Enhanced Geothermal Systems (EGS) technology was successfully demonstrated in the Australian context with the operation of the Habanero 1MW pilot plant. This project aims to determine the optimum power plant design for the geothermal parameters found at the Habanero pilot plant. In order to achieve this, a techno-economic optimisation of an Organic Rankine Cycle (ORC) was undertaken.
The EGS conditions used in this work are a brine production well head temperature of 220 oC, and minimum brine temperature of 80oC in order to limit scaling formation in the brine heat exchanger(s). The production well head pressure is 35 MPa and the required reinjection pressure is 45 MPa in order to maintain the desired mass flow rate of 35 kg/s through the EGS resource.
A significant source of parasitic power consumption in ORC systems occurs in the condensing system. In order to avert this parasitic power consumption Natural Draft Dry Cooling Towers (NDDCTs) were investigated as the condenser for the ORC. A one dimensional NDDCT model was developed and integrated into the cycle design process to analyse and design for the coupled nature of NDDCT performance with the power cycle. As a base for comparison a one dimensional Mechanical Draft Air Cooled Tower (MDACT) model was developed and each cycle was also analysed with MDACT as the condenser.
A wide range of organic working fluids and several cycle configurations were evaluated in the preliminary analysis using a simplified NDDCT model. The cycles were optimised for maximum net power generation and the highest performing cycle configurations were progressed to the techno-economic design point optimisation stage. The cost of each of the major equipment items in the plant was estimated using cost correlations based on historical equipment cost data. The condensing system geometry for both NDDCT and MDACT, heat exchanger geometry and cycle parameters were optimised to find the lowest Specific Investment Cost (SIC) in AUD/kWe for each candidate cycle. The cycle configurations with the lowest SIC from the design point analysis were evaluated across the range of ambient temperatures expected at the site. The mean annual net power generation for each cycle was calculated based on site temperature data and this was used in determining the annualised SIC values, the measure by which the optimum plant configuration was selected.
The recuperated, regenerative and basic ORCs were found to be the cycles that obtained the highest net power generation in the preliminary analysis with butane, butene, isobutene, R152a, isobutane, R123 and isopentane the highest performing fluids. The highest net power generation found in the preliminary analysis was 2.688 MWe.
The NDDCT model developed in IPSEpro was investigated in isolation to find the optimum design configuration which gives the lowest SICcd, in AUD/kWth of heat rejected. The tower geometry ratios selected were: aspect ratio (tower height / base diameter) of 1.4, diameter ratio (outlet diameter / base diameter) of 0.7, and 𝐴𝑓𝑟/𝐴3 (the proportion of heat exchanger coverage of the base of the tower) of 0.65. With these geometric ratios fixed, the effect of tower size on cycle performance was investigated in a basic cycle model, by varying the number of heat exchanger bundles, and it was found that an NDDCT of 52.5 m in height and 37.5 m in base diameter gave the lowest SIC for the cycle.
The detailed cycle design stage optimised the 15 cycle configurations selected from the preliminary analysis with both indirect NDDCT and direct MDACT condensers. The cycles were optimised for SIC and it was found in all cases that, despite their higher TCI, the NDDCT condensed cycles produced lower SIC values, due to the higher 𝑊̇𝑛𝑒𝑡. The highest performing cycles in ascending order of SIC were the recuperated cycles with isobutene, butene and butane, basic butene, recuperated R152a and then the regenerative butene and regenerative butane cycles. These cycles were selected to progress to the annual performance analysis along with one of each cycle type with an MDACT condenser, in order to allow comparison of NDDCT and MDACT performance variation versus ambient temperature.
The selected cycles were first analysed across the range of ambient temperatures expected at the site, based on temperature data from the Australian Bureau of Meteorology. Next they were subject to a diurnal performance variation analysis for four sample cases for each of the seasons; significant variation of net power generation was found with variation of up to ∓ 20% from the mean on a daily basis and 25 to 35% change in the mean net power generation from summer to winter, depending on the cycle. Finally, the annual performance analysis used daily temperature data for 2012 to calculate the mean daily net power generation for each of the finalist cycles and this was used to find a mean annual net power generation. The NDDCT cycles were found to achieve 3% to 5% lower SIC than their respective MDACT condensed cycles. The optimum cycle according to the annualised SIC was found to be the recuperated supercritical butene ORC with an NDDCT.