L'Air Liquide S.A. ($AI)

Good afternoon, ladies and gentlemen, and welcome to the Air Liquide Conference Call on 2025 Sustainability Performance. [Operator Instructions] I will now hand over to the Air Liquide team. Please begin your meeting, and I will be standing by.

Thank you, Sandra. Good afternoon, everyone. This is Aude Rodriguez, Head of Investor Relations. Thank you for attending the call today. On our website, on top of the CSRD disclosure, you will find our 2025 integrated annual report, a press release detailing our extra financial performance and the presentation we are sharing today. Joining me today are Diana Schillag, member of the Executive Committee, notably in charge of sustainability, along with Louis-Francois Richard and Guillaume De Smedt, respectively. Louis-Francois and Guillaume will join Diana for the Q&A session. Let me now hand you over to Diana.

Thank you very much, Aude, and welcome to everyone. It's clearly a great pleasure to be with you to present our 2025 sustainability performance and as well draw perspective on what is coming next. 2025 is an important milestone for Air Liquide. It marks the successful conclusion of the advanced strategic plan, which, as you know, was the first to combine financial and nonfinancial performance. Before we dive into the specific indicators and projects, let me share some elements of context with you. The advanced cycle started in a time when sustainability matters benefited from a wide consensus across markets and as well geographies. And this is a time when a lot of net zero commitments were taken. Now 5 years later and following a number of geopolitical events, the landscape has considerably evolved with very diverse and fragmented dynamics in markets, geographies and as well regulations. And of course, with rising importance given to energy security and sovereignty. Now despite these complexities, we have delivered on our commitments, strengthening the group's resilience while delivering measurable positive impacts for the planet. And I would like to very warmly thank all the Air Liquide teams for this success. Please turn to Slide 2. Now let's take a look at our agenda for today. We will begin with the sustainability performance results, especially as many of our advanced commitments reached that term in 2025. We will then focus on our climate transition plan and share with you where we stand regarding our decarbonization levers. And to make it concrete, we will talk you through several flagship projects that will illustrate how we are turning ambition into reality. And finally, of course, we will open the floor for a Q&A session. Please turn to Slide 3. Now let's start with our environmental performance and more specifically with CO2. Compared to 2020, the group reduced its absolute CO2 emissions by 13%, reaching by that our first objective of inflection point actually 1 year ahead of schedule. That's what we told you already last year because it was already -- the inflection point was already achieved last year. So this represents a reduction over the total time until 2025 included of 5.1 million tons of CO2 over the period, much ahead of what a linear decrease would require to reach the 2035 objective of minus 33%. Furthermore, we decreased our carbon intensity compared to 2015 by almost half, more precisely, 46%. And this is significantly surpassing our 2025 target of minus 30%. What about water? Well, we met our objective. Water management plans were implemented at all of our 75 priority sites in high water stressed areas, another element enhancing our operational resilience. Let's turn to Slide 4. Moving now to Health. We continue to see solid progress and concrete impacts. 64% of our 2.3 million patients at home now benefit from personalized care plans. And as populations age and chronic diseases rise, we are responding to the growing patients' preference for being serviced at home, providing the right level of care at optimal cost. At the same time, in low- and middle-income countries, access to oxygen still remains a challenge in many areas. Since its launch in 2017, our Access Oxygen program has expanded with a population of about 3.5 million people living in regions now covered by this initiative. And let me highlight that we are the only industrial gas company supporting the World Health Organization's ambition in this field. Please turn to Slide 5. Now I would like to share our social achievements. Beyond our environmental contributions, we are dedicated to social progress for our employees and striving to bring positive contributions to the communities we serve. Let's start by our absolute priority, which is safety. In 2025, we achieved a record low accident frequency rate of 0.4 for our employees. This represents a remarkable 60% reduction in just 2 years. What it means is that about 100 people went home safe, thanks to preventing lost time accidents. Moving on to diversity. We now have 34% women in management and professional roles. It marks a plus 4% increase since 2020 and positioning us as a leader in our industry. Why is this specific number so important? Because research consistently shows that the tangible benefits of diversity being innovation, adaptation and of course, performance really begin to materialize once the minority reaches 1/3 of the total population. Finally, we are also proud of how we support our people and our communities. Our common basis of care program intends to guarantee the same level of social coverage for our 65,000 employees worldwide. It was fully deployed at the end of 2024, 1 year ahead of schedule. On top of that, we have also fully deployed a minimum living wage across the entire group. This ensures that every employee everywhere in the world receives an adequate salary that allows them to live with dignity with his or her family. Last, we continue the positive impacts we bring to society through our Citizen at Work corporate citizenship program. It's now deployed in all countries and affiliates. In 2025, we had more than 7,000 active employees volunteering their time and expertise to support local community projects. Please turn to Slide 6. Let me now go through our climate transition plan and its levers. Please go to Slide 7. As you certainly recall, our decarbonization road map is built on 3 strategic levers that combine environmental performance with operational resilience. First, low-carbon energy sourcing. By securing massive amounts of renewable and low-carbon electricity, we are effectively protecting against energy price volatility and strengthening energy security. Second is asset management. By improving energy efficiency, reconfiguring assets and electrifying, we are structurally increasing the group's operational performance and resilience. Last, carbon capture and storage, what you know as CCS, which is a way to decarbonize our assets, but also the only large-scale solution to decarbonize hard-to-abate industries in the near term. While this market is still maturing and relies on evolving regulations, our technology is ready and the first projects are materializing. Please go to Slide 8. Now how did this translate into numbers? This waterfall chart illustrates how we achieved our 13% absolute emission reduction since 2020, driven by 2 major levers. Low-carbon energy sourcing has been clearly the most significant driver, accounting for 53% of the decrease. We accelerated our shift to low-carbon power, bringing the share of low-carbon power to our global mix to 40% in 2025. This alone contributed to a reduction of 2.7 million tons of CO2. The second lever is asset management, contributing 35% of the decrease. We continue to drive operational efficiency across our production sites to reduce energy consumption and fossil fuel reliance. This includes the strategic electrification of steam-driven air separation units, for instance, in China, alongside several asset optimization and reconfiguration projects. Collectively, these actions have structurally lowered our footprint by 1.8 million tons of CO2 over the period. Finally, while CCS is not yet reflected in the 2020 to 2025 reduction numbers, the technology is mature and ready with our first project to capture CO2 from our largest hydrogen plant in Europe being built as we speak. Please turn to Slide 9. I would like to share with you one illustration for each of our levers. And here, a flagship example of our first lever is clearly South Africa. Air Liquide is running the world's largest oxygen production site in Secunda, South Africa with 47,000 tonnes per day, and that had the opportunity that actually I had the opportunity to visit last year. Air Liquide and Sasol joined forces and have signed 1 gigawatt of PPAs of renewable power for the site in combination of new solar, wind and battery facilities. For Air Liquide alone, this is 460 megawatts that will bring 1.4 million tons of CO2 emission reductions every year. So this project is a major contributor to the group's decarbonization. Let's turn to Slide 10. For the next illustration, let me take you to China. In Tianjin, where we have carried out a major modernization of 2 large-scale air separation units in the country's fourth largest city. This is part of a contract renewal to supply 4,000 tons of oxygen every day to a chemical customer. In many Chinese industrial basins, steam is generated by burning coal. By replacing steam with electricity and sourcing low carbon and renewable electricity, we are avoiding 800,000 tons of CO2 emissions every year. This project perfectly illustrates Air Liquide's business model, whereby investment is backed by long-term commitment by the customer. Energy transition projects target the same standard rate of return than any other project. On top of having reduced the group's CO2 emission, this project also cuts the site water usage by 50%. Given the site of the site -- the size of this site, this represents minus 8% of the group's total withdrawals. And with its green leap forward, China is certainly the place where energy transition is happening at an accelerated pace, and Air Liquide is proud to be actively contributing to this trend. Please turn to Slide 11. Let's now look at the concrete application of our third lever, CCS. The Porthos project in the Port of Rotterdam is evidence that when the right ecosystem exists, CCS projects move from concept to reality. Starting in the next few quarters, the Porthos Consortium will store 2.5 million tons of CO2 annually from the different partners. For Air Liquide, this will allow us to reduce the Scope 1 emissions of our largest hydrogen plant in Europe by 400,000 tons per year. Porthos succeeded, thanks to 3 factors: industrial synergies with the shared infrastructure, proven technology, our proprietary Cryocap solution and policy support with a clear regulatory framework in the Netherlands and carbon contract for difference. We are convinced that the first CCS projects in Europe will open doors for wider adoption as CCS is the only available option to decarbonize hard-to-abate industries in the near term. Let's turn to Slide 12. And now I would like -- I suggest that I conclude with a few last messages. So please, Slide 13. Well, in conclusion, despite a complex international context, we have delivered a strong sustainability performance during the advanced period. Our achievements also contributed to enhancing the group's resilience. At Air Liquide, sustainability is not just a responsibility. It's a fundamental strategic choice and a driver of performance. As we enter 2026, we stay the course on energy efficiencies, low-carbon sourcing and climate adaptation with a pragmatic and focused approach, always translating our ambitions into concrete impacts for our customers, the planet and society. Looking forward, we are convinced that advocacy for low-carbon products must continue, especially for hard-to-abate sectors. The carbon measures initiative that we joined at the end of last year is a typical example of a platform aiming at promoting low-carbon product markets. I'll stop here, and we are now happy to take your questions. Thank you very much.

[Operator Instructions] We will now take the first question, which is coming from the line of John Buckland from W1M Investment Management Limited.

Can you hear me right?

Yes, very well.

I'm just asking because I'm using the computer to call you rather than a phone. This is the modern way. Anyway, thank you for your presentation and the progress you're making. But I do have some more detailed questions about future progress because when you look at the details of emissions and energy use, the impact of your renewable purchasing is not really seen very strongly. For example, Scope 1, Scope 2 emissions, both market and location base are pretty much the same, which means you're not doing much better than the grid, local grid. And then when you look at the total renewables energy used, it's still around 14% of the total. So it's -- renewables went up 6.4% versus total energy consumption of 3.2%. So -- and it shows on the table that you provided today still around 14%. So it's not really clear that you're making a huge amount of progress on purchasing renewable energy. I wonder if you can talk about that, please.

Yes, of course. And so I will give maybe just a short reminder on our Scope 2 with actually low-carbon power and then hand over to Guillaume, who can go into the details of renewable electricity and PPA purchase that we have done. So just as a reminder, we have actually accelerated the shift to low-carbon power, especially through the pluri-annual power purchase agreements. In 2025, the share of low-carbon power purchased by the group reaches 40%, so 40, but that's for low carbon power. And globally, the low-carbon electricity sourcing has already prevented the emissions of 2.7 million tonnes of CO2 per year compared to 2020. And that is actually already delivering quite a measurable impact, especially in the carbon-intensive geographies that we operate in. Maybe, Guillaume, do you want to add a word on the Scope 2 for renewables?

But that -- can I just interrupt? You said 40%. What you're saying is that of the low carbon energy that you're purchasing, 40% is PPA. Is that right?

No. What I'm saying is that the share of low carbon power purchased by the group out of the total power purchased is 40%. So low carbon represents 40, 40% out of the total power purchased.

Why is that not shown in the table of -- in your environmental table that provided the Excel farm I've downloaded. It's said here that total renewable energy consumption is 9.5 million megawatt hours and the total is 67 million, and that's only 14% of the total. So 14% versus 40% doesn't make sense to me. And perhaps I'm being stupid, but perhaps you could explain.

No, I think what we need to clarify is that I'm talking about low-carbon electricity. And I think you mentioned the renewable energy, so purely renewables. And of course, low-carbon includes as well nuclear energy, for instance. So the renewable piece within low carbon is certainly the number that you have in the table.

Yes. Maybe I can elaborate a bit first on the numbers because, in fact, the table you have is a mandatory table we have to publish as per the CSRD reporting through the climate and energy standard, ESRS E1, where basically the 14% is the share of renewable energy out of the total energy procured by the group, knowing that the total energy procured by the group is both the electricity, the power we use to run our air separation unit, but also the natural gas we purchase for operations, which dilutes a lot in the numbers in the figures you have on the 14% the share of renewable power. If you look at power only, and I think maybe it's in the text, but it's not in the table, you have the 9.5 terawatt hour of renewable power. When you add the nuclear power, you go through to 15.6 terawatt hour, which represents 40% of our power purchase.

Excluding natural gas?

Excluding natural gas for the share, which is used as a fuel for our hydrogen production units.

Maybe to help your communication, it would be useful to expand this table to make it clearer.

Thanks for the suggestion.

And on the location base versus market-based point, it happens and that's part of our decarbonization and that's a key lever in our decarbonization that we still have a handful of air separation units, which are driven by steam, knowing that the emission factor of steam is much higher than the emission factor of the grid. So that's why when we electrify a steam-driven ICU like the project that was presented, we save and we reduce a lot our emissions. One electrification, it's several hundreds of thousands of tons of CO2 per year, which is reduced.

So how many air separation units globally still use steam? And when -- what is the program to getting that to 0?

We have, let's say, we don't disclose the precise number of assets, but we have a dozen of such assets out of a fleet of 400 -- more or less 400 air separation units worldwide. So it's really a very, very small amount and the program is to work with our customers in order to shift them to electrical power when the access to power in the location permits.

And that is why actually the electrification is one of the levers that we have highlighted as well in our presentation and as well in the example that we have chosen. So that's exactly what you have seen in the Chinese example that I just mentioned in Tianjin.

[Operator Instructions] We will now take the next question from Keith Lee from Lombard Odier Investment Managers.

I had 2 questions. One is on the topic of Scope 3. So I know I obviously see that you have disclosed your Scope 3 emissions, and there is some detail on the strategy to reduce them. I was wondering what are the considerations around setting a Scope 3 target, please? And whether there's a plans to do so? Or if not, why not? And then secondly, with regards to water management, and obviously, congratulations on achieving your target for 2025. I'm wondering what do you -- what are the plans for setting new targets? What kind of KPIs will you be disclosing on water risk moving forward, please?

Okay. Thank you very much. So yes, on Scope 3, you have certainly seen that we are actually year after year reinforcing the Scope 3 categories and becoming more and more clear on the elements that we share on Scope 3, working as well on the methodology for estimations. Scope 3, it's always a hard work on methodology to really solidify our estimations on Scope 3 because it is, of course, as you all know, linked to the emissions of, on the one hand side, our suppliers and on the other side, downstream, our customers. So it always takes a bit of effort to actually identify that. For the moment, we are still in that phase of clarification what we include in our Scope 3 emissions. And that is where we are -- what we are focusing on our energy on for the moment. So we have not planned short term to fix Scope 3 objective. Now when it comes to water, water is clearly a very -- a second big topic for us. We have achieved to establish 100% of our -- to equip 100% of our 75 priority sites with water management plans. And this is clearly a big achievement. It was a big step forward. Now of course, given the importance for the group, we will not stop here. Now we are not in the face of revealing a new strategic plan. So as we -- this is actually work in progress. But I can -- without revealing a secret, I can already tell you that water will clearly be one element where we will continue to move forward. Maybe one first element that I could already share is that in -- when we started our effort on water, the Aqueduct referential, we used the Aqueduct referential at that time. That's how we came to the 75 priority sites. Now when you look at our disclosure in the URD, we have actually disclosed the total amount of water withdrawals for all the sites under the new Aqueduct guideline. So the number of our sites has actually risen to 112. And going forward, this is, of course, the number that we are tackling. So as a minimum, we will certainly target to equip all of those 112 sites with water management plan. But again, more to come on water.

Great. And sorry, if I may, just a quick follow-up on the Scope 3 because obviously, in the absence of emissions target itself, on our side, we certainly welcome proxy targets if they're credible. And there, we note that you have this objective to have 75% of your top 50 customers set a 2050 net zero goal by 2025 and 100% by 2035. I'm just wondering how you I guess, how do you consider the extent to which you've actually influenced your customer to do so, just bearing in mind a context where increasingly large numbers of companies are setting net zero by 2050 goals. And if you ever consider also pushing them to set interim goals, which, as you know, are actually really important, right, because we do need emissions to come down soon and the shape of the curve in terms of emissions reduction really does matter in terms of overall emissions reductions. So interim goals are just as important as a long-term 2050 goal.

Yes, I couldn't agree more. And I totally agree as well with you on the pace at which decarbonization is moving forward, which, from our perspective, is as well much too slow compared to the need that we see. So fully aligned with what you're saying. Now how -- what are we looking at? We are, of course, following our top 50 customers in a regular basis and as well tracking their net zero objectives. We are looking as well into shifts between -- there was one shift moving -- some of them moved to 2030 targets because it's especially for the European ones because it's as well in the CSRD one of the KPIs that was requested. But we see as well some of them actually moving back or moving out and from the midterm, short-term targets and rather pushing it out to 2035. So overall, our mix of customers is evolving. We still see a good engagement in terms of net zero. And maybe I'll let Louis-Francois or Guillaume add a word on the customers.

Yes, I can do, Louis-Francois speaking. I think in terms of influence, what we can do is more to influence on the medium term. You know what I call medium term, [ 2030, '35 ] because this will be the term of contract renewals. So things we are discussing today or in the years to come. So yes, we can have an influence on our customers discussing techno solutions, decarbonization solutions for their long-term commitments, I think it's very difficult for us to influence because this is very much driven by their markets, their customers. And so let's be humble. I think our ability to influence is limited. So I would make the difference between the short to medium term and the long term.

We will now take the next question from the line of Siti Griffiths from Federated Hermes.

Just 2 questions and one of which is probably very basic. But just wanted to understand the difference between the baseline years that are used for the carbon intensity versus the carbon emissions reduction in terms of absolute terms, that 2015 versus the 2020 baseline. I wanted to understand that approach where you've used different years for that. And then the second question relates to the -- on Slide 7, where it talks about carbon capture and storage in terms of having only large sizes available for hard-to-abate sectors. Do you have any plans in terms of having small and medium-sized sort of solutions that still align with carbon capture and storage?

Okay. So for the objective fixing, maybe, Guillaume, you know the history. So 2015 was actually the year that was set as a basis when -- because we actually fixed that carbon intensity target before the advanced strategic plan. And that was enhanced and taken on continued along advance. Guillaume, maybe you want to say a word about the carbon intensity indicator.

Yes. Indeed, to make a bit of history, in November 2018, we were the first in our industry to announce the global climate objectives. At that time, we took the objective to reduce our carbon intensity, so CO2, Scope 1 plus 2 on EBITDA by 30% in 2025 versus 2015. So we took a previous year. Then in 2020, we developed a new set of objectives that were announced in March 2021, shifting to a global absolute emission reduction objective on Scope 1 and 2, so the minus 33% 2035 versus 2020 baseline. And in 2020, we decided to use -- in 2021, when we announced our objective, we used the 2020 baseline because it was the last year of reporting where we had reliable data. And that's a good practice in target setting for climate to use recent years. So we would not take an objective in 2021 on a 2015 baseline. So that's the reason why we have these 2 track. And basically, the intensity objective was instrumental in the group to onboard our operations and to understand how to manage global climate objective as a group that allowed us to take much more ambitious objective in March 2021.

And for the CCS, Guillaume, do you want to continue? Explaining the CCS projects, just perspective on those?

On CCS, if I get correctly the question is that today, indeed, we target large-scale project. And in particular, we target large-scale project in hubs in order, in fact, to build volume to be able to develop the infrastructure, which is the first step. Once we will have developed the first hub and the first infrastructure to export the CO2, then you can target smaller project. We have basically the technologies to do so, but it's more a matter of economics that to deploy the first of its kind project for CCS, it's easier economically speaking, to target large-scale projects because you have more volume, so you can amortize basically the capital on more volumes.

We will now take the next question from the line of Mary Gauthier from Comcast.

I had 2 questions. The first question is maybe if you could share a little bit more on your participation in carbon measures. Maybe what you find is missing today in current accounting protocols such as the GHG protocol on Scope 3 that led you to join this initiative and hence, what you expect out of it? What is your hope out of joining this initiative? And then the second question is more relating to Scope 1 and 2. On your 3 levers are very clear in the presentation. I was just wondering if you believe that the efforts to decarbonize are actually going to be harder in the future for lever 1 and 2, so on energy sourcing and asset management. And if you're expecting hence, lever 3 to compensate for lever 1 and 2, if the low-hanging fruits have basically already been reaped on lever 2 on ASU electrification.

Okay. Thank you very much. So I suggest I take the first one on carbon measures, and then I'll let Louis-Francois or Guillaume complete on the levers evolution. So maybe on carbon measures, I think we have already mentioned during the first question that was raised the context and actually we feel that the world is moving only at about half the speed it needs to meet the 1.5 degrees target. So for us, it's clearly not enough and it is not fast enough. What has changed as well significantly is the approach to energy transition. The strong consensus we were seeing 4 to 5 years ago has clearly fragmented. So in summary, things are still moving, but not at the same pace and with the same enthusiasm around the world. So to address the global warming challenge, we believe that manufacturing industries must decarbonize. That is a reality, and it will not go away. And as private sector players, we believe that we need strong, clear and consistent market signals to invest. And that's exactly what's behind carbon measures. Just as a reminder, carbon measures dual objective is, first, to advocate for product level carbon intensity standards as those critical climate policies as they are absolutely critical for climate policies to drive decarbonization. And second, to develop the carbon intensity letter-based accounting standard that is needed to underpin such standards and as such, ensuring actually the right accuracy of data. Now why are we a founding member? Well, it's because despite the climate frameworks and protocols that have been enacted across the globe, global emissions still keep growing. And that's where carbon measures is calling for a major shift in the type of climate policies to drive real actions, especially in terms of demand for energy and energy-intensive materials, it will continue to increase in the coming decades. So let's face the fact, demand for low-carbon product remains limited to small voluntary markets today because they often cost more to produce without offering a competitive market advantage. And that's where CA aims to fix it by creating market signals by which low-carbon products are legally differentiated and rewarded. And this is actually what we believe is essential to decarbonize especially the hard-to-abate sectors. So in a nutshell, we want to advocate for mandatory product carbon intensity standards with pragmatic technology-neutral regulations that allow to reduce emissions by establishing clear conditions for market access. So products to be differentiated based on their carbon intensity with manufacturers of products that have a carbon intensity above the mandated threshold having to buy credits from overperformance. This transfer then of value to low carbon intensity products will drive investments in low-carbon processes. And as such, decarbonization becomes a competitive advantage rather than a cost burden. And it would, we believe, unlock the demand necessary to scale technologies like CCS and low-carbon hydrogen. So we are today 25 members in carbon measures. We are -- as you probably know, we had the launch in November last year during the Sustainability Business Corp. And we are now building up the expert panel to work, especially on the ledger for carbon accounting as well as the product mandates. And here, we -- our intent is to focus on the 70% of products that actually have -- represent the highest emissions. So much for carbon measures. So now handing over to the levers.

Okay. So on the outlook for the transition plan continuation, first of all, I would like to recall that we have EUR 2 million of backlog of energy transition projects, so still to start up and that will bring emission decrease. We expect the low carbon electricity sourcing to remain strong, in particular, because we signed since 2021, a large amount of PPAs, 5.6 terawatt hour, and they have not all started and they are not all at full ramp-up, specifically in China and South Africa. So this will feed, I would say, the first lever of the transition plan. Second, asset management. So we have either electrolyzers to start up, either electrification project. We have one additional one in China. So that will feed the asset management lever, which is more a modernization, reconfiguration of our assets. And the last on CCS, so we have the Porthos project in the Rotterdam [indiscernible] that is due to start, let's say, in the year to come. So that will also feed the CCS lever. So I believe that in the next few years, we'll certainly have a more balanced contribution of the 3 levers. But of course, looking forward and beyond 2030, we will need to have more projects so that we can continue the decarbonization. But this is more or less -- these are more or less the perspectives for the next few years.

[Operator Instructions] We will now take the next question from the line of John Buckland from W1M Wealth and Investment Management.

Just some follow-up. On Scope 3, purchased goods and services, capital goods is important of sold products in the numbers. So I wondered if you could talk about suppliers as well as you've been talking about customers and the relationship there and what you're encouraging to do. But also, I just wondered when you make your assumptions about use of sole product, what assumptions are you making about the decarbonization of your customers' energy consumption? I mean -- or are you just saying we're assuming that where we are now continues in future? Because obviously, that makes a big difference if you're talking about long use products.

Yes. So actually, on the supplier side, we have very early on engaged with very concrete actions to drive down Scope 3 emissions coming from suppliers. That is a program where we had as well internally objectives fixed, and we are driving that forward. So you're totally right. Of course, we do not only focus on customers, but remain very active as well on suppliers. Now again, coming to your question regarding the customers, Guillaume, do you want to add some elements or Francois?

Or we can make a mix of answers. On the use of sold products. So at the moment, we are disclosing numbers for the CO2 that we sell as a product. We are disclosing numbers on the nitrous oxide that we are selling in particular to hospitals. And we are working also on fluorinated gases, but we are not yet disclosing numbers. You have qualitative disclosures in the URD. So for these -- any of these products, the Scope 3 that we disclosed is a combination of the global warming potential of each molecule multiplied, of course, by volumes, but using also an abatement rate because depending on the application, take the example of an electronic fab, you have scrubbing systems that are collecting the exhaust from the process. And so the molecules are being eliminated. So there is a strict and audited calculation that is, let's say, compiling all this data. Also worth mentioning, there is an important move within hospitals to move away from nitrous oxide for anesthesia. You have this big time in Europe, but also in the U.S. So our teams are working also on reducing volumes for this type of applications because you can find alternatives for the hospitals. And I think this is where we are in terms of use of products sold in the Scope 3. Maybe what is worth having in mind is that we have internal action plans to reduce some of these categories. So we are not necessarily disclosing objectives, but we have internal plans to work on these emissions.

What percentage of your revenue is covered by the -- by your disclosure and use of sold products.

The products that I mentioned are very marginal. The large -- I would say, the bulk of Air Liquide sales are hydrogen, air gases, oxygen, nitrogen. So there we are referring to CO2, nitrous oxide, fluorinated gases. This is very limited sales for the group.

Okay. And that's going to remain the case. There's no areas where you would make future disclosure is going to be a much bigger number.

Yes, absolutely.

May I ask another quick question on -- you talked about water and you explained that you've had a major reduction in China. Could you just explain exactly what -- how you've achieved it and how that may be can be used in other places?

Yes, that's the cooling water for the steam condensates. Any Air Liquide plant, we are compressing a large amount of air in particular. So this is creating heat and you need to cool down your machines in particular and the process air. And so this is why we need cooling water. So in the case of the Chinese plant, this was the water to cool down the steam condensates. And so any time we'll find this configuration, we'll have a very significant water withdrawal reduction. But we don't have this in any of the steam-driven plant. So that will depend on the asset-by-asset case.

So sorry, just to understand. So you said you've got 12 plants which use steam?

Yes.

But is that -- but you're also talking about cooling the condensate. So is that saving related to your steam ASUs? Or is it all -- is it independent of that?

Can you ask the question again? Maybe I'm not too sure.

Well, I'm just -- you're talking about using water to clean -- sorry, using water as a coolant. You're also talking about steam condensate.

Yes.

So the question I was sort of relating -- I was trying -- maybe this is wrong, but you have steam ASUs, right, which you talked about earlier. Is there a relationship between the 2, the water and the getting rid of the steam ASUs. But presumably if it's not, then every single ASU has a cooling requirement.

It's not the same water. You are right. Any air separation unit has cooling requirements. So we have cooling water in various process and form to cool down these units. In the case of steam driven, you have an additional steam, but steam is being recuperated through the condensate. So this is not creating any water loss, and it's not the same water.

Maybe if I may add, in fact, on steam-driven ICUs, in some cases, and it depends, in fact, on the technical settings, you have to cool also the condensate themselves to be able to recycle them back in the water system. So that's why on some specific steam-driven ICUs, you have higher water usage.

So you had a big -- so I'm just -- sorry, just being late at this point now. You said you had a big reduction in China. Can what you did in China be applied to other areas or I'm just being a bit slow. I didn't understand how that could be an initiative throughout rest of the world.

In other steam-driven ICUs, in fact, it will depend on the type of cooling circuit that is installed on these plants. The savings when we electrify a steam-driven ICU in terms of water depends, in fact, on the type of cooling circuit that is already installed as of today. And there are several types that have been designed at the buildout of the plant.

And the massive reduction does not apply to electrical drive air separation unit.

We will now take the next question from the line of Lionel Heurtin from Ofi Invest Asset Management.

I read in the Excel spreadsheet that 48% of your Scope 1 is covered by regulated emission trading scheme. I assume this is across geographies, not only the EU. How many million tonnes did you receive free of charge? Is the trend down? And how many million tons do you need to buy as the trend for the ton you need to buy be up? And what is the total cost of purchase? And what is the average cost per ton of CO2 emission cost you?

You are right that many of our assets that have Scope 1 direct emissions are covered by ETS, about half our emissions. We do not disclose the total cost or the average cost or the amount of allowance that we buy. But what I would like to remind is that as per our business model, this is passed through to our customers, in fact, the CO2 cost.

Because this is a good way for us to put a price on what is the cost of CO2 emission for Air Liquide, even if you pass it to your customer, we can still think it's a risk.

Actually the CO2 cost, Guillaume was referring to is passed through to our customers. So we have no direct risk exposure. We could say we have an indirect the competitiveness of the customer. But within our P&L, we are not bearing CO2 cost or risk because in the form of our contracts, environmental attributes at large are passed through to the customers. Then it's up to the customer -- to our customer to either structure it into its product price, and that's what they do usually.

We will now take the next question from the line of Tsitsi Griffiths from Federated Hermes.

I'm conscious that we've spoken a lot about carbon emissions and climate. Just wanted to touch on health and safety, just looking at the lost time accident frequency rate and how that's come down as well as other metrics looking at fatalities and other metrics as well. Just wanted to understand because there still seems to be a gap that persists between Air Liquide's own employees and subcontractors. So 0.4 for the company's own employees, but you're looking at 0.7 for subcontractors. So I just wanted to understand why that gap still persists in terms of that accident rate and if there might be a difference in how health and safety is perceived by subcontractors that work for Air Liquide?

Yes, you're totally right. So first of all, safety is our first and foremost priority, and it is clearly very strongly driven forward for our employees, but as well for our -- for temporary employees and subcontractors. So the engagement that we have is really about all individuals and the actions that we take, may it be on the culture, may it be in terms of system and procedures. And of course, ultimately, in terms of mindset, we include the subcontractors into that. It is for us as well a license to operate. And this is where in our procurement initiatives, safety is part of the selection criteria for subcontractors. Nevertheless, it's very often separate companies who are actually employing our subcontractors. And of course, they might have a very different view on that. That's what we try to limit or we try to limit the negative impacts of accidents on our subcontractors by selecting the right subcontractors who are compliant with our expectations. And then, of course, to make sure that wherever they operate, if they operate on our site, that they are well trained on our processes and on our procedures on how we would like to operate. We have as well introduced what we call the stop work authority, which is basically giving every employee, every operator, internal or external, the right to stop work if they perceive a major risk. And this is clearly something that we are pushing forward to all our operators on site, may it be our own employees and our subcontractors. But you're fully right on subcontractors, we still have room for improvement. And of course, let's be very clear, our ambition is still 0 accidents. So even for our employees, despite the fact that we have made tremendous progress, we are not yet there where we would like to be.

I would now like to turn the conference back to the Air Liquide team for closing remarks.

So I would like to thank all of you for your participation and the active questions. It has been a pleasure sharing those elements with you. Thank you very much for your attention, and I suggest we close now the session for this year. Thank you very much.

This concludes today's conference call. Thank you for participating. You may now disconnect.