Peter Cunningham
Analyst · Kaan Peker from RBC
Thanks, Jakob. Good morning, good evening, everyone. I'm really pleased to have the opportunity to present this set of results because we've had good operational momentum with a steady improvement in our performance in the Pilbara, where we delivered iron ore shipments at the upper end of our guidance. We also had a strong start to underground operations at Oyu Tolgoi and Kitimat has returned to full production.
But we do still have a lot of work ahead of us. Firstly, we have some assets where we need to stabilize production. In 2023, IOC and Kennecott in particular, face some challenges. And secondly, we need to push on with the implementation of the safe production system to deliver continuous productivity improvement in our operations. In summary, there is significant value remaining to be unlocked from our existing assets.
On a net-net basis, our underlying EBITDA declined 9% to $23.9 billion. Cash flow from operations remained strong at $15.2 billion, but we do need to bring down inventory. Free cash flow was $7.7 billion after capital expenditure of $7.1 billion. Following dividends paid and funding of the Matalco transaction for just over $700 million, we ended the year with net debt of $4.2 billion, virtually unchanged from 2022.
Overall, we delivered a healthy return on capital employed of 20% on underlying earnings of $11.8 billion. This underpinned our decision to continue our 8-year record of declaring a 60% payout on the ordinary dividend, equating to $7.1 billion.
We did have some one-off items. As I presented at the half year, we made an adjustment to the carrying value of our Gladstone refineries. In the second half, we increased the closure estimates for a number of closed assets, in particular, ERA. As ever, markets are the biggest determinant of annual volatility in our financials and 2023 was no different.
Overall, the price impact was negative. Although it is important to call out the stability of iron ore markets during the period. Despite the Platts Index being broadly flat, our realized iron ore price was actually 2% higher due to higher relativity of lower-grade products. The copper market was largely stable year-on-year with prices declining 3%.
We've recently seen some disruptions in mine supply, about 1 million tonnes, resulting in much stronger concentrate markets. We're also seeing the effects of the energy transition on demand coming through, particularly from the EV market. Aluminum demand continues to increase, although at a lower rate. We saw our realized price come down by 18%, with lower LME price as well as market and product premia.
The behavior of the aluminum price reflects its increased exposure to consumer markets. Let me now provide some context to the iron ore price stability. Critically, 2023 was the fourth year with Chinese steel production above 1 billion tonnes. The big driver was a significant increase in net steel exports to 84 million tonnes, mainly to Southeast Asia.
China is also experiencing a fundamental change in demand. As shown by the chart on the left, since 2019, we've seen a steady rise in its share of finished steel demand going into infrastructure, the energy sector and manufacturing with properties share declining.
Turning now to the EBITDA movement. In aggregate, commodity prices lowered EBITDA by $1.5 billion, primarily driven by aluminum. Weaker currencies in Australia and Canada offset this by about $600 million. The real positive in the period, though was the 3% rise in copper equivalent production.
The increase in Pilbara output was a big factor behind this growth and added $600 million. In copper, we benefited from the Oyu Tolgoi underground ramp-up, but there was some offset at Kennecott due to a conveyor failure in the first half and the planned rebuild of the smelter in the second and third quarters.
aluminum production was 9% higher as Kitimat returned to full production. However, we're not yet seeing the extra metal volume flow into higher earnings due to the additional costs of the ramp-up. Reducing these is going to be a key focus area for 2024. Somewhat counterintuitively, we're showing a negative volume variance for aluminum, which reflects lower value-added product sales of around $100 million.
Our ongoing exploration and evaluation expenditure in 2023 was $900 million, which compares with guidance of around $1 billion. We saw a significant step-up in activities at Simandou, which we continue to expense until the end of the third quarter.
Net-net, E&E was around $300 million higher than last year. More broadly, however, other options are progressing. And at the front-end of the pipeline, we now have the best exploration portfolio we've had for some time, having consistently invested in this area over the years.
The cost picture is covered by several bars in this chart, but let me try and summarize what we're seeing in broad terms. Firstly, as foreseen at the half year, we did see the reversal of some market-based costs, particularly aluminum raw materials. You can see this in the first section of the chart.
Secondly, many of our costs are under contracts, which renew periodically. As a consequence, the spike in inflation was only reflected in 2023 on renewal of these contracts. This process now looks to be largely complete. Thirdly, we continue to see some cost pressures from tight labor markets, particularly the Pilbara, Quebec and Utah. Again, these are in the unit cost variance.
We have separated out the effects of the operational disruptions at Kennecott and IFC. You can see on the chart that they drove up our unit costs to the tune of $600 million. Overall, we do believe a lot of the forces driving up costs are now starting to moderate, and we expect to see more stability in the cost base going forward.
Our business continues to be highly cash generative. This chart reconciles EBITDA and cash flow. Our cash conversion ratio was 63% compared to 61% in 2022 when tax payments were substantially higher. At the half year, I did say I expected working capital to reduce in the second half, but instead, it stayed roughly flat. We saw reductions in some areas such as raw materials, but the extended Kennecott smelter shut and softness in the TiO2 market meant that the aggregate balance of inventory did not come down as expected.
This was compounded by the rise in the iron ore price late in the year, increasing balance sheet receivables. These were turned into cash in early 2024. We also had lower dividends from equity accounted units, mostly related to Escondida. Finally, the major driver of provisions is closure. We have a number of active projects underway with just under $800 million spent in 2023.
Looking forward, we expect to spend around $1 million per year as we advance activities at the various sites. Spend will vary year-to-year as we execute individual programs of work and we continue to look at structural opportunities to reduce our closure exposure.
Onto product group performance. Iron ore had a strong year, its second highest on record for shipments. Gudai-Darri is at nameplate capacity, and we're extracting more volumes from the Safe Production System with a 5 million tonne uplift in 2023. We're targeting another 5 million tonnes this year with the combined 10 million tonne benefit delivering significant incremental value to the business.
We expect a small increase in unit costs in 2024, reflecting ongoing tight labor markets in Western Australia and costs associated with material movement and maintenance in our system. We're building a much stronger aluminum business. It was a tough year as the price dropped and margins compressed. However, as I said, Kitimat is now back to full capacity and we're making investments in North America that really strengthen this business for the future. These include investing in the AP60 technology and in Matalco, with the latter giving us exposure to recycled products.
As Jakob mentioned, it is really positive to see the Oyu Tolgoi mine investment starting to pay off with the ramp-up of production from the underground. And at Kennecott, our focus is to stabilize the operation following the completion of the smelter rebuild.
Lastly, it was a challenging year for minerals from both an operational and market perspective. IOC lost 1 month of production in June due to wildfires and we had some operational impacts in the third quarter. Whilst at our iron and titanium Quebec operations, 3 furnaces remain off-line in response to weak market conditions.
Moving on to capital allocation. You've seen this slide showing our approach many times. My key message today is that nothing has changed. Sustaining capital, higher returning replacement projects and decarbonization remain our first priority where we're forecasting around $7 billion of spend per year, unchanged from previous guidance.
That is followed by the ordinary dividend and then compelling growth. We believe that $3 billion remains the right level for us to invest in growth, and our largest project is expected to be our equity share of Simandou. While CapEx at Oyu Tolgoi underground will wind down as we complete key infrastructure investments.
We expect the remainder to be mainly invested in copper and lithium projects, some of which are yet to be sanctioned. But as I've said many times before, we will remain very disciplined. Our investments in growth are highly dependent on the timing of commitments, but most importantly, by our ability to generate value.
Just turning now to the key financials for Simandou. As previously guided, we saw $900 million of spend incurred on the project in 2023, $500 million of which is our share and $400 million will be refunded by our Simfer JV partner, Chalco Iron Ore holdings. This includes $300 million of qualifying costs, which we started to capitalize from 1 October.
In 2024, we expect our share of spend to be around $2 billion. Now I was very pleased actually to have the opportunity to visit the project last month. And I must say, I was pretty impressed to see the progress being made on the ground.
Finally, the dividend. We have declared a 60% payout for the full year, which equates to $7.1 billion and attractive dividend yield of more than 6%. We remain very consistent with our shareholder returns policy with a 60% payout on ordinary dividends and 71% total payout across the last 8 years. This highlights our continued discipline. Our net debt is unchanged year-on-year, and this financial strength, means we can accelerate our decarbonization investment, reinvest for growth and continue to pay attractive dividends through the cycle. And with that, let me hand back to Jakob.