Welcome to this presentation of our half year results for H1 2021. Our performance is robust and promising, riding on the back of excellent operating performance and a stronger price environment. We're optimistic when it comes to the next few months. About this presentation, I will start with a short introduction, then I will share with you the progress that we've made in terms of our CSR roadmap. Corporate social responsibility is gaining pride of place in terms of our general roadmap. Our CFO, Thomas Devedjian, will present our financial performance, and then I will be back to discuss our operating performance as well as the inroads we have made in terms of our strategic roadmap, and then the conclusion. By way of introduction. [Non-English content]. Obviously, let us zoom in on the performance of our first half. Strong performance. Revenue is up 11%, so close to EUR 1.9 billion.
82% of that revenue comes from M&M, mining and metals. Manganese accounts for 47% of that business. EBITDA is up strongly, so 2.5 times what EBITDA was in H1 2020, so close to EUR 300 million. Our operating income is close to EUR 160 million. Net income group share is EUR 53 million. Our free cash flow is really good. As you know, there is an unfavorable seasonal effect in terms of our mining operations in Q1. Despite that, there was excellent cash generation, so EUR 111 million in cash generation. This means we've been able to reduce our net debt, and we're now back on our bank covenants. Within the meaning of our covenants, our gearing has been brought down to 92%. Excellent performance, as I said, which is first and foremost the result of our inroads.
The inroads we have made in-house, our intrinsic performance accounts for over two-thirds of our improved EBITDA. As you can see, our intrinsic performance exceeds EUR 110 million, this mostly ties in with the growth in volumes and also cost adjustments, particularly within A&D. This is true despite the extremely poor operating conditions in H1 in New Caledonia. By and large, we've made excellent inroads within the group. Markets have kicked back up, which means that prices have kicked back up in most of the countries where we operate, but not all. This is one specific feature of our group. As you will see, manganese ore prices have not increased. The economic recovery has also led to a strong increase in the price of some of our inputs, including freight costs.
Obviously, this has weighed down our cost structure, and also the dollar has depreciated against the euro, which has affected us. By and large, our external factors have improved our EBITDA only to the tune of EUR 60 million. EUR 60 million is something, but it's not as much as the progress we've made in-house. As I said, positive free cash flow of EUR 111 million with a high contribution from M&M, Mining and Metals Division. Also, this is due to the fact that our HPA division has strongly reduced its cash burn relative to last year. Cash burn has dropped by more than three-fold. Net debt is EUR 1.2 billion. Also we have continued to make significant inroads in terms of our CSR roadmap, which is really important to the group.
Getting back to our economic circumstances, it's important that we drill down on that. We have enjoyed a favorable price environment this half year, except for manganese ore, which, as you know, is the most important commodity for our group. Flat manganese ore prices relative to H1 2020, +2% in dollars, but -7% in euros because the dollar depreciated over the period. However, we have a strong rise in manganese alloys prices over the half year. I'll get back to that later, the increase range is 20%-30%. If we look at manganese alloys alone, the impact is in excess of EUR 64 million. Manganese prices are up and also nickel ore. Because there's a shortage of nickel ore on the market at the moment, prices are up significantly, up 40%.
By and large, the price environment has had a positive impact to the tune of EUR 190 million, which is a considerable impact despite the fact that manganese alloy prices stayed flat. Input costs have increased significantly, including freight unit costs. We're talking EUR 55 million, and we're expecting this trend to continue to go up significantly over the next few months. Other input costs, such as raw materials and energy, those input costs are higher to the tune of EUR 24 million. There's an unfavorable Forex impact, EUR 45 million. When you add it all up, the impact of net external factors is favorable, but only EUR 60 million over the half year. As we will see in our forecasts, we expect a much higher impact, a much more positive impact of external factors in H2.
In terms of our different businesses, M&M is still on an upward track. Operating performance is improving, except in New Caledonia. Very poor weather conditions there, difficult operating conditions earlier this year. If we zoom in on manganese, organic growth continues at a steady pace. I'm referring to the Okouma mine in Gabon, so up 13%. We're talking 3.1 million tons of ore produced in Gabon. We shipped only 2.9 million tons of ore out of the 3.1 million tons, and this compares with a very high level of shipping or transport. We actually reduced inventories by transporting a lot in H1 2020. This year, passenger traffic resumed. I'll get back to that in a minute. We also had railway problems in Q2, which means that transport was impacted. Like I said, the baseline that we're comparing ourselves with is very high.
Manganese alloys drove this trend up 16%, and the product mix is much more favorable. In terms of nickel, there are two very different situations. Weda Bay, on the one hand, is in top form, enjoying remarkable growth. In H1, the mine produced seven million tons of nickel ore, which is twice as much as what we produced over the full year last year. A huge ramp up, an extraordinary ramp up of mining operations at Weda Bay. The plant is operating at full capacity. The plant developed 20,000 tons of NPI, or nickel pig iron, in H1 this year. This is in line with its nominal capacity. The situation is very different for SLN. You may remember that SLN started the year in very difficult circumstances.
The mines were blocked earlier this year and at the end of last year, at a time when we're supposed to put together inventories at the plant for the dry season. Because of those blockades, we have been unable to replenish the inventories necessary to cope with the rainy season. This was true for all Caledonian operators. The rainy season was extremely unfavorable. Rainfall was twice what it usually is, and the rainy season lasted twice as long as it usually does, up until June. Operating conditions were very difficult. This means we weren't able to increase mining production as much as we wanted, but it's still up 5% despite the poor weather conditions. Exports are slightly up 2%. It's mostly the plant that suffered quite a bit because it was purely fueled and supply was poor, inventories were low. Ferronickel production is down 22%.
In terms of mineral sands, we still operate at a very high level, hitting operating performance records at GCO. By increasing efficiency, we're now able to maintain a production despite the lower grade of deposits. Also operational performance is strong at GCO. We're producing quite a bit of CP slag at TTI Norway. We're also making inroads in terms of our corporate social responsibility roadmap, which is very, very important. It is core to our model. It's also very much in line with our new corporate purpose, which has now been enshrined in our statutes, in our articles of association. CSR has long been at the heart of our roadmap, and our roadmap is based on three main pillars. We are a corporate citizen committed to a sustainable future.
We have a commitment to people, to the planet, and to a responsible economy. 13 objectives for these three pillars. With quantitative operational indicators, we have made a lot of inroads, made a lot of progress in terms of these 13 indicators. For two years now, we have onboarded the indicators of this roadmap into our LTIP, our long-term incentive plan for the group's executives. Also, we recently onboarded two CSR indicators or targets in our bank documents. I'm referring to CO2 intensity and the need to reduce it, and also the need to reduce our LTIR. I'm now going to review every single one of these 13 targets. What's really important is safety. Safety of our employees and of our contractors. We're still getting better on this front. The TRIR has dropped by 40% relative to last year.
Since 2016, there's been a six-fold reduction in the TRIR. The dynamics are really good. This is true for all of our peers, and it's kind of a difficulty for the entire sector. We're still experiencing severe accidents, sometimes even fatal accidents among our contractors. We have a strong focus. Our management is strongly focused on daily operations for safety routines. We pay close attention to the safety of all our employees as well as all of our external providers and contractors operating on site. This is quite a challenge because we have a lot of subcontractors. In those countries where safety standards aren't always upheld, we need to move forward and make things better. That's going to be a challenge for the group over the next few months.
If we look at our other key performance indicators in terms of our CSR roadmap. Let me say a few words about the climate. We're not just on track, we are ahead of schedule in terms of achieving our 2018-2023 CSR roadmap. We were planning to reduce our product CO2 intensity, last year we increased our climate-related targets. These targets have just been approved by the SBTi, the Science Based Targets initiative. We have received validation from SBTi. We have a credible roadmap. We're planning to reduce by 40% our scope one and two CO2 emissions in absolute terms by 2035, also achieve carbon neutrality by 2050. In terms of the communities, we have made a lot of progress. We have set up two CSR funds working together with Gabonese authorities, and they are working really well.
In H1, they focused quite a bit on infrastructure such as access to water, roads, et cetera. We also broke ground on the very first biodiversity foundation in Gabon. It is called the Lékédi Biodiversity Foundation, an Eramet foundation. Also, we try to maximize our positive impact on local communities. We create jobs around our activities, and we also make positive contributions to healthcare and education. These are strategic priorities in our CSR roadmap. I'd like to remind you that 100% of our mining sites enjoy ISO 14000 certification. Now, our CSR performance has been recognized by non-financial rating agencies, and our ratings are improving year-on-year. We always stand on the podium. Look at our different ratings. We have achieved a score A with the MSCI corporate rating. This means we rank among the top 30% of mining and metals operators.
In terms of ISS ESG, we ensure a prime status. We are part of the first decile in ISS ESG mining and metals. We enjoy a score B, but we also rank among the leading companies in the sector. I'm referring to the Carbon Disclosure Project, climate change targets, and also Vigeo. We also stand on the podium here. We are third out of 44 companies in this panel. We have achieved an advanced level. We have a score of 66 out of 100, and we are very proud to have received external recognition for the progress we have made. Once again, making such progress is so important to us. It is part of our strategic model, it's part of our management priorities. Now, our financial performance. Handing over to our CFO, Thomas Devedjian.
Thank you, Christel. Good morning to all. Results that are strongly up versus H1 last year, as Christel indicated. Sales up 11%, EBITDA risen up over 150%, close to EUR 300 million as against EUR 120 million last year. Income group share, positive EUR 53 million. Last year, we had a major depreciation. Net debt down by just EUR 100 million. Our gearing covenant, we asked for two waivers in June and December last year from our banks. Our gearing, under the definition of our documentation, falls below 100%, no need for a waiver. ROCE goes from 3%- 11%. These are metrics that are all improving. Group EBITDA is sharply improved. We have both Mining and Metals divisions that is up strongly. EUR 337 million EBITDA alone. We were at just over EUR 200 million last year.
First off, we have manganese production up sharply, and also the price of manganese alloys that are rising sharply, set to continue, and we're going to benefit fully from that effect in H2. Very strong performance of mineral sands, our mine in Senegal and our plant in Norway. On the nickel front, we're sharply reducing the nickel loss, primarily thanks to improved nickel prices, even if we suffered, as Christel said, from particularly poor weather conditions this half, with exceptional rainfall in New Caledonia. Lastly, our high-performance alloys division, that is to say, Aubert & Duval, Erasteel, dividing its loss by six. We go from EUR -66 million- EUR -10 million EBITDA loss. That's a strong improvement. This EBITDA improvement comes two-thirds from our internal performance and one-third to external performance. Internal performance shows an improvement across our businesses, except for SLN, where ferronickel volumes are down.
If we leave aside that drop in ferronickel, we're up across the board with volumes that are growing. Aubert & Duval, significant cost reduction and activity improvement. The external factors, well, in manganese, essentially an increase in the price of manganese alloys, and we have this strong increase in nickel price. However, we are suffering from the increase in cost of freights across our seaborne operations. Maritime transport costs are a lot more than sharply up, so that deteriorates our EBITDA by EUR 55 million, and the slide in the dollar acts in our disfavor, because we went from about $113- $123 rate, so that penalizes us for about EUR 45 million negatively on EBITDA. Net income group share EUR 53 million. Last year, we booked, as said, in performance for EUR 426 million and the lithium mothballing costs this year.
In terms of one-off events, we have the cost-cutting reduction plan at Aubert & Duval with a charge of some EUR 20 million on that, and the benefit, and that's consolidated at equity. The performance of Weda Bay, that impacts us very favorably, very positively. That's share in income from associated companies, EUR + 77 million on the half. Of course, we have our taxes paid in Gabon and Norway and the minority interest, Comilog positive and SLN negative. Turning to our industrial CapEx that are really under control in 2020. We slowed the pace of CapEx, and that was felt fully in Q1. We had a reduction in CapEx through H1 2021. On growth CapEx, and I exclude lithium here, that was stopped last year. We're about on a par with last year's level, with a split between the growth of Comilog.
To increase Comilog volumes, we need to invest and, of course, the renovation of the railway in Gabon. Once we produce the ore, we need to transport it. We have an investment plan on the railway. Current CapEx down 32%. There will almost certainly be a catch-up significantly in H2, where we'll return to more normal CapEx levels. This leads us to revise our CapEx guidance down. We were at EUR 500 million at the start of the year. We're revising our CapEx guidance down to between EUR 400 million-EUR 450 million. The strong free cash flow generation over the half, EUR 111 million, to be compared with a cash burn in the first half of last year of EUR 210 million, is obviously due to the strong performance of the Mining and Metals Division, but the sharply improved performance of assets that underperformed last year.
That's to say SLN and to a far lesser extent, Sandouville, SLN, essentially thanks to price improvements. On the high-performance alloys front, we divided the cash burn by more than three, we have about EUR 50 million of cash burn on HPA, whereas last year we were EUR 165 million. A sharp improvement there. Of course, the major contributors remain manganese and the mineral sands, to a lesser extent. Lastly, we see appearing a third big cash generation engine, which is Weda Bay, that generates EUR 70 million of free cash flow over the half. Our growth plans as well as addressing difficult assets is paying off, and it's reflected in the free cash flow level. Where does the improved net debt situation come from? Of course, first and foremost, from the improved operating performance.
The Mines and Metals Division generated EUR 250 million operating cash flow, whereas the High-Performance Alloys Division consumed some EUR 40 million in operating cash flow. On both divisions, significant improvements have been recorded. Next we have the Weda Bay dividends, because we're a minority shareholder to the tune of 43% in a company that holds Weda Bay 80%, because we have an Indonesian partner. We're redistributing the results generated by Weda Bay between shareholders, very regular basis. That's the bulk of the improvements recorded. Our cash position remains very strong. We were at over EUR 1 billion at the end of December, and at the end of June, we were over EUR 1.9 billion gross cash. In July, we did two things. Given this improved cash position, we seized the possibility of an early repayment of the TiZir bond, a coupon of 9.5%, very costly.
We reimbursed it a year early. We reimbursed EUR 240 million in debt that was costly. Furthermore, we reimbursed over half the drawing of our RCF line that was drawn in total at EUR 980 million. In July, we repaid EUR 500 million of that. The pro forma liquidity of this line that remains drawable is at over EUR 1.7 billion to date. No major debt maturity short-term, given the TiZir bond early repayment. The major milestones are in 2024, no significant maturity before then. Our net debt including IFRS 16, is of the order of EUR 2.5 billion at June 30th, and at about 80% of which at fixed rate, excluding the RCF. Back to Christel.
Thank you, Thomas. Let's look at our operational performance, starting with Mining and Metals, M&M, and our biggest business, which is manganese. The underlying factors include carbon steel. Output in H1 boomed up 13% around the world in H1. This ties in with the strong recovery in the economy, so up 12% in China, which accounts for close to 60% of global carbon steel output. That's because the automotive and construction sectors in China have been extremely dynamic. Europe and the U.S. are up between 16% and 18%, starting from really low levels. They have yet to return to pre-crisis levels, but the recovery is extremely strong there. The downstream segment is extremely dynamic. Global manganese ore consumption is up 10%, so 10.5 million tons. Manganese ore output followed, not necessarily at the same level.
It is up 9%. We're talking 9.8 million tons. As you can see, there's a slight shortfall between ore production, ore output, and ore consumption. In gray and blue on this curve, you see the ore inventories, which reached record highs, and as a result, the deficit had very little impact on prices. Manganese ore prices remained relatively flat, that's the blue line on the graph. $5.1 per DMTU in Q1. Barely 2% more in dollars than last year and +7% in euros. Inventories are slightly down but remain relatively high, 11 weeks of consumption. On this side, however, you see that the red line, that's the manganese alloy prices, soared because of strong demand and a shortage in supplies. Problems with capacity due to COVID and inventories went down throughout the value chain.
It's hard to get the entire chain working again. The market is undersupplied, which has pushed prices up. As I said before, refined products prices are up 30%, and commodity prices are up about 25%. What do we expect? We expect prices to remain very high over the next half year. I'll get back to that in a minute. Price levels have already been established in Q3, and they look really promising for Q4. If we look at our own performance, as I said before, we're still growing our manganese ore output, up 13% in Gabon. I talked about the transport situation, which was slightly more difficult but remained higher than last year. While last year's baseline was relatively high because of COVID, passenger traffic was brought to a standstill, which freed up a lot of capacity for freight transport on the TG railway.
Our baseline was already high, and there were railway problems in Q2. As a result, only nine million tons were produced, or rather, were shipped. In terms of the railways, significant works are being done at Setrag level in order to improve reliability of the tracks and increase transport. The seasonal impact is getting better. This is the dry season at the moment in Gabon, and as a result of our modular investments into the mine, which will improve output levels due to the favorable seasonal impact and also progress at Setrag. Our ore production target is seven million tons for the mine, and at the very minimum, we're targeting 6.5 million tons of transported and shipped volumes. Volumes transported and shipped by Setrag. As you can see, we're already seeing improvement in transport levels in June and July.
Now, there is one aspect that had a significant impact on our ore business, freight. At this juncture, freight is up significantly, and these are the fundamentals of our own transport costs. I'm talking about freight costs per se and fuel costs. These are the two aspects that we need to address in terms of our shipping costs. As you can see, the blue line is up, fuel cost is up 22%, and freight costs themselves, if we have to ship by road, West Africa to China, this is up 153%. You can see the curve, the line is growing throughout the half year. Levels will remain high in Q3, and the impact on H2 will be significant. Regarding H1, by and large, when we factor in these two aspects, transport costs are up 60%.
If we look at manganese alloys, there are two positive impacts. First of all, price stability in terms of manganese ores, and also selling prices have surged. Obviously, this is good for our profit margin when it comes to our alloy businesses. We're going to maximize that by improving output. Our capacity, we usually operate at full capacity, but we've been able to improve output by 7% from one half year to the other. We've also been able to improve the most profit margin, has jumped from 49%- 55% of the total. That's a 20% improvement from half year to half year when it comes to manganese alloys. As you can see, we're seeing on the right-hand side the price of alloys that is soaring and the price of ore that remains relatively flat.
This trend continues into Q3. We've already saw the volumes in Q3, and prices are very high, and Q4 is looking pretty good. We expect this maximum margin impact to continue. We hit a record level at the end of Q2. Now, nickel. Again, the market fundamentals are excellent. stainless steel output has soared across the world, up 28%, which is amazing. It is true that stainless steel was hard hit by the COVID pandemic in H1 2020. Demand for primary nickel is up by 28%, and primary nickel output is up by only 12%. 12% isn't bad, but only 12%. This is mostly driven by the rapid growth in Indonesian NPI, nickel pig iron output, so up 81%. By and large, overall, throughout the world, if we look at Indonesia and China, NPI accounts for 50% of nickel output globally, which is a huge share.
The fact that demand is so strong, and the fact that output is keeping up, but not at the same level, there's a 66,000 tons shortfall in the first half of the year. This has led to a 40% increase in nickel prices, LME prices, and also a reduction in listed inventories. Ferronickel prices are also increasing by 43%. In terms of LME and SHFE inventories, we're talking more than nine weeks consumption declining. There's something you need to bear in mind. For about a year now, ferronickels or ferronickel alloys have been sold at a discount relative to LME prices because there's a glut. There's a glut of ferronickels on the market. Much NPI is available, and even though ferronickel prices have gone up, there's a discount, and you can see it on the screen.
The blue line is the LME nickel price and the red line, that's the NPI price. That's a listed index, we're able to report it. As you can see, there's a significant discount. Even though they follow the same trend, there's still a gap between the two lines. This discount was particularly strong. See the growing gap in Q1. We're expecting this discount to wane in H2, we believe this trend is going to continue for some time still. Nickel ore is also at high levels, growing from one half year to another, up 40% almost. Very high levels, $100 approximately per ton. This level was reached late last year. Ore stores in Chinese ports have reached historically low levels, this is due to the unfavorable seasonal impact both in the Philippines and in New Caledonia.
We bore the brunt of that, but at the moment, there is a shortage of nickel ore on the free market. Today, inventories come to one month's consumption, which is very low, and this is why prices are up that way. Unfortunately, we at SLN didn't fully enjoy those benefits. As I said, by way of introduction, SLN was severely impacted by the blockades in December 2020 and early January 2021, which is when we put together inventories and we increased our stocks to cope with the rainy season. We started with very low inventory levels, almost nil, and the season was disastrous, and everybody suffered. As a result, our mining output is up by only 5%. Up 5% versus 16% last year. It's still an increase, but much lower than last year. Also this affected plants, which received very little ore and low-grade ore.
Plant output volumes are down 22%. Obviously, this exploded the cash cost. As a result, the drop in ferronickel, we're talking EUR 0.70. Also, a negative Forex impact. The dollar depreciated relative to the euro, and this accounted for EUR 0.50 in terms of cash flow deterioration. The situation is currently improving. Finally, the dry season has arrived. Since the last two weeks of July, SLN has been operating pretty much normally, and we're expecting the cash cost to significantly improve in H2. This being said, the future of SLN is still undetermined. There are things we still need to secure. First of all, the export authorization. As you know, we have a permit to export four million tons of ore, and we're close to saturation this year. If it hadn't been for the disruptions, we have been pretty close to four million.
Right now we're maintaining 3.5 million tons in exports, and we're expecting to clear that bar next week. We have not received that authorization yet. We only have for about five months. Since February, well, the previous government fell, and the new government was put together only mid-July. Now finally, New Caledonia has a government that can approve that additional export capacity. We have resubmitted our request. We have called upon the new government's sense of responsibility because this request is absolutely vital in terms of SLN's future. Also, reduction in electricity costs. That's something we need to do. We have plans to build a new power plant for Doniambo. The consultation process is underway, things are proceeding at pace, and we are expecting firm bids that we will look at in Q4 2021.
I would like to remind you that when it comes to export authorizations, considering current prices, one million WMT of nickel ore means an additional contribution of EUR 30 million to EBITDA. Nickel is very successful thanks to the market conditions looking up. Weda Bay, we're very happy with our nickel business in Weda Bay. We beat record mining production levels, and I would like to remind you that Eramet operates this mine as part of a JV with Tsingshan, a Chinese group. Eramet operates the mine, and the plant is operated by Tsingshan. It's a joint venture. We're maintaining record production levels of seven million WMT of nickel ore produced in H1. We're supplying not just our own plant, but all of the other plants that are currently being built.
There are six of them now that are part of the Weda Bay Industrial Park downstream from the mine. These are operations that are extremely competitive with high contributions. Metallurgical operations as well. The plant is operating at full capacity. Needless to say, considering this excellent performance, we are revising our guidance. 12 million tons for the year. We already secured seven million in H1, but Indonesia is in the middle of the rainy season, which may slow down production a little bit. We're pretty confident that we will exceed the 12 million tons over the full year. I'd like to remind you that cash flow from Weda Bay is EUR 57 million, as Thomas said, in terms of dividends paid, and the rest is the trading business because demand for zircon output increases at a slower pace than demand. This leads to a market deficit in H1.
This was not really felt in this H1 because the baseline last year was very high. It's a market in which you negotiate prices on a quarterly basis and sometimes on a half-year basis with some customers. There is a lag between market price rise in Q3. We're looking forward to excellent zircon business in H2. Same thing for TiO2, titanium dioxide, manufactured in our plant in Norway. Here again, prices do not reflect the market's positive trend. They are expected to increase in H2. Demand is good. Supply does not match demand. Our operations are working really well. I won't get back to that. GCO is beating records year-on-year in terms of OEE, overall equipment efficiency.
This is an opportunity to largely offset the fact that, obviously, when a mine starts operating, it starts operating the richest deposit and then moves on to less rich areas or deposits. Output is good, production is good at TTI in Norway. We fully feel the positive impact of favorable market conditions. Moving on to HPA, high-performance alloys. The market conditions for A&D is obviously very different. We're still bearing the brunt of the aerospace crisis. There's a huge decline on the aerospace market. Air traffic is only 60% of what it was prior to the COVID crisis. In H1 last year, 70% of A&D sales was aerospace, and now it's only 58% in H1 2021. Despite the good news regarding single-aisle aircraft, that's good news for the future. It's not yet reflected in our order book.
There may be good news in terms of single-aisle aircraft. The long-range aircraft accounts for 70% of our A&D aerospace sales. As you know, long-range aircraft are struggling. We're not seeing any recovery anytime soon. Defense, nuclear, and energy, those markets are improving. This helped offset the aerospace crisis. If we look at what this means in terms of sales and EBITDA for A&D, there's a strong contrast between aerospace sales, which are down 24%. 52%, actually, when you compare it with H1 2018. It's a very strong drop. Again, a 24% drop between H1 2021. The beginning of the year wasn't impacted by the crisis yet in H1 this year. However, energy and defense sales are up 74%. In large, sales are down by 9%. The division has worked really hard to address the cost structure.
This is starting to pay off. A 35% decrease in commodity costs. Personnel costs are down 15%, and this has led to EUR 33 million in savings on labor costs. By and large, our EBITDA is still in negative territory, but much less so than last year, EUR -40 million versus EUR -52 million last year. Free cash flow is still in negative territory, EUR -38 million. This includes the sale of a small affiliate, a small A&D affiliate called Brown Europe. There was a positive impact of that sale of EUR 30 million. A&D suffered from the need to upgrade quality processes. This was extremely time-consuming. This hampered production. We're finally seeing the light at the end of the tunnel, and by the end of the year, we're hoping to be done with that. We're hoping to be done with this quality processes review process.
Hopefully by the end of the year, we will be out of the woodwork. When it comes to Erasteel, strong momentum in terms of sales. This is due to the market, but other factors as well. Sales are up 20% from one half to the other. The underlying markets, mostly automotive, increased less than that. There's an underlying market on the one hand, but also shareholding acquisitions, and this means a significant increase in sales. EBITDA is back in positive territory. Free cash flow is still in negative territory, but this is due to the very strong increase in sales, which has increased the working capital requirement despite strong efforts to reduce the number of days of revenue. The reduction is 38 days. Okay. Despite the -38 days of sales improvement.
Strong momentum in terms of Erasteel's recovery, and we're confident when it comes to what will happen in the next few months. So much for our operating performance. Let me say a few words about our strategic roadmap. We're still making inroads. There are three pillars in our CSR roadmap. Pillar number one, as you well know, we need to reposition or fix our least performing assets. As Thomas rightly said, we have made progress in terms of improving EBITDA of such assets in H1. We have almost reached the given point. That's not our final target, but we are on track. SLN, I talked about it, but SLN's business model has proved relevant. In H2 last year, and the prices weren't what they were today, and yet SLN made money. We need to manage to operate SLN properly.
New Caledonian authorities hold all the cards. The ball is in their court. Management has done its job. Eramet has supported its leadership team, its business model. It has funded this subsidiary. For some time now, we said that we would not reinvest into SLN. SLN now needs to fend for itself. It needs to secure its own future. We hope that the New Caledonian government will give SLN the wherewithal that it needs to implement this new model, which is fully efficient. When it comes to A&D, selling this asset remains the preferred option, and we are currently working to make this happen. This is something that we are focusing our efforts on. In parallel, of course, we're adjusting A&D's cost structure to match the market situation so as to improve its performance and reduce its cash burn. The Sandouville Refinery.
A decision has been made to divest this asset. We discussed this in February already. We are in the advanced stages of negotiation with a potential buyer, and we're hoping that the divestment will be finalized sometime soon. When it comes to Erasteel, I talked about the recovery process, the strategic review. We need to see how we can reposition Erasteel outside the group, and obviously, that's not our top priority right now. When it comes to the middle pillar, cash generation and the need to grow our most attractive businesses. Now, this is fully in line with our roadmap. We're actually ahead of schedule. Manganese ore is up, manganese ore output is up. Weda Bay is beating all output records and all are exceeding expectations. Mineral Sands are on track to meeting their targets. We've made a lot of inroads regarding the cash generation priority.
When it comes to pillar number three, expansion. Expanding our portfolio to include metals for the energy transition. Last year, we mothballed the lithium project. We're currently looking at how we can restart that project because there's strong demand for lithium and prices have increased significantly. We need to do this in a way that won't weigh down our balance sheet. No decision has been made yet when it comes to the lithium project, but clearly, this is an excellent project for the group. If we are able to restart that project without weighing down the group's balance sheet, we will do so. We are moving forward in terms of a partnership with BASF. Pre-feasibility study is underway in terms of diversifying Weda Bay operations to include the cobalt salts in particular, and also lithium-ion battery recycling.
The R&D program is ongoing. 2021 will be a key time for repositioning our pillars. We will continue to make progress until the end of the year. We've made excellent inroads in terms of pillar number two. This has much improved the financial situation so that gradually we can restart working on pillar number three once all the conditions have been met. By way of conclusion, thank you for your patience. I know these are long-winded explanations. We created strong momentum in H1. Market conditions are positive. Prices are up pretty much across the board in H2. There's a favorable seasonal impact in H2. We expect to deliver excellent performance throughout the year. Our plans for organic growth and the favorable seasonal impact are expected to boost the mining output in H2. Weda Bay is expected to continue delivering high output and cash flows.
Manganese Alloys will operate at excellent margin levels in H2. Pillar one, that's something that we are currently addressing, as I said before. As a result, we have confirmed our volume targets for ore manganese, so seven million tons to be produced, and also 3.5 million WMT of nickel ore exports at SLN. We have increased our output targets for Weda Bay. Over 12 million WMT of nickel ore produced at Weda Bay. Our EBITDA target is above EUR 850 million in 2021, considering this buoyant price environment, both in terms of market conditions and operations. That's about it for us, and of course, we are now on hand should you have any questions.
We'll take our first question from the room. There's a question there in the center. [Non-English content]. My question, you have EUR 27 million of non-recurring costs. Could you detail their utilization? I also had a question on Weda Bay to know how far you can go in terms of mining production without having major investments. Then far more forward-looking on the battery, there's more and more talk about replacing nickel and cobalt with manganese. There are batteries based on manganese. You working on that? That's a great many questions. I hope Thomas jotted those down because I didn't have my pen with me. First question, because the microphone wasn't switched on in the first part. First question was on manganese alloys production lower than in Q1 versus Q2. There are two reasons for that. We stopped a furnace.
I mean, that was programmed, that furnace stop because we have to realign our furnaces on a regular basis. There was a program furnace shutdown that occurred in Q2. The mix is important. That's to say that last year and early in the year, we did a lot of commodities. That's a lot of volume, not necessarily margins. Refined alloys, perhaps a little less volume, but more margin. We really want here to push when we can the mix to refined alloys that generate far more than commodities. That's in response to your first question. Second question concerned how far we can go on Okouma without using the track. Oh, sorry. That's without the washing unit. Kleber, can you speak to that?
Well, for Okouma, actually, there's no limits. That is, we can continue have the dry processing and the wet treatment goes with it.
That's the strategy that delivers the seven million tons and going forward. First dry treatment that is maintained. That really is the bulk of our increase in the modular washing units that will come on stream, the first at the end of 2022. Well, in fact, we can ship everything that is dry process, store what has to be washed, there are limits to what we can store. When the washing units are there, we process the portion that must be concentrated and wet processed. There was a question for Thomas. Yeah, there was a question on the non-recurring costs, if I understood the question here. Other operating income and expenses, the minus EUR 27 million, is that what you're referring to precisely? The detail is essentially, as is put in the note, the Aubert & Duval adjustment plan for just over EUR 20 million.
There are the remaining costs of the lithium project. That's the bulk of it. There was a question on Weda Bay, how far we can go without expanding major CapEx. We can go quite far is the answer to that. I mean, it's a mine opening, so the CapEx is building roads, exploring mining planning, et cetera. We're going to continue to make progress without major investment, especially given the size of Weda Bay. I mean, they're really quite minor investment. Just to follow up on that point. The fine thing is that as Christel said, the investments are marginal. It's a huge mine reserve with a profile very close to the surface. That is that these mines that give tonnage with not big increase in coverage rates and huge investments. We can go quite far. Excellent quality, very good source of reserves and the associated costs.
I mean, the profile is very good. The final question on the batteries and battery technology today. I think we have Philippe in the room with us who'll be able to Philippe is our head of strategy, business development, innovation. He's really on top of these issues. The various battery technologies, there are a lot of R&D worldwide, and there are lithium-ion batteries. Lithium, some non-lithium, the takeaway number is between 20 and 25. We take into account the various battery technologies, global lithium consumption will increase five-fold. For the cathode material, nickel, cobalt, iron, manganese, we plan a mix of various battery technologies, depending on the utilization capacities and the performance batteries.
LFP, lithium iron, that's developed a lot in China for buses, for e-bikes, and conversely, in Europe we're seeing the NMC, nickel manganese cobalt technology, but little lithium manganese alone for reasons of battery performance and all that. When we project the future mix in light of those various techs into 2030, nickel consumption will increase six-fold between 2020 and 2030. Cobalt consumption with a very cautious assumption on cobalt will increase four-fold. Let me reassure everyone, it's not times six global nickel consumption, it's nickel consumption in batteries. Steel will continue to represent a bulk of that. Just to conclude on this, we do factor in our forecast and the various players, the OEMs, automotive or upstream for the metals for that. We take account of all these mixes for various batteries. I think we've answered all your question. There's another question in the room.
Yes. Yes. Hi. Alain William from Oddo. I have three questions, if I may. The first, I'd like to know the procedure of amicable settlement from SLN. Do you have an update for us on that? The second concern, the reduced CapEx guidance. That's pretty significant. Just wanted to know where it comes from, also, what is the level of maintenance CapEx of the group to distinguish the two buckets. The third question on Aubert & Duval. I'd like to know why you rejected the first offer. The press states that discussions have resumed. What's changed, can we expect a deal by the end of the year?
I'll answer about the amicable settlement on Aubert & Duval, Thomas will speak to the CapEx. Amicable settlement stopped. It was halted in May because an amicable settlement is limited in time.
It lasts four months. In fact, it started at the end of January, aimed at attempting to move the dial in terms of the local authorities because the conciliators clearly saw that the real drivers to improve the SLN situation wasn't really in the hands of management, but rather in what authorizations via power. Conciliators were faced with an absence of government throughout the amicable period. They closed it in May without really. There was an ad hoc mandate where a diagnosis was made, everything was shared with the locals, but the amicable settlement procedure was unable to unfold because of absence of negotiating partners. Doesn't mean it won't be resumed. The price has risen. SLN, even if it's burned some cash, with all its problem, the cash burn wasn't that significant thanks to the nickel price.
The financial situation isn't alarming to date, but problems are unresolved. Their conciliation is over and can be resumed at any point in time if SLN deems that necessary. Part of the answer that Thomas will give you is that in our CapEx, we had ramp up CapEx for the six million tons at SLN, and these CapEx, we didn't get the green light. We're not doing those. Some of the SLN CapEx that haven't happened are a part of the reduced CapEx guidance. Thomas, I'll let you answer the other points. Maybe there's a slide. We can return to the CapEx slide. Yeah. On CapEx, we've said that we would do EUR 500 million in 2021, and we're going to deliver between EUR 400 million and EUR 450 million. As I indicated last year, we gave stricter cash rules.
That kicked in cash containment rules that impacted fully the first. Then the six million tons of SLN, we don't commit CapEx if we don't have the go ahead. That's fairly logical. Traditionally, H2 is far stronger than the first in terms of CapEx commitment. That's fairly straightforward. The bulk of our growth CapEx, as I said, are on Comilog to support its ramp up and to support the rail transport that goes with that. In terms of split, it's a marked majority of maintenance CapEx. That's what I can say. On this slide, or shown on screen, you have the slip because what we have the recurring CapEx, well, they are essentially the maintenance CapEx, et cetera. Not necessarily these CapEx that have the most delay.
On Aubert & Duval, I'm not going to give you all the details as you'll, of course, well appreciate. We rejected the offer because it really wasn't satisfactory and not just for questions of valuation. There were a great many terms and conditions in the offer that were unacceptable to us, quite simply. Simply put, it doesn't, of course, prevent us from working to resolve a number of the outstanding issues reflected in the offer. That's why I'm saying we're working on the conditions of the disposal of the divestment so that we can have a divestment in better conditions over the coming months. There are no further questions from the room. We can take questions on the web. Yes. First question on manganese alloys, what are the sectors that are driving demand and prices? Answer.
On manganese alloys, while manganese alloys are sold in steel, carbon steel, so that's in steel, notably, you saw the very sharp production increases of steel output in the U.S. and Europe, and it's the areas where we sell 80% of our manganese alloys, so we're very geared. Our plants are in Norway, in France, in the U.S., and a small plant in Gabon. We're essentially selling in European and North American markets. These are markets that have boomed with local players who had more difficulty than we did to get started again. Eramet is truly seen as a very reliable supplier for manganese alloys and really the benchmark supplier of steel producers of Europe and U.S. We benefited strongly from the strong recovery of the restocking of Europe and U.S. on manganese ore.
Given the health situation in South Africa, are we going to face supply issues, problems from South Africa? Answer, on the face of it, no, but Kleber, maybe you can speak to supply. Well, we're very vigilant on port loading, but you've seen that South Africa has significantly upped its output. We're not in a situation that we are at during COVID, where ports were totally shut down. This year there were a few days of disruption of loadings, but to my knowledge, we haven't had problems of supply of our Norwegian plants from South Africa the past few months. I'm confirmed that that is indeed the case. Do you have any more specific guidance on the cash cost of SLN and on the contribution of Weda Bay to Q2 as compared to Q1? Thomas, would you like to answer those questions?
Answer, on the cash, we weren't any more specific than that. You must realize it. We've carefully calibrated our guidance and will add the cash cost for SLN is obviously very linked to the output volume and should improve significantly in H2. That's our target because output volumes are also set to rise in H2 and on Weda Bay. We also expect an increase. Your debt began to decline in H1. What's your debt, your de-leveraging target at the end of the year? Have you begun to consider your dividend policy? On net debt, well, we haven't given any guidance on that. We said that we would do in excess of EUR 800 million EBITDA and between EUR 400 million and EUR 450 million in CapEx. That gives you a first idea about cash generation.
EBITDA. You add the free cash flow generation from Weda Bay. I'll let you estimate that on the basis of everything I just said. In light of all that, one can expect a reduction in net debt. That's fairly logical. We won't give any more guidance than that. The question of the dividend, well, we'll start off by looking at the results at the end of the year and the balance sheet situation is really to consolidate it before asking that sort of question, I believe. Concerning your Sandouville business, can we hope for a disposal between now and the end of the year? Possibly we could indeed is the answer to that. Thank you. No further questions on the web. Very good.
If there are no further questions in the room, I'd like to thank you all very much for your attention and for those of you who are going on holiday, have a great summer. Thank you.