Gacko II, Bosnia and Herzegovina

In December 2017, a Memorandum of Understanding was signed between state-owned utility Elektroprivreda Republike Srpske, the China Machinery and Engineering Corporation (CMEC) and Emerging Markets Power Fund, to build a new 350 MW lignite power plant in Gacko, near the town’s existing plant. After years of stagnation, in August 2022 it was reported that the Czech company Witkowitz was considering investing in the project. However, it is still unclear where financing would come from.

As is too often the case in southeast Europe, there has been no convincing analysis proving that this plant is needed or that it would be the best way to provide Bosnia-Herzegovina’s energy supply in the coming years.  

In April 2018 an analysis by economist Damir Miljević showed that fatal flaws in the input data make it highly likely the plant will generate losses. Three out of the main data inputs for the official feasibility study – the price of coal, electricity sales price, and the price of CO2 – are unrealistic: 

• A realistic price of coal is mentioned in the study as just over EUR 18 per tonne – yet the amount used in the calculation is much lower, around EUR 13.3 per tonne.
• The Study foresees export of all the electricity generated, at a price of EUR 50 per MWh, except in exceptional cases when 30% would be sold on the domestic market at EUR 19.90 per MWh. There no evidence that the electricity would find a market in the long term and that it could be sold at this price. Moreover, the scenario including 30% of electricity being sold domestically is not even examined in the calculation – if it was it would show that the plant is unprofitable.
• A CO2 price of EUR 5 per tonne is mentioned in the text, but not included in the feasibility calculation. Including even this very low CO2 price in the calculation would take the plant into the realm of unprofitability. 

The analysis concludes that although the official feasibility study for Gacko II claims it would generate profit of around EUR 23 million per year, with more realistic input data, a loss of minimum EUR 1.15 million per year looks more likely.