The development of mine IoT requires an advanced computing system managing diverse sensing devices in mine scenario. Consequently, the heterogeneous computing system(HCS) which has been widely used to perform complex and massive computation, was introduced to mine scenario. For multiple workflows processing in HCS, how to make a reasonable tradeoff between schedule length(SL) and energy consumption is a key issue. Toward this direction, the dynamic voltage and frequency scaling(DFVS) technique which was proposed to make compromise between SL and energy, is employed by most energy-efficient scheduling algorithms. This paper studys workflows that can be represented by directed acyclic graph(DAG) and addresses the issue of minimizing the total SL with a given energy consumption constraint by depicting two state-of-the-art algorithms. Simulation results show that the two algorithms can reduce energy consumption prominently compared to a widely-used heterogeneous earliest finish time(HEFT) algorithm.