[home] [Personal Program] [Help]

---

Session: Poster session & Sponsor Exhibition
Session starts: Monday 07 October, 14:00

---

Tiao Yuan Wu (China Steel Corporation)
Pai Hsiang Wang (China Steel Corporation)
Chun Da Chen (China Steel Corporation)

Abstract:
Organic Rankine Cycle (ORC) generation system can convert heat to electricity and has been widely applied to in various heat sources with different temperature ranges. This technology helps to save energy consumption, reduce cost, and make more benefit. Especially, it is one of the few technologies that could efficiently recover low-temperature waste heat. In China Steel Corporation (CSC), some technologies such as co-generation have been employed to recover about 40% of waste heat, but no suitable systems could be used in the low-temperature region. Therefore, CSC developed a 10 kW ORC pilot plant and investigated the optimal operation to get higher efficiencies of power generation and net output power. The operation of ORC system involves some controllable and un-controllable parameters, and there are some constraints in the operation, for instance the maximum power generation capacity, and the maximum temperature of working fluid. Besides, some power consumptions from working fluid pump, hot water pump, cooling pump, and cooling tower fan should be considered to get the maximum efficiency of net output power; these will make the problems much complex. To get the maximum efficiencies of power generation and net output power, CSC employed a set of optimization methodology, including Design of Experiments (DOE), Response Surface Methodology (RSM) and Sequential Quadratic Programming (SQP), in a 10 kW ORC pilot plant. The DOE can help to plan the experimental parameters which will be used to effectively build the response surface. The response surface is a model that can predict the performance accurately and is usually used to parametric analysis or optimization analysis. The SQP is an optimization programming and can be employed to search global optimal values with some specified constraints in the response surface model. In the present study, the controllable parameters and the parametric boundaries are working fluid flow rate (20~40 kg/min), hot water temperature (100~120oC), hot water flow rate (70~150 LPM), and cooling water flow rate (130~310 LPM), respectively. By using this set of optimization methodology under the constraints of parametric boundaries and power generation < 10700 W (maximum power generation capacity), the maximum efficiency of power generation increases from 7.48% to 8.03%, about 0.5% better; the maximum efficiency of net output power increases from 3.88% to 4.48%, about 0.6% better. These improvement can be achieved easily only with changing the operating parameters and without any more cost.