underlying measurement will handle the differences as a formatting problem, which is infinitely cheaper and far less error-prone.

Now, what do you actually do, and by when. The deadlines stack in a specific order, and missing the early ones makes the later ones painful. Starting now — and now means the week you read this, September 2026 — you should be capturing emissions records for every CBAM-relevant shipment you handle. Not next quarter. Now. The regulation says importers must keep emissions records, and the expectation is that the habit is formed before the charge bites. Build the collection process into your existing customs entry workflow so it happens automatically rather than as a heroic monthly catch-up. Identify which of your product lines fall inside the five sectors — aluminium, cement, fertiliser, hydrogen, iron and steel — because if a product is not in that list, none of this applies to it and you should not waste a minute on it. For the ones that are in, tag them in your ERP so nothing slips through unnoticed.

The next hard date is 1 January 2028, when registration with HMRC opens. You cannot register before then, and you do not need to pay the charge through the registration portal in 2027 — but you do need the records behind you to be clean by the time you do. Treat 2028 registration as the moment the spotlight turns on, and make sure the year and a half before it is spent building an audit trail, not scrambling to reconstruct one. The firms that sail through will be the ones whose 2027 records look like a steady, boring, consistent log. The firms that struggle will be the ones whose 2027 looks like a gap with a few heroic reconstructions taped over it, and auditors can smell that from a mile away.

In between, the repricing work has to happen for 2027 contracts. This is the part your sales team will resist, because adding an invisible line to a quote is awkward with a customer who is used to a simpler number. But the alternative — quoting 2026 prices for 2027 delivery and eating the carbon cost yourself — is a quiet tax on your own margin that compounds with every order. Put a carbon line in the quote, even if you label it provisional carbon adjustment subject to HMRC rates. Customers in these five sectors are all getting the same shock from every supplier, so it is not a competitive disadvantage to be early; it is a credibility signal that you understand the regime. And keep the assumption explicit in your own files so that when HMRC publishes the real rate you can true it up without renegotiating the whole contract from scratch.

Worth naming the five sectors individually, because the pain is not evenly spread. Aluminium and iron and steel are where the tonnage and the carbon intensity both run hot, so they will dominate most importers' exposure. Cement and fertiliser are lower volume for many traders but ferociously carbon-intensive per tonne, so a single shipment can carry a charge that looks disproportionate until you remember how much CO2 a tonne of clinker represents. Hydrogen is the wildcard — its embedded carbon swings wildly depending on whether it was made by electrolysis with renewable power or by steam methane reforming, so the same product arriving from two different plants can land in two completely different charge bands. If you touch hydrogen, the provenance of the molecule is now a commercial variable, not just a technical spec.

The liability sits with the importer, and that is you if you are bringing goods into the UK, not the overseas factory that made them. This trips up a lot of firms who assume their supplier will handle it. Your supplier can give you the data, and they should, but the legal obligation to hold records, to register, and to pay is yours. That means your commercial agreements need to specify who supplies the emissions evidence and on what timeline, because a supplier who sends you a number nine months late is handing you a default-value charge for an entire year. Put the data-delivery obligation in the contract now, while you still have leverage, rather than discovering the gap in 2028 when the relationship is already strained.

Your freight forwarder or customs broker is going to become a more important conversation than they used to be. The carbon charge rides alongside the customs entry, and the firms that wire it into their broker's process from the start will treat it as routine. Sit down with them this quarter and ask what their CBAM workflow looks like, what data they need from you, and how the charge will appear on your statements. If the answer is a shrug, that is useful information about whether you need a different broker before 2027, not after. The cost of switching now is a phone call; the cost of discovering your broker cannot handle it in February 2027 is a nasty surprise on a live shipment.

Do not underestimate the cash-flow angle either. A carbon charge is a real payment owed to HMRC, and across a year of volume it is money leaving your account on a schedule you do not fully control yet. If your margins are thin and your working capital is tight, a three or four percent surcharge that you failed to pass on is not a rounding error, it is the difference between a good year and a bad one. Build the charge into your cash-flow model now as a line item with a stated assumption, even if the number is provisional, so that when the real rate lands you are adjusting a model rather than discovering a hole. Finance teams hate surprises more than they hate bad news, and a modelled provisional charge is just news you can plan around.

One way to stress-test the worked example is to turn the assumptions the other way. If the real UK carbon price lands closer to the EU level, which has traded well above my fifty-pound placeholder, that twenty-tonne aluminium shipment could carry two or three times the charge, and suddenly the surcharge is not three percent but eight or ten. The point of modelling with explicit assumptions is precisely this — you see which knob moves the number most, and you learn that the carbon price is the lever, not the tonnage. That tells you where to focus: locking in lower-carbon supply matters more than shaving freight, because freight is a rounding error next to a volatile carbon price that you do not set.

Do not wait for a verification certificate to appear magically from your supplier either. The data you need is an embedded-emissions figure per tonne, ideally with the methodology behind it, and for many producers that number already exists inside their environmental reporting — it just has never been asked for by a customer before. The earlier you request it, the more time they have to dig it out, and the less likely you are to fall back on a default value that punishes you for their silence. Treat the request as routine procurement correspondence from this month onward, not as a 2028 fire drill, because the firms that normalise this conversation now will be the ones with clean files when HMRC finally asks to see them.

There are alternatives and workarounds worth knowing, and also pitfalls that will bite the unwary. The obvious pitfall is the EU overlap I mentioned — if you serve both markets, do not assume compliance in one counts in the other. They are separate filings, separate registrations, separate verification, and pretending otherwise is how penalties start. Another pitfall is leaning on default values as a strategy rather than a fallback. It is tempting to think you will just let the default apply, saves the hassle of measuring, but defaults are priced to the industry's worst performers and they will cost you more than verified data almost every single time. Use the default only when you genuinely cannot get supplier data, and work to replace it the moment you can lay hands on a verified figure.

A genuine alternative for some firms is to shift sourcing toward lower-carbon producers before 2027 locks in. If your aluminium currently comes from a coal-powered smelter and a competitor sources hydro-powered metal, the carbon charge widens the gap between you every year. That is a procurement decision with a trade-policy cause, and it is worth raising with your sourcing team now while contracts are still negotiable. Another angle is product mix: if you import both primary aluminium, which is high in embedded carbon, and aluminium scrap or low-carbon secondary product, which is much lower, the regime quietly rewards shifting toward the latter. None of this requires you to become a climate advocate — it is just reading the price signal the government has now attached to carbon and positioning your buys accordingly, the same way you would react to a change in a tariff schedule.

One more thing that does not get said enough: the rates and default values being published later is not a reason to wait, it is a reason to prepare the inputs so you can react the day they land. When HMRC drops the numbers, the firms with clean, product-level emissions records will be able to compute their exact exposure in an afternoon. The firms without will spend weeks guessing and still get it wrong. The information asymmetry here rewards the prepared, and the preparation is cheap if you start now and expensive if you start in 2028 when everyone else is also panicking and the consultants are booked solid.

So the shape of the thing is simple even if the paperwork is not. A carbon cost is now attached to five categories of goods entering the UK, it starts in 2027, the records start now, registration is 2028, and the rates are coming. Your job between now and then is to measure, to reprice, and to stop treating embedded carbon as someone else's problem. The exporters among you have the extra layer of juggling multiple regimes, but the underlying data is the same and the discipline of measuring once carries across borders. Do that, and the UK CBAM becomes a line you quote and explain instead of a surprise you absorb.

That is the whole game. The dates are fixed, the sectors are fixed, the direction is fixed. Only your readiness is up to you, and readiness is built in the quiet months before anyone is watching.

• Start logging embedded-carbon records for every CBAM shipment now — September 2026, not next quarter.
• Tag all aluminium, cement, fertiliser, hydrogen and iron-and-steel SKUs in your ERP so none slip through.
• Put a provisional carbon line in every 2027 quote and keep the assumed rate explicit in your files.
• Contractually oblige suppliers to deliver verified emissions data on a fixed timeline.
• Build one product-level emissions dataset you can slice for UK, EU and North America filings.
• Book a CBAM workflow review with your freight forwarder or customs broker this quarter.
• Model the carbon charge as a cash-flow line item now, even with an illustrative rate.