Centre for Energy Technology

1. Centre for Energy Technology Ricoh Clean Energy Scholarship Fibre optic broadband: a pathway for remote geothermal energy to market? Ashok A Kaniyal* Prof. Graham ‘Gus’ J Nathan Prof. Jonathan J Pincus 34th IAEE International ConferenceStockholm, Sweden – 19-23 Jun 2011 School of Mechanical Engineering

2. Geothermal Resource Exploitation in Australia Electricity transmission network • Engineered geothermal systems (EGS): recoverable resource is vast > 1.9 x 1025 J; • Considerable barriers: • Cooper Basin resources are >500 km from electricity grid, • Uncertainty in subterranean dynamics of heat flow, • Above ground plant performance uncertainty; • Not one unit of fully operational electricity generating capacity. Highest temperature geothermal resources near the Cooper Basin

3. Why fibre optic network? • Interconnection to a market for low capital outlay; • Can attract data centres to bring forth a revenue stream for geothermal proprietors where… • Data centre energy demands can be met by geothermal energy of: • Low temperature (~120oC): Direct Use (DU) for absorption refrigeration • High temperature (~240oC): Combined Heat and Power (CHP). Natural gas could supplement electricity demand in both cases. • Potential economic viability of geothermal systems  from receiving retail electricity prices  opportunities for incremental capacity expansion.

4. The Concept • Geothermal data fibre linkto National Broadband Network @ US$60m (DBCDE, 2010). NBN Co Fibre optic links Geothermal-data fibre link Fibre optic node Geothermal-data fibre link~US$60m DBCDE, 2010.

5. Aims • Define the energy consumption profile of a data centre unit. • Develop assessment scenarios based on the most significant uncertainties affecting geothermal energy extraction. • Describe the geothermal system designs that could meet the data centres’ energy demands: • Low temperature direct use (120 deg. C) • High temperature CHP (240 deg. C) • Assess the economic viability of investments in energy and fibre optic common use resources.

6. Cooling load~189 kWr NBN market Data Centre Co. hv Geothermal-data fibre link (1000-1500 km) IT electrical load~252 kWe Data centre energy requirements • Energy consumption characteristics of a modular 350 kWe data centre unit (Barroso, 2007): • Refrigeration load ~ 189 kWr • Electrical load ~ 252 kWe

7. Systems and scenarios • Main uncertainty with respect to flow • Geo-fluid flow rate systematically varied: • Three scenarios examined • Explorers aim to generate higher flow rates than assumed: • Petratherm P/L: 70 kg/s and Geodynamics P/L: 100 kg/s

8. Total IT electrical demand6.65/4.90/3.15 (H/M/L) MWe Geo-fluid flow rate High (H)/Med (M)/Low (L) outcomes Low temperature direct use Data centre serviced 14/19/9 units NBN market hv Geothermal-data fibre link (1500 km) mhw_geo(kg/s) Natural gas (NG) co-genH: 6.10 MWth/6.65 MWe M: 4.90 MWth/5.10 MWe L: 3.05 MWth/3.15 MWe Li-Br absorption/ electric chiller mhw_NG(kg/s) Direct use system • Three flow outcomes enable: • High  70 kg/s  19 data centres serviced per production well • Medium  50 kg/s  14 data centres serviced per production well • Low  30 kg/s  9 data centres serviced per production well

9. Geo-fluid flow rateHigh (H)/Med (M)/Low (L) outcomes Total IT electrical demand6.65/4.90/3.15 (H/M/L) MWe High temperature CHP: 1.50/1.09/0.65 MWe NBN market Data centre serviced 19/14/9 units hv NBN augmentation(~1500 km) mhw_geo(kg/s) Natural gas (NG) cogenH:5.15 MWe / 4.85 MWth M:3.80 MWe / 3.65 MWth L:2.66 MWe / 2.45 MWth mhw_NG(kg/s) Li-Br absorption/ electric chiller Combined heat and power system • Three flow outcomes enable: • High  70 kg/s  19 data centres serviced per production well • Medium  50 kg/s  14 data centres serviced per production well • Low  30 kg/s  9 data centres serviced per production well

10. Additional assumptions • Total of four production well increments examined: • Total of six wells - production and injection. • Data centre units assumed to be attracted to install facilities at a rate of 6 per year • Rate increases by one data centre unit every two years.

11. Energy systems – Total costs

12. Results of assessment of Economic viability

13. Direct use system – IRR (%) Cost of capital

14. CHP system – IRR (%) Cost of capital

15. Fibre optic network augmentation: economic modelling assumptions

16. Fibre optic augmentation: economic viability • IRR of 6% achieved when at least: • Medium flow outcome – 14 data centres serviced per production well • NPV < 0: Low flow outcome < 9 data centres serviced per production well

17. Attracting data centres • Advantages of remote geothermal location for a data centre: • Obvious CO2 emissions mitigation benefit. • Lower cost of energy than a competing urban site. • Cost of land is negligible in comparison to an urban site. • Disadvantages of remote location: • Remoteness!! BUT almost all data centres are operated remotely. • Energy is only one of the factors effecting locational decision making • This research addresses two critical issues associated with attracting modular data centres but... • Investors’ heterogeneous preferences – public subsidy difficult to quantify.

18. Conclusions • Economic viability of servicing a data centre with geothermal energy shown for: • Low temperature direct use system if geo-fluid flow > 30 kg/s • High temperature CHP system only if geo-fluid flow > 50 kg/s • Fibre optic link only viable if > 50 kg/s @ IRR = 6%. • Synergies that enable – high capital productivity, renewable resource to displace a high value product and at low scale. • Significant value in this approach: • Opportunity for incremental investment – extremely important for • resource characterisation and • incremental reductions in drilling expenditure. • Cash flow to enable geothermal industry to pave its own way

19. THANK YOU! For your attention

20. Comparison of capital costs Source: Gawlik and Kutscher - NREL (2000)