mwambachilufya/Zambia_Copper_Production_Project

Data Science project that investigates the factors that drive Copper production at FQM Zambia and whether the country can achieve the 3m tonne target set for 2031 # FQM Zambia — Copper Production Forecast 2025–2031 ### Sentinel FQM & Kansanshi Mine | Scenario Modelling, Tableau Visualisation & Machine Learning --- ## Dashboard. The dashboard shows the FQM copper prodution predictions starting from the actual at 401 721 in 2024 all the way to 2031. The bar chart shows the three 2031 scenario totals for both mines combined. There is the Best Case at 485,416 tonnes, Optimistic at 624,315 tonnes, and Pessimistic at 350,347 tonnes. The pie chart on the right shows how Kansanshi's Optimistic 2031 output is divided based on the type of ore of which mixed ore contributes more at 184,275 tonnes. The line chart at the bottom shows how each scenario trends from the 2024 actual value of 401,721 tonnes all the way up to 2031. The dip on the Best Case line at 2028 is a data entry error in the original excel file which was corrected using three dfferent machine learning algorithms for validation purposes. --- ## 2031 Scenario Comparison Three scenarios were modelled for the combined FQM Zambia operations across Sentinel and Kansanshi. The difference between optimistic and pessimistic scenario is 274,000 tonnes. The reason for this is because of the mixed ore at Kansanshi. Under the pessimistic case, mixed ore contributes 41,652 tonnes. Under the optimistic case it contributes 184,275 tonnes. That is a difference of 142,623 tonnes. Therefore mixed ore is the most important contributing factor in this prediction thus making it a key performance indicator. --- ## Production Trajectory 2024–2031 All three scenarios start from the same 2024 actual value of 401,721 tonnes and trend over seven years. The Optimistic case assumes the mine achieves recovery rates, ore grades, and throughput volumes that are very ideal which is why the machine learning models later predict more cautiously on this scenario than the formula-based Excel forecast does because in the real word things are far from ideal. The Best Case line dips sharply at 2 …