The purpose of this paper is to present the novel long-term energy mix optimization model eMix, which takes into account daily requirements of power system operation (primary, secondary and tertiary reserve) and how should they react to increasing RES penetration. Until now there has been no computational effective, long-term optimization model of energy mix with individual representation of power generating assets that allows for RES development incorporation and consider the daily requirement of power system operation. Existing unit commitment models, although comprehensive, are not appropriate for long-term, large-scale problems because of computational limitations. The eMix links long-term planning (investments) with short-term requirements of power system operation reserves (primary, secondary, tertiary) provided by dispatchable units. Additionally, development of the RES introduces stronger requirements to the reserve's level in order to maintain the flexibility of the power system operation. This relation has significant impact on the final energy mix and was modeled in the eMix. The costs are calculated annually for each individual unit (given to the application of Mixed Integer Linear Programming), to reflect the real conditions of energy sector operation, where annual repayment of loans, amortization as well as operation and maintenance costs are applied. eMix is also fully scalable and allows for freely definable interval and horizon of the calculation. To prove the quality of presented model, the authors calibrated it with the data from Polish power system for different scenarios. Results indicate that the consideration of financial costs and required reserves increase the reliability of the optimization outcomes. (C) 2016 Elsevier Ltd. All rights reserved.