skip to main |
skip to sidebar
Concept of geothermal power for Singapore
Geothermal Energy Concept for Singapore
Grahame J.H. Oliver, PhD, BSc
SEAPEX Visiting Senior Fellow, Department of Geography, National University of Singapore.
Introduction
Most Singaporeans know there are hot springs and plenty of granite on their island. Sixty million people living on plate boundaries around the world already obtain their electricity from hydrothermal sources in young magmatic rocks. Many commentators see the return of the USD100-plus barrel of crude oil as inevitable with future gas prices tracking that rise. Eighty percent of Singapore’s electricity is generated from imported natural gas. Geothermal exploration of buried hot granite terrains in continental interiors is now attractive: in Australia and Alaska, exploration has now passed into the appraisal phase. Australian State Governments have committed USD90 million for research and demonstration and another USD750 million has been allocated to works programs for the period 2002 and 2013 (1). In May 2009, President Obama announced a USD 350 million stimulus boost for US geothermal energy(2).
The main heat releasing isotopes in rocks are uranium-235 , uranium-238; thorium-232 and potassium-40. Granites usually have more of these elements than most other rocks. Granites with more than 10 ppm U can be classified as “hot” and provided the granite has been allowed to heat itself up under a thermal blanket of overlying rocks, significantly high temperatures can build up over millions of years. For example, one cubic km of buried granite at 250 o C has the stored energy equivalent of 40 million barrels of oil. New technology means that boiling geothermal water or steam is not required. The commercial binary cycle Chena Power Station in Alaska boils R-134a refrigerant with 74oC geothermal water extracted from Mesozoic hot granite, and produces 2-300 kWh at US 5 cents/kWh(3). This is in contrast to the US 30 cents/kWh cost of diesel generators previously used at Chena(3). The Geysers region in California generates electricity at 5 cents/kWh(4). US 5 cents/kWh is competitive against all forms of power generation except coal.
Geology and hot springs of Singapore
Singapore lies inside the stable Asian continental plate called Sundaland. The island is composed mainly of Middle Triassic I-Type granite and minor gabbro, intruded into a km thick blanket of contemporary acid volcanics and partly covered by Upper Triassic and Lower Jurassic sediments (Fig. 1). There are three confirmed hot springs situated at or near the coasts (although the precise locality of the one on the SW side of Pulau Tekong is uncertain because of land reclamation). Seeping “steam” (sic) has been reported to me but not confirmed from another location on Sembawang Singapore Air Force base (Fig. 1). The best known hot spring at Sembawang has been drilled down to 100m into a 50 m wide fault zone in granite: temperatures of 70.2 o C were measured(5). Chemical analyses classify it as a potable neutral chloride spring with total dissolved solids (TDS) measured at 914 mg/l and a Cl content of 431 mg/l(5). I have applied various geochemical thermometers(6) using Si/Na/K/Ca concentrations which indicate underground reservoir temperatures between 122 and 209 o C, i.e. hot enough to produce steam at the surface. Cold surface ground water probably mixes with the geothermal water, so the calculated reservoir temperatures are minimums.
Granite bedrock is not normally considered to be permeable unless it is well fractured, jointed and faulted. I have investigated the granite quarries around Bukit Temah and Pulau Ubin and close spaced jointing is common (see Fig. 2). The average maximum horizontal stress (s1) in ~100 m bore-holes in central Singapore has a 13oE orientation(7). My analysis of the stress vectors in the Sumatra section of the Australian Plate collision in the Sumatra-Java trench and subduction zone to the SW of Singapore, suggests that the maximum horizontal stress in the hangingwall of the subduction zone is oriented 30oE. The present day stress map of SE Asia(8), based on earthquake fault solutions, shows that the maximum horizontal stress is orientated NE/SW in neighbouring Sumatra. Furthermore, my lineament map of Singapore, based on topography, geological map and satellite data shows a very strong NE/SW trend (35oE, see Fig. 1).The important implication is that any joints (of whatever age) that are orientated NE/SW will be open in this stress regime.
I have measured joints and fault orientations in various Singaporian granite and gabbro quarries and there is a NE/SW correlation with the lineament map (see Fig. 2). The prediction is that granite with a strong NE/SW orientation of open joint sets will preferentially channel ground water in a NE/SW direction away from the watersheds. Figure 1. shows that the confirmed hot springs in Singapore are indeed NW of their associated watershed maxima.
Ground water model for Singapore
The USGS groundwater model for islands with an unconfined aquifer surrounded by seawater can be applied to Singapore(9). Assuming seawater has a density (Ps) of 1.025 g/cc compared with fresh water (Pf) of 1.0 g/cc, then according to the Gyben-Herzberg relation,
z = Pf . h
Ps - Pf
a head (h) of 100m above sea level in the centre of the island will drive cold fresh water down 4 km below sea level (z). I have confirmed that there is a permanent (cold) spring at an altitude of 120 m on Bukit Temah, the highest hill in Singapore (164m), indicating that such a high water table is probable. Assuming that the average geothermal gradient for the Singapore region is 35 o C/km(10), ground water at 4 km depth will reach 140 o C. Because of the high rainfall (2.4 m/year) and the +100m head, the hot ground water will be driven along the fresh water/seawater transition and up to the surface at the coast (Fig. 3). As described above, this hot water is likely to be preferentially channeled along NE/SW orientated joints and fractures. The Sembawang hot spring is 3.7 m above sea level which was at sea level during the warm interglacial periods at ~80 ka and ~125 ka(11). The Pulau Tekong hot spring is at 0.5 m above sea level, and from photos, it looks to be in the mangrove transition.
Exploration and Appraisal
Pump testing of 100 m deep wells at the Sembawang hot spring produced up to 400 l/min at a constant 70 o C for many days(5). However, this rate is not high enough to support a commercial power plant. Deeper production wells are required to intercept hotter water and these need to be coupled with injection wells to supplement the artesian flow.
Geophysical surveys: e.g. gravity, magnetics, 3D seismics, electrical resistivity, and heat flow surveys (i.e. infrared radar and 0.5 to 100 m bore holes) are required to locate a 2 km deep exploration well that will hopefully intercept significantly hot artesian water. Ground water age dating (3H/3He) plus SF6 and CFC analysis would be useful to model artesian flow rates.
Proof of Concept
Proof of concept for a 50 MW power station, providing power for 50,000 homes by tapping 150oC water at 2 km depth might cost USD 19 million: i.e. two 3 km long L-shaped wells at USD 3 million/km, plus USD 1 million for a 1 km hydro-fracture job. The horizontal part of the wells should be 1 km long and orientated NW/SE so as to maximize intersections of NE/SW trending joint and fracture sets (Fig.4). If this was proved as successful, another pair of L-shaped wells could be drilled from the same platform, at 180o to the first pair.
Development Costs
Development costs for the 200 kW binary generation system based on the commercial Alaskan shallow (200 m) hot wet granite binary system using massed produced refrigeration components were US$1300/kW(3), i.e. US$ 1.3 million/MW. Production costs at 5 US cents/kWh and selling at a domestic rate of 12 US cents/kWh generates a profit of 7 cents/kWh, i.e. US$70/MWh. Based on these figures, in one year a proof of concept 1 MW power station (for 1000 homes) might make 70 x 1 x 365 x 24 = US$ 0.613 million “profit” before tax. The development costs could be written off in 1.3 / 0.613 = 2.12 years and save the equivalent of 10,000 barrels of oil/year (with crude oil at US$100 oil/barrel this is a saving of US$ 1 million/year)
Development costs in California Geysers district, using conventional steam generation from 2 km depths are US$2.9 million/MW(4). A 50 MW power station (providing for 50,000 homes or a good proportion of the power used on Singapore Mass Rapid Transportation railway system would therefore cost 50 x 2.9 = US$ 149 million. Production costs at 5 cents/kWh and selling at a domestic rate of 12 cents/kWh generates a profit of 7 cents/kWh, i.e. US$70/MWh. In one year a 50 MW power station might make 70 x 50 x 365 x 24 = US$30.6m “profit”. Write off would take 149/30.6 = 4.9 years and 50 MW saves the equivalent of 0.5 million barrels of oil/year which at US$100/ bbl = US$50 million/year.
Environmental Impact
An environmental impact assessment would be a high priority. Geothermal energy is viewed as renewable, clean and green with a small carbon footprint during the construction phase. No doubt, a rig capable of drilling 3 km long directional wells would be disruptive in Singapore’s mainly urban environment. All the drilling for a power station could be conducted from one location. The hydraulic fracture job might create micro-seismicity, but Singapore already experiences micro-seismicity from the Java/Sumatra subduction zone. The infra structure for a 1 MW ‘proof of concept’ power station would fit into 2 or 3 tennis courts. A 50 MW power station could perhaps be located on an area the size of three or four football fields. Pipe work and high tension transmission lines would be placed underground. Water supply in Singapore is an issue and there would be an initial requirement to augment the working hydrothermal fracture system with fresh or storm water, but not sea water which could cause scaling. Once the system was pressured up and if the injector/production well connections were efficient, the requirement for augmented water would drop. Air rather than water cooling might be installed to condense turbine vapour for recycling. Sembawang hot spring has virtually no smell but Pulau Tekong hot spring is reported to be H2S-rich.
Summary
The Singaporean method of geothermal power generation would involve the drilling of 3 km deep directional wells into hot, wet, fractured granite under a thermal blanket and the generation of electricity either from70 o C hot water through binary cycle turbines or from ~150 o C conventional steam turbines, with the cooled water being recycled down injection wells. Proof of concept for a 50 MW power station might cost USD 19 million. Development costs could be written off in 5 years. There is a “local” market of 4.8 million people and the generating costs would remain static whilst the cost of natural gas varies. This is ‘renewable’, clean, green power generation and of strategic importance for a country that is viewed conventionally as having no natural resources.
References: web sources were accessed from February to August, 2009.
(1) http://www.searchanddiscovery.com/documents/2008/08181goldstein/index.html.
(2) http://useconomy.about.com/od/candidatesandtheeconomy/a/Obama_Stimulus.htm.
(3) http://www.chenahotsprings.com/geothermal-power/.
(4) www.geyser.com.
(5) Zhao, Chen & Cai (2002) Bull. Eng. Geol. Envir., 61, 59-71.
(6) Bowen (1986) Groundwater Technology and Engineering. Springer.
(7) Zhow (2001) Engineering Geology and Rock Mass Properties of the
Bukit Timah Granite. Report, Singapore Defence Science & Technology
Agency.
(8) http://www-wsm.physik.unikarlsruhe.de/pub/stress_data_frame.html
(9) http://pubs.usgs.gov/circ/2003/circ1262/#heading156057192.
(10) Mazlan et al. (1999) In : PETRONAS. The Petroleum Geology and
Resources Malaysia.
(11) Lambeck et al. (2002) Links between climate change and sea levels for the
past million years. Nature, 114, 199-206.
Grahame J.H. Oliver, PhD, BSc
SEAPEX Visiting Senior Fellow, Department of Geography, National University of Singapore.
Introduction
Most Singaporeans know there are hot springs and plenty of granite on their island. Sixty million people living on plate boundaries around the world already obtain their electricity from hydrothermal sources in young magmatic rocks. Many commentators see the return of the USD100-plus barrel of crude oil as inevitable with future gas prices tracking that rise. Eighty percent of Singapore’s electricity is generated from imported natural gas. Geothermal exploration of buried hot granite terrains in continental interiors is now attractive: in Australia and Alaska, exploration has now passed into the appraisal phase. Australian State Governments have committed USD90 million for research and demonstration and another USD750 million has been allocated to works programs for the period 2002 and 2013 (1). In May 2009, President Obama announced a USD 350 million stimulus boost for US geothermal energy(2).
The main heat releasing isotopes in rocks are uranium-235 , uranium-238; thorium-232 and potassium-40. Granites usually have more of these elements than most other rocks. Granites with more than 10 ppm U can be classified as “hot” and provided the granite has been allowed to heat itself up under a thermal blanket of overlying rocks, significantly high temperatures can build up over millions of years. For example, one cubic km of buried granite at 250 o C has the stored energy equivalent of 40 million barrels of oil. New technology means that boiling geothermal water or steam is not required. The commercial binary cycle Chena Power Station in Alaska boils R-134a refrigerant with 74oC geothermal water extracted from Mesozoic hot granite, and produces 2-300 kWh at US 5 cents/kWh(3). This is in contrast to the US 30 cents/kWh cost of diesel generators previously used at Chena(3). The Geysers region in California generates electricity at 5 cents/kWh(4). US 5 cents/kWh is competitive against all forms of power generation except coal.
Geology and hot springs of Singapore
Singapore lies inside the stable Asian continental plate called Sundaland. The island is composed mainly of Middle Triassic I-Type granite and minor gabbro, intruded into a km thick blanket of contemporary acid volcanics and partly covered by Upper Triassic and Lower Jurassic sediments (Fig. 1). There are three confirmed hot springs situated at or near the coasts (although the precise locality of the one on the SW side of Pulau Tekong is uncertain because of land reclamation). Seeping “steam” (sic) has been reported to me but not confirmed from another location on Sembawang Singapore Air Force base (Fig. 1). The best known hot spring at Sembawang has been drilled down to 100m into a 50 m wide fault zone in granite: temperatures of 70.2 o C were measured(5). Chemical analyses classify it as a potable neutral chloride spring with total dissolved solids (TDS) measured at 914 mg/l and a Cl content of 431 mg/l(5). I have applied various geochemical thermometers(6) using Si/Na/K/Ca concentrations which indicate underground reservoir temperatures between 122 and 209 o C, i.e. hot enough to produce steam at the surface. Cold surface ground water probably mixes with the geothermal water, so the calculated reservoir temperatures are minimums.
Granite bedrock is not normally considered to be permeable unless it is well fractured, jointed and faulted. I have investigated the granite quarries around Bukit Temah and Pulau Ubin and close spaced jointing is common (see Fig. 2). The average maximum horizontal stress (s1) in ~100 m bore-holes in central Singapore has a 13oE orientation(7). My analysis of the stress vectors in the Sumatra section of the Australian Plate collision in the Sumatra-Java trench and subduction zone to the SW of Singapore, suggests that the maximum horizontal stress in the hangingwall of the subduction zone is oriented 30oE. The present day stress map of SE Asia(8), based on earthquake fault solutions, shows that the maximum horizontal stress is orientated NE/SW in neighbouring Sumatra. Furthermore, my lineament map of Singapore, based on topography, geological map and satellite data shows a very strong NE/SW trend (35oE, see Fig. 1).The important implication is that any joints (of whatever age) that are orientated NE/SW will be open in this stress regime.
I have measured joints and fault orientations in various Singaporian granite and gabbro quarries and there is a NE/SW correlation with the lineament map (see Fig. 2). The prediction is that granite with a strong NE/SW orientation of open joint sets will preferentially channel ground water in a NE/SW direction away from the watersheds. Figure 1. shows that the confirmed hot springs in Singapore are indeed NW of their associated watershed maxima.
Ground water model for Singapore
The USGS groundwater model for islands with an unconfined aquifer surrounded by seawater can be applied to Singapore(9). Assuming seawater has a density (Ps) of 1.025 g/cc compared with fresh water (Pf) of 1.0 g/cc, then according to the Gyben-Herzberg relation,
z = Pf . h
Ps - Pf
a head (h) of 100m above sea level in the centre of the island will drive cold fresh water down 4 km below sea level (z). I have confirmed that there is a permanent (cold) spring at an altitude of 120 m on Bukit Temah, the highest hill in Singapore (164m), indicating that such a high water table is probable. Assuming that the average geothermal gradient for the Singapore region is 35 o C/km(10), ground water at 4 km depth will reach 140 o C. Because of the high rainfall (2.4 m/year) and the +100m head, the hot ground water will be driven along the fresh water/seawater transition and up to the surface at the coast (Fig. 3). As described above, this hot water is likely to be preferentially channeled along NE/SW orientated joints and fractures. The Sembawang hot spring is 3.7 m above sea level which was at sea level during the warm interglacial periods at ~80 ka and ~125 ka(11). The Pulau Tekong hot spring is at 0.5 m above sea level, and from photos, it looks to be in the mangrove transition.
Exploration and Appraisal
Pump testing of 100 m deep wells at the Sembawang hot spring produced up to 400 l/min at a constant 70 o C for many days(5). However, this rate is not high enough to support a commercial power plant. Deeper production wells are required to intercept hotter water and these need to be coupled with injection wells to supplement the artesian flow.
Geophysical surveys: e.g. gravity, magnetics, 3D seismics, electrical resistivity, and heat flow surveys (i.e. infrared radar and 0.5 to 100 m bore holes) are required to locate a 2 km deep exploration well that will hopefully intercept significantly hot artesian water. Ground water age dating (3H/3He) plus SF6 and CFC analysis would be useful to model artesian flow rates.
Proof of Concept
Proof of concept for a 50 MW power station, providing power for 50,000 homes by tapping 150oC water at 2 km depth might cost USD 19 million: i.e. two 3 km long L-shaped wells at USD 3 million/km, plus USD 1 million for a 1 km hydro-fracture job. The horizontal part of the wells should be 1 km long and orientated NW/SE so as to maximize intersections of NE/SW trending joint and fracture sets (Fig.4). If this was proved as successful, another pair of L-shaped wells could be drilled from the same platform, at 180o to the first pair.
Development Costs
Development costs for the 200 kW binary generation system based on the commercial Alaskan shallow (200 m) hot wet granite binary system using massed produced refrigeration components were US$1300/kW(3), i.e. US$ 1.3 million/MW. Production costs at 5 US cents/kWh and selling at a domestic rate of 12 US cents/kWh generates a profit of 7 cents/kWh, i.e. US$70/MWh. Based on these figures, in one year a proof of concept 1 MW power station (for 1000 homes) might make 70 x 1 x 365 x 24 = US$ 0.613 million “profit” before tax. The development costs could be written off in 1.3 / 0.613 = 2.12 years and save the equivalent of 10,000 barrels of oil/year (with crude oil at US$100 oil/barrel this is a saving of US$ 1 million/year)
Development costs in California Geysers district, using conventional steam generation from 2 km depths are US$2.9 million/MW(4). A 50 MW power station (providing for 50,000 homes or a good proportion of the power used on Singapore Mass Rapid Transportation railway system would therefore cost 50 x 2.9 = US$ 149 million. Production costs at 5 cents/kWh and selling at a domestic rate of 12 cents/kWh generates a profit of 7 cents/kWh, i.e. US$70/MWh. In one year a 50 MW power station might make 70 x 50 x 365 x 24 = US$30.6m “profit”. Write off would take 149/30.6 = 4.9 years and 50 MW saves the equivalent of 0.5 million barrels of oil/year which at US$100/ bbl = US$50 million/year.
Environmental Impact
An environmental impact assessment would be a high priority. Geothermal energy is viewed as renewable, clean and green with a small carbon footprint during the construction phase. No doubt, a rig capable of drilling 3 km long directional wells would be disruptive in Singapore’s mainly urban environment. All the drilling for a power station could be conducted from one location. The hydraulic fracture job might create micro-seismicity, but Singapore already experiences micro-seismicity from the Java/Sumatra subduction zone. The infra structure for a 1 MW ‘proof of concept’ power station would fit into 2 or 3 tennis courts. A 50 MW power station could perhaps be located on an area the size of three or four football fields. Pipe work and high tension transmission lines would be placed underground. Water supply in Singapore is an issue and there would be an initial requirement to augment the working hydrothermal fracture system with fresh or storm water, but not sea water which could cause scaling. Once the system was pressured up and if the injector/production well connections were efficient, the requirement for augmented water would drop. Air rather than water cooling might be installed to condense turbine vapour for recycling. Sembawang hot spring has virtually no smell but Pulau Tekong hot spring is reported to be H2S-rich.
Summary
The Singaporean method of geothermal power generation would involve the drilling of 3 km deep directional wells into hot, wet, fractured granite under a thermal blanket and the generation of electricity either from70 o C hot water through binary cycle turbines or from ~150 o C conventional steam turbines, with the cooled water being recycled down injection wells. Proof of concept for a 50 MW power station might cost USD 19 million. Development costs could be written off in 5 years. There is a “local” market of 4.8 million people and the generating costs would remain static whilst the cost of natural gas varies. This is ‘renewable’, clean, green power generation and of strategic importance for a country that is viewed conventionally as having no natural resources.
References: web sources were accessed from February to August, 2009.
(1) http://www.searchanddiscovery.com/documents/2008/08181goldstein/index.html.
(2) http://useconomy.about.com/od/candidatesandtheeconomy/a/Obama_Stimulus.htm.
(3) http://www.chenahotsprings.com/geothermal-power/.
(4) www.geyser.com.
(5) Zhao, Chen & Cai (2002) Bull. Eng. Geol. Envir., 61, 59-71.
(6) Bowen (1986) Groundwater Technology and Engineering. Springer.
(7) Zhow (2001) Engineering Geology and Rock Mass Properties of the
Bukit Timah Granite. Report, Singapore Defence Science & Technology
Agency.
(8) http://www-wsm.physik.unikarlsruhe.de/pub/stress_data_frame.html
(9) http://pubs.usgs.gov/circ/2003/circ1262/#heading156057192.
(10) Mazlan et al. (1999) In : PETRONAS. The Petroleum Geology and
Resources Malaysia.
(11) Lambeck et al. (2002) Links between climate change and sea levels for the
past million years. Nature, 114, 199-206.
Followers
Blog Archive
About Me
- Grahame Oliver
- Singapore, Singapore, Singapore
- Although I am employed to teach petroleum geosciences, I can’t help thinking that Singapore is an attractive geothermal energy exploration target. This is a summary of a presentation I recently gave at a seminar at NUS.