Experimental characterization of scroll expander for small-scale power generation in an Organic Rankine Cycle

Felipe Airoldi Accorsi, Purdue University

Abstract

Organic Rankine Cycles (ORC) are thermodynamic power cycles designed to generate work from low temperature sources. The low temperature heat input, typically between 80 °C to 270 °C, allows them to recover energy from industrial waste heat, exhaust gas from diesel, engines, solar systems, geothermal systems, and others. This technology already is commercialized in large-to-medium-scale power plants; however, it is still in development for small-to-micro scale. The operating principle of ORC comprises four major steps. First, the working fluid evaporates at high pressure using the heat transfer from the low temperature heat source. Then, the working fluid reduces its enthalpy in an expander producing mechanical work, which is turned into electricity by a generator. The low-pressure working fluid leaving the expander outlet is liquefied in the condenser. The working fluid is then pumped, pressurizing the liquid to the high pressure and restarting the cycle. The efficiency of an ORC is highly dependent on the expander characteristics. This research experimentally evaluates the performance of a scroll expander with a nominal capacity of 5 kW and a built-in volume ratio of 3.5. Tests were conducted in an ORC test-rig using R245fa as the working fluid. Two temperature sources of 85°C (185°F) and 110°C (230°F) and five expander speeds, from 800 RPM to 3000 RPM were tested. The scroll expander achieved a maximum isentropic efficiency of 0.58 at a volume ratio of 6.12, expander speed of 1600 RPM, and a temperature source of 110°C (230°F). The same temperature source also registered the maximum expander power output of 3.75 kW at a volume ratio of 6.55 and an expander speed of 2500 RPM. An empirical and a semi-empirical model were compared with the data to assess their competence in predicting the ORC outlet power and isentropic efficiency. Both models matched experimental data accurately. The empirical model has average deviation errors of 3.22% and 3.20% for the expander isentropic efficiency and power output, respectively. While a mean deviation in the order of 4% was registered using the semi-empirical model. The semi-empirical model suggests friction and leakage attenuate efficiency more than any other source.

Degree

M.S.M.E.

Advisors

Groll, Purdue University.

Subject Area

Mechanical engineering

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server
.

Share

COinS