On 3 September 2025 EDF and Lafarge France announced a contract that allocates to the cement maker a share of the output of EDF’s operating nuclear fleet for more than 10 years. The announcement names what the two sides share, costs and the risk on the volumes the fleet actually produces, and it names no price.
Industrial energy has no single profit pool: the money is made through contracts that allocate different risks, a regulated allowance on invested capital, a long-term share of costs and volumes, a margin on equipment, a fee for running a service, a premium paid only after certified production, and the useful question for a candidate is which of those risks the work measures, allocates or prevents.
This piece opens the second industry of the CareerOn Industry Atlas, energy for industry. It reads ten dated documents published by a regulator, a grid operator, an energy company, four listed groups, the European Commission and the International Energy Agency. It asks one question of them: when an industrial site pays for energy, where does the money stay, and on what condition? It reads the contracts first, then the four company measures, then the difference between a signature and a revenue, then the case against our reading, and it ends with three pieces of work a candidate can rehearse.
Start with the contract, not the logo
The instinct, when you ask where money is made in an industry, is to line up the largest companies and compare their margins. In industrial energy that instinct misleads, because the same euro paid by a factory can pass through several very different agreements before it becomes anyone’s profit. Each agreement answers a different question: who puts up the capital, who carries the price, who carries the volume, who is paid on delivery and who is paid only if something works.
The Commission de régulation de l’énergie, the French energy regulator, answers the first question for the distribution grid in a decision dated 13 March 2025. It says the regulator sets the remuneration of Enedis, the operator of most of the French distribution network, by a normative method that ensures a reasonable remuneration of the capital invested. The same decision records that Enedis plans a sharply rising investment programme, EUR 25,573 million in total between 2025 and 2028.
That is a precise and limited finding. It tells you that on the network, money is earned by being allowed to recover and remunerate capital under a framework the regulator writes. It does not tell you that the network is the most profitable layer of the industry, and a tariff allowance is not realised profit. We print the mechanism and nothing larger.
The EDF and Lafarge France contract answers a second question. Here the money is earned by allocating a slice of existing capacity to one industrial buyer over a long period, with the costs and the risk on produced volumes shared between the two. The press release discloses the structure. It does not disclose the price, the margin or whether either side expects an attractive return, and we do not guess at them.
Two further contracts are visible in company reporting: selling equipment, which is how Schneider Electric and Nexans earn most of their money, and running a service over time, which is how Veolia and, for its gases, Air Liquide earn theirs. The fifth is the newest and the most conditional, and the European Commission documents it exactly.
One euro can travel through five different contracts, each carrying a different risk
| Contract | How money is earned | Risk the document assigns |
|---|---|---|
| Regulated network (CRE, Enedis) | Reasonable remuneration of invested capital, set by the regulator | Not stated as profit; an allowance only |
| Long-term supply (EDF, Lafarge France) | A share of nuclear capacity over more than 10 years | Costs and produced-volume risk shared |
| Equipment delivery (Schneider Electric) | Sales of products and systems | Not stated |
| Operating service (Veolia) | Running water, waste and energy services | Not stated |
| Production-conditional support (Innovation Fund) | Fixed premium per kilogram of certified hydrogen | No payment before production |
Sources 1 Commission de régulation de l’énergie (CRE) · 3 Électricité de France · 4 Schneider Electric · 7 Veolia · 8 European Commission
Read across the five rows, the map does not say which contract makes the most money. It says that each carries a different risk, and that the documents are explicit about some risks and silent about others. Where a document is silent, the cell says so. That silence is itself information: the equipment and service groups publish results, not the risk terms of the contracts behind them.
Four measures that cannot be ranked
The listed groups do publish percentages, and those percentages are the easiest thing in this story to misuse. Each company measures its operating economics in its own way, over its own perimeter, against its own denominator. Placing them on one axis would invent a league table that none of the filings supports.
Schneider Electric reported first-half 2025 group revenues of EUR 19.3 billion, up 8% organic, and adjusted EBITA of EUR 3.5 billion, up 7% organic. Its adjusted EBITA margin was 18.2%, a figure the group itself describes as reflecting currency headwinds and seasonality. It is a group measure, not an energy-product measure, and it is not defined under IFRS.
Nexans, the cable maker, reports on a different base. Its first-half 2025 standard sales were EUR 3.8 billion against current sales of EUR 4.7 billion, up 4.9% organically. Standard sales remove the effect of metal prices, so its adjusted EBITDA of EUR 441 million, up 7.0% year on year, gives a margin of 11.7% of standard sales. That denominator is chosen precisely because copper prices would otherwise distort the ratio, and it cannot be compared with another company’s reported revenue.
Air Liquide reports recurring operating income. For the first half of 2025 its group revenue was EUR 13,722 million and its recurring operating income EUR 2,737 million. The group margin was 19.9% and the margin of its Gas and Services activity 22.0%. Those are two scopes with two denominators, and neither isolates hydrogen or a single industrial supply contract.
Veolia reported first-half 2025 revenue of EUR 22,048 million, up 3.8%, and EBITDA of EUR 3,367 million, up 5.5%, a margin of 15.3%. That margin covers water and waste as well as energy. It is not a district-heating margin, and we do not present it as one.
Four companies, four different measures: none of these percentages can be ranked
| Company | Measure, as published | Denominator | Scope |
|---|---|---|---|
| Schneider Electric | Adj. EBITA margin 18.2% | Group revenue | Whole group |
| Nexans | Adjusted EBITDA margin 11.7% | Standard sales (metal effects removed) | Whole group |
| Air Liquide | OIR margin 19.9%; Gas & Services 22.0% | Group or segment revenue | Two scopes |
| Veolia | EBITDA margin 15.3% | Group revenue | Water, waste and energy |
Sources 4 Schneider Electric · 5 Nexans · 6 Air Liquide S.A. · 7 Veolia
What the four panels do establish is narrower and more useful. Each of these groups has operating economics that it reports as positive and growing. None of the reports tells you where inside the group those economics arise, which customers produce them, or how much of the result comes from industrial energy rather than from the rest of the business. A candidate who reads these numbers as a ranking of employers learns the wrong lesson. A candidate who reads them as four different definitions of profit learns how finance teams in this industry actually talk.
A signature is not a revenue
The third question is timing. Energy for industry is full of announcements, agreements and awards, and each is reported as news. The documents themselves are careful to separate the states a project passes through, and the reader should be equally careful.
The clearest example is public support for renewable hydrogen. On 7 October 2024 the European Commission reported that the winners of the first EU-wide renewable hydrogen auction had signed their grant agreements. The Innovation Fund supports those projects with a fixed premium per kilogram of certified and verified renewable hydrogen produced, and no payments are made before the projects start production. The total support comes to EUR 694,521,237, disbursed over ten years.
Read literally, that record establishes a signed agreement and a conditional payment rule. It does not establish that any of the projects is producing, that any premium has been paid, or that any project is profitable. The sum is a commitment conditional on future production, not hydrogen revenue already earned.
The EDF and Lafarge France contract passes the same test. The announcement and the signature are documented. Deliveries under the contract, and whatever each party earns from it, are not documented in the source, so our exhibit leaves them as not documented rather than filling them in.
A signature is not production, and production is not revenue
| Record | Announced | Signed | Producing | Revenue realised |
|---|---|---|---|---|
| EDF and Lafarge France supply | Documented | Documented | Not documented | Not documented |
| Innovation Fund hydrogen grants | Documented | Documented | Not documented | Not documented |
Sources 3 Électricité de France · 8 European Commission
This is not pedantry. Across the energy transition, the distance between a signed contract and a realised revenue is where projects are delayed, renegotiated or abandoned, and it is where much of the analytical work in the industry sits. Keeping the four states apart is the first discipline of anyone who models these projects.
The demand behind the contracts
All of these contracts sit on top of industrial demand for electricity, and the latest national record is mixed. RTE, the operator of the French high-voltage transmission grid, reports that in 2024 the aggregate consumption of large industrial and tertiary users connected to its high and very high voltage network rose by 2.4% compared with 2023. The same document records that this consumption remained 12.7% below the average of the 2014-2019 period.
Both numbers matter, and neither is a revenue figure. The first says demand from the largest sites recovered in 2024. The second says it has not returned to its earlier level. Neither tells you the price those sites paid, the contracts they signed, or how many people the recovery will employ.
The counter-case
The strongest objection to this reading is that it is too cautious to be useful. An investor would say that regulated networks are plainly among the most durable businesses in the sector, that equipment makers with margins in the high teens are plainly doing well, and that a map which refuses to rank anything leaves the reader with nothing to act on.
We take the objection seriously, and the answer is in the documents. A regulated allowance can recover costs and remunerate capital without establishing realised profit. A ten-year contract can allocate risk without revealing whether either party earns an attractive return. A group margin can show that a company has healthy operating economics without showing where inside the group they arise. An awarded grant is not revenue before its production condition is met. And industrial consumption can recover while staying below its earlier baseline.
We also refused one hypothesis outright. It is often said that flexibility, storage and efficiency capture the value of costs the system avoids. None of our ten documents defines a flexibility settlement, a storage spread or a measured-savings contract, so that claim stays out of this piece until a primary document supports it.
What would change the reading
Three kinds of document would move this map. A published contract price or margin for a long-term industrial supply agreement would let us say which side of that risk-sharing earns more. A segment disclosure that isolates industrial energy inside one of the listed groups would turn a group measure into a sector measure. And a record that one of the hydrogen projects has started certified production would move it from signed to producing. Until then, the map stays at the resolution the evidence allows.
What this means for your career
If the money in industrial energy is made through contracts that allocate risk, then much of the valuable work is the work of measuring, allocating or preventing those risks. The documents point to three such tasks, and CareerOn has a simulation for each.
The first is the risk inside a long-term power contract. The EDF and Lafarge France agreement shares costs and volume risk over more than a decade, and advising an industrial buyer on that kind of structure is the work of an energy-transition consultant. The source documents the structure. It does not tell you the price, the margin or how common such contracts are.
The second is concentration in the battery supply chain. The International Energy Agency’s report on the global supply chains of electric-vehicle batteries finds raw material processing highly concentrated, with five major companies responsible for three-quarters of global production capacity in one segment and four companies for half of global anode material capacity. Those are 2021-era global capacity figures, not current shares, not procurement costs and not the exposure of any one European buyer, and a supply-chain analyst’s first task is to know that difference.
The third is lost output at solar plants. RTE estimates that in 2024 about 0.8 TWh of French solar production was lost because part of the fleet was economically incited to lower its output during negative-price episodes. That is one documented category of loss at national scale. It is not the performance ratio of any plant, and it is not a breakdown of soiling, temperature, inverter or availability losses, which is the work a solar performance analyst does plant by plant.
Three risks, three rehearsals, each with its limit printed
| Risk | Rehearsal | The evidence cannot tell you |
|---|---|---|
| Cost and volume risk in long-term power | Industrial energy-transition consultant | Price, margin or how common such contracts are |
| Supplier concentration in battery materials | Battery supply-chain analyst | Current shares or procurement cost (2021-era baseline) |
| Curtailed output (about 0.8 TWh, France 2024) | Solar performance analyst | A plant performance ratio or loss waterfall |
Sources 3 Électricité de France · 9 International Energy Agency · 10 RTE (Réseau de Transport d'Électricité)
If you want to advise a factory on how it buys energy, start with the industrial energy-transition consultant simulation. If you want to test where a battery supply chain can break, start with the battery supply-chain analyst simulation. If you want to find out why a solar plant produces less than it should, start with the solar performance analyst simulation. None of these documents tells you how many such jobs exist or what they pay, and we do not pretend otherwise. Each simulation lets you do the work before you choose it.