Geothermal electricity generation

From Climate Etc.

by Chris Morris

Geothermal power stations are mature technology with proven performance, reliable operation and ideal for baseload generation. The units are synchronous, so they support the grid.  The production from them is considered by most to be renewable. They do not use fossil fuels to provide the heat. It is not “carbon free”, but no generation truly is. It has a relatively small footprint, environment harm is low, and it can coexist with farming or industrial development. Most developments have a cheaper energy cost than onshore wind, using published accounts for analysis. For countries or areas where the resource is there, geothermal generation is very viable.

The resource

Geothermal power stations are very much a niche generation source (only about 15GW worldwide,  from 673 units at 198 fields according to Google), totally dependent on locality. They are mainly associated with plate boundaries, particularly the Pacific Ring of Fire. Compare the plate boundaries and volcanic activity in Figure 1 with station locations in Figure  2

Associated with the plate boundaries and other weak points in the earth’s crust, the deep underlying heat in the mantle can find its way to the surface easier. “Bubbles” of magma can push up to relatively shallow depths. These may force their way to the actual surface as volcanoes with their lava. With the distortion and earth movement from this activity, the crust’s rock formations are deformed and cracked – earthquakes.  Groundwater can enter all the fault cracking in the rocks. This will be heated up by the hot magma, even if that has solidified.

Geothermal resources exploited for power production are the plumes of hot water formed from the heating of this deep groundwater. In geologic terms, such convection systems are short lived – generally lasting between 200 and 450 thousand years. They end because the heat source has gone or the cracking has been filled by precipitated minerals from the circulating water as it cools. The world is full of solidified magma (granite) and prehistoric geothermal systems. Many of the latter are now mined for gold and other precious materials.

Fig 1                 A simplified map of the Pacific mid-ocean spreading ridge, the plates, the subducting trenches and the volcanoes in the Pacific Ring of Fire. The named active volcanoes are the famous ones. In reality, and depending on how you define geologically active or what is part of the ring (Antarctica? Indonesia?), there are around 500

Fig 2     The localities of geothermal power stations around the world. There are sometimes many plant associated with one dot like New Zealand or Indonesia. They are a combination of conventional separated water steam turbines and binary plants

At a conceptual level, geothermal resources generally comprise four main components: a heat source, a fluid filled permeable rock structure (reservoir), a near impermeable cap, and surface features. The heat source comprises a localised body of molten or hot rock deep in the crust. This body heats deep circulating groundwater – in this context, deep is generally greater than 6 km. The resulting buoyancy causes the hot water to rise towards the surface through the cracked rock as a plume.

As the water is heated up, it dissolves rock. This enlarges the size of cracks, enhancing permeability (the interconnectedness of voids allowing fluid to flow through). At 300°C, gold is soluble. The water rises up cracks in hot rocks towards the shallow section of the crust, maybe less than 3 km deep. As the water is cooled nearer the surface (less than 500 m depth), the minerals in it precipitate and alter the structure of surface rocks. Sandstones and particularly mudstones can easily be turned to clay and have naturally low permeability. These actions form a cap on the reservoir. Inevitably, some of the geothermal reservoir fluid or heat leaks past the alteration cap to the surface, forming hot springs, geysers and the like.

Geoscientists locate conventional geothermal resources by mapping surface features and measuring the geophysical properties. The rock at the top of the reservoir has low earth resistance compared to the cold surrounding rock. The geothermal fields that exist are often shown by those boundaries (Figure 3). Wells drilled inside the boundary will be hot, but may not have permeability. Wells drilled outside it will be cold.

Fig 3     A map of the Wairakei Tauhara geothermal field with its limits defined by the earth resistance boundary. The map is out of date. There is an extra station at the Huka site and a new one about 2km east of there will more wells for both.

 The dissolved mineral concentrations in the resource are dependent on the temperature and rocks the water flows through. Typically, the major component is salt, but there is also silica. There are dissolved gases, mainly CO2 but some hydrogen sulphide. There will also be environmental nasties like arsenic and mercury. At the boundaries of the plume where the fluid is cooled, the minerals will precipitate out. These effectively seal the hot region from the surrounding rock. The salty fluid gives that very distinctive low earth resistance helps define the size of the resource.

Extracting the resource

Into these prospects, wells are drilled using oil & gas rigs but modified for the hot conditions and pressurised water. The wells generally have larger diameter production casings (200mm or bigger) and are open hole below the casing shoe, which is set below the cap rock and at the top of the hot zone. When the valves on the wells are opened, they can discharge the geothermally heated water. (Figure 4)

Figure 4            A modern drilling rig capable of drilling a 3km deep deviated well, a schematic of the drilling operation and a vertical discharge of a new well to blow all the drill chips and debris out of the hole before the well can be hooked up to pipelines. The discharge from this well was about 50% boiling water

To understand how this fluid can be used, it is necessary to understand the thermodynamics, particularly enthalpy – the practical engineering side, not the theory.   Enthalpy is the heat content of the fluid but it also relates to the phase (water/ steam), the temperatures and pressures. The temperature at which water boils changes with pressure; the saturation line. However, one has to add a lot of heat to get from just boiling water to dry steam (no water present) At sea level, this is about seven times as much as to get it from ice water up to boiling. If the heat content is between these two points at a given pressure, the fluid is a mixture- two phase. As the boiling pressure rises, the heat and density difference between the hot water and dry steam at that temperature decreases. Conversely, below atmospheric pressure, it diverges.  If the temperature of steam is above the saturation line, it is superheated. When the temperature and pressures are high enough, there is no difference between the physical properties of steam and water – the critical point.  Both of those conditions aren’t relevant for existing geothermal. Almost all the work is in the difficult two phase region. Note one can change water to steam and vice versa just by changing the pressure, without adding or taking away heat.

Most geothermal plants in the world run on separated steam as the deep fluid is hot water (>220°C) at very high pressure. As the fluid comes up the well bore, the pressure drops and steam flashes off so the two phase mixture at the surface is both steam and water. Steam mass fraction depends on enthalpy and wellhead pressure but is typically 20-30%.  This steam has to be separated from the water in surface plant before it can be used. It is done in cyclone separators (Figure 5).  The reason it needs separation is two phase fluid is very difficult to deal with. The flow regime is unpredictable. Pipelines carrying the steam water mix are subject to heavy shock loading, even in normal operation.  It is easier to separate into its component parts and deal with each separately.

The steam out of separators is always just below saturation line as they are not 100% efficient. There is no energy available to superheat it. To remove the mineralised carryover water in the steam, the practice is to use the long pipelines from separators to station. This allows gravity separation or condensation, then removal of the water at special drains. Modern practice is to build separators closer to stations, spray in clean water to wash the steam, and dry the steam in a scrubber using centrifugal force to fling the water to the walls for drainage and discharge.

Fig 5     This is how the cyclone separators work for geothermal fluid and their actual physical size nowadays. Each of the vessels is rated for about 400t/h steam flow  at 5bg.

To improve plant efficiency for high temperature resources, the separated water can be passed through a control valve to a lower pressure (Figure 6). The steam that is flashed off can be separated and fed into either another turbine, or a port part way through the main turbine (Figures 7 & 8)

Fig 6     A simplified process flow diagram for a cascading triple flash system feeding three pressures of steam into a turbine. This is what is at geothermal stations like Ngawapurua and Tauhara

Because of environmental concerns about the heavily mineralised separated water and to minimise the deep pressure decline of the resource, the hot (90-130°C) separated water is reinjected. This is done into dedicated wells located at the field margins. The ideal site is somewhere with a deep pressure communication with the resource but not close enough to quench the hot fluid. The industry rule of thumb is the water should stay underground for at least six months before being discharged. In that time, it should have been heated up enough by the rock it passes through not to affect the production enthalpy.

The water is generally supersaturated with minerals, particularly silica. That will precipitate in pipelines and wells, clogging them up.  To stop this happening, the water is acidified.  

At some fields, the pressure decline has allowed a steam pocket and two phase zone to form above the liquid and under the cap rock. Relatively shallow wells can be drilled into this, giving higher enthalpy, even dry steam discharges.

Steam Turbines

Most of the geothermal power stations use steam driven turbines.  They are proven technology. There is only one moving part, the rotor. They are very reliable. A station I work at operates some turbines that have done over 450k running hours and much of their componentry is still original. However, because the steam is only saturated (not superheated) at the inlet, the design details are significantly different to those on conventional boiler plant.

 A turbine is just a heat engine, where some of the enthalpy in pressurised steam is converted to velocity as it passes through a narrow nozzle to a region of lower pressure. This high velocity steam hits the blade of a rotor, forcing it to rotate. The heat energy has been converted to rotational energy. The slowed steam is expanded again (another enthalpy drop) for more energy extraction. The steam temperature drops with pressure but the enthalpy drop makes the steam wetter. Power output is proportional to mass flowrate and enthalpy change. If the turbine was running on compressed air rather than steam, the output would be significantly lower.

Each set of stationary blading then rotor blading is called a stage. Turbines typically have 5 to 12 stages, depending on the inlet and condenser pressure.  them with their relatively low inlet pressures, most of the power is extracted in the last three or four stages using sub-atmospheric pressure steam.  A big limitation is high strength steels can’t be used in a turbine as hydrogen sulphide makes them crack. That restricts maximum size of the rotor blades which limits their output to 60-150MW range, depending on inlet pressure. More powerful units than that need parallel steam path doubled turbines. On a boiler plant in a comparable sized turbine hall to the biggest geothermal units, but with the higher pressures and use of special steels, there are 4-700MW units which have 40-50 stages spread over two or three  turbines in series.

Wet steam all through the geothermal unit makes it different to boiler plant where only last few stages are wet. However for both, by the last stage of blading about 10% of the steam has been condensed to water. No energy can be extracted from the water and the high velocity droplets are very damaging to componentry. This water lowers efficiency, increases maintenance costs and reduces plant life. The water needs removal so careful capture and drainage systems are designed and built into the rotor blades and casings for the wet steam region.

Fig 7     the rotor from a dual flow triple flash steam turbine plant – still new coming out of the box. The high pressure steam is fed into the middle. It passes through 4 stages of blading. More steam is added and it goes through 3 more stages. Then low pressure steam is added for another 4 stages before it exhaust into the condenser.

Fig 8     Showing how the steam supply as in fig 6 for the rotor is arranged – rotor at top of picture, casing at bottom. A central HP annulus, with IP either side and LP ones outboard of that again. The cavity on the far right is a casing drain to remove condensate from the steam.

The turbine generators running on separated water fields are base-loaded. This is because there is minimal cost for the “fuel” and stable operation reduces manning requirements. The turbine design is optimised to perform best at full load.  If load reduction is needed, the steam has to be vented until wells can be shut in. Increasing the output from a throttled up well by opening the valves has to be done slowly to allow downhole conditions, surface two phase flow and the chemistry to stabilise.

At a few stations, Geysers in California is one, the fluid out of the ground is near dry steam that doesn’t need separators and can be supplied directly to the turbines. On these fields, plant can load follow, ramping up and down as dispatched. However, they are the exception.

Binary Plant

There is another type of conventional geothermal plant gaining in popularity, the binary ones. They are particularly good for lower enthalpy fields where conventional plant would be uneconomic. Their process uses a circulating/ working fluid like conventional boiler plant but rather than water, a lower boiling point organic fluid is used. This is often one of the pentanes. Instead of an actual boiler, there are a series of shell and tube heat exchangers through which the geothermal fluid is cascaded to boil the pressurised working fluid. It is even slightly superheated. This vapour is then expands through the turbine (similar to those on steam plant)  and is converted back to liquid in the condenser before being pumped for recirculating through the process (Figure 9). These plants invariably have air cooled condensers.

Material properties of the vapour means turbines are half grid speed, so less inertia. It also limits their maximum size. The lower enthalpy available means higher mass flow rate needed when compared to a steam turbine. That is more pumping. They have a proportionally higher parasitic load with all the motor driven fans and pumps, lowering their nett output.. Because of their design and control systems, they are also baseload with no significant ability to load follow without wasting energy.

Fig 9                 A simplified process flow diagram for a binary plant

For higher enthalpy resources, they are less efficient that steam turbines. However, they are cheaper and faster to build. In an era where lower capital cost and speed of installation dominates the economic modelling, these are significant benefits. They are also small (typically 5-25MW) and modular. This means the field can be initially developed with one or two units, then more added if the production shows it can take a higher energy extraction rate. Another advantage of these plants are they come as near complete packages from the manufacturer. They just need the wells and pipelines connected up.  It is a significantly easier task to design and build steamfield surface facilities with separation plant and pipelines than it is to design and build a power station. 

Why it has limited future expansion potential

As well as positives for geothermal development outlined above, there are the negatives. For many developments, the costs are hard to predict and over which an organisation can have little control. The regulatory environment, both national and local, can be challenging to navigate. There is a long lead time between field investigation beginning and electricity generation starting for the issues described below. 

Most geothermal developments are small, 10-50MW. Countries are often looking for 500-2000MW stations like they can get from gas turbines or coal. Many of the remaining best sites for development are in 3^rd world countries. Unless the country is prepare to have the development being proven, designed, built and run by expats (and pay for that privilege), they haven’t the educated professional workforce to do that. Iceland and New Zealand universities run geothermal training programmes for graduates from those developing countries, but many of them gravitate to better paying jobs in countries with existing plants.

Any new field has to be proven to have both permeability and sustainability before power station building starts. This needs an extensive well drilling and testing programme. Deep wells are expensive and permeability is elusive so even infill wells have a significant failure rate. The deep water temperatures need to generally be greater than 200°C otherwise the field is uneconomic, needing subsidies for development and exploitation.

Turbines and balance of plant are bespoke, needing the field output to be known before it can be sized. The industry history is of plant too big for the resource. One company in NZ operates a turbine purchased second-hand which had sat unused in a San Francisco warehouse for a decade, because it was for a field that couldn’t supply enough steam to the existing plant.

Once production starts, design failings and operational problems can occur. These often need extensive alterations and outages (lost generation income) to correct. Turbines and equipment working in a wet gassy steam environment where hard steels can’t be used is challenging with many lessons needing to be relearned. Big name plant manufacturers still get it wrong.

There is a continuing need for new well drilling and workovers as the field changes under exploitation. If the field enthalpy drops, the steam flow and wellhead pressure decreases, so turbines need to be derated to maximise output.. Finding suitable re-injection formations can be a very expensive exercise. The plants are often below rating because of steam shortfall, waiting for enough downhole work to accumulate to justify a drilling programme.

There are real localised environmental risks that mismanaged exploitation will damage natural geothermal features, even though these themselves are geologically speaking fleetingly transient. Excessive nett mass withdrawal can cause dewatering of cap rock formations. This may cause ground deformation, even significant subsidence. There was 15m! in a very localised area of Wairakei.

For efficiency of plant, it depends on what story you want to tell. Geothermal steam for electricity generation is a low value product. Depending on inlet and condenser pressures, it is 5-10t/h/MW. Contrast that with a boiler plant where it is about 2t/h/MW. Detractors of geothermal point to a very low 20-30% on First Law thermodynamics principles. Advocates prefer the Second Law (isentropic) efficiency which is generally over 80%. Binary plant is generally up to 10% lower than these figures but as they are often on lower temperature resources, that isn’t necessarily a true apples for apples comparison.

Like all energy production, society has to balance the costs with the benefits. For countries where the resource is there, they are a very good, albeit niche, electricity generation investment.

Proposed expansion

There are three developments that promoters push as the future for major expansion, making it mainstream. This alternative energy investigation has been supercharged by being funded by governments wanting to be seen to be doing something about climate change. They are low temperature resources,  Enhanced Geothermal Systems (EGS) also known as hot dry rock, and supercritical geothermal. 

Note that the internet is full of PR and academic writings about major breakthroughs that will change the face of the geothermal power station industry. Yet just a few years later, those pronouncements haven’t come to pass and the promises have sunk without trace. Theory is just that. Reality is cruel. It needs something working and has been doing so reliably for five years. That is proof of successful technology which will then relatively quickly be adopted. Until then, it is invariably just something seeking government funding.

It will be interesting to follow what happens to geothermal development in the USA from the new government and its change in energy production direction. Where the US goes, the rest of the world will follow.

Low Temperature Resources

As well as the high temperature (>200°C) resources able to be used by existing conventional power stations, there has been a push to exploit the wider availability of elevated ground temperatures. This has been promoted by maps like the one shown in Figure 10. The map is a misleading guide to potential viability. First the fluid is deep so would need significant drilling capability. Second, the permeability is unknown. Third, the wells generally won’t sustain a discharge so they have to have downhole pumps to bring the hot fluid to the surface. Fourth, the Carnot cycle efficiency is an economics killer. As a practical example of this point: for a massflow rate of 3500t/h 280°C water,  170MW is supplied to the grid at one station. Another on a reinjection system takes 3000t/h of 130°C water and  produces only 14MW.  

With sufficient flow rate, a binary plant using a lower temperature geothermal fluid can run and generate more power than it consumes to operate. The working fluid used in the plant can be tailored to suit the actual temperatures. These are often various refrigerants. But invariably, the plants aren’t viable without subsidies or there are specific advantages, like it is for an isolated community and the station is replacing diesel engines.

Fig 10             Map showing the deeper (>3km) rock temperatures for continental USA. The colouration makes one think that there is a lot more potential than there really is. Red appears to be temperatures >90° though the legend does not put temperatures directly on it. If the rock fracturing could be controlled, wells in those areas could be used for direct heating but not economically for electricity production.

However, if there are subsidies available like those in the USA, particularly for Californian electricity supply, plants can operate. There are about 1GW of pumped well power projects mostly in the western US states.  With all the pumps and their reliability issues, the load factor is lower than what conventional plant do. As the resource is lower temperature, the output in summer markedly declines – that pesky Carnot cycle again. On the map shown in Figure 2, there are two power stations shown in Central Australia. These are micro stations on low temperature resources and they don’t work. 

NREL in the 2023 report as well as reporting on economics of stations also sees a major use of the lower temperature fluid as district or process heating. Most agree this direct heating is is a lot more efficient use of the resource. Ground source heat exchangers are more effective if in the water table. In Taupo NZ, the shallow heat is directly used for domestic and public facilities including several large open air year-round swimming facilities. Geothermal heat is used for drying timber and wood pellets. The heat from a reinjection water line is used to grow tropical prawns (Figure 11). At a nearby geothermal resource, it is used to provide process heat for a dairy factory and greenhouses. The author’s home uses the hot underlying ground water to heat up town supply water through a U- bend heat exchanger in a shallow well to provide hot water for his household.

Fig 11   Ponds growing tropical prawns heated by water from a re-injection line. Behind the vent steam plumes is a 14MW binary plant running on 130°C water. The water is from the discharge of steam separators.

EGS

The theory of EGS is simple. It was designed for places where the rock is hot but there is poor permeability. Two wells are drilled side by side 1-200m apart. The rock between them in the reservoir formation is hydraulically fractured to give the permeability. Cold fluid pumped down one well is heated up by the rock and discharges out the other well where it can be used, then disposed of down the cold well. Figure 12 is a schematic.

The problem is the desired controlled rock fracturing can’t be done. In most cases, there is no significant increase in permeability. Where the wells have been close enough to get communication, there has been thermal breakthrough and the cold fluid has rapidly quenched the rock. There have been no successes with the world littered by failed projects, but promoters are undeterred, wanting to continue.

Figure 12          The type of schematic used by promoters of Hot Dry Rock proposals. Simple in theory. It hasn’t worked yet in practice.

 Supercritical Geothermal

Underneath geothermal fields at depths greater than 3km (about the limit of current geothermal well drilling technology in hot volcanic rock) the science says the fluid will be a lot hotter, maybe above the critical point ~400°C,  The theory is if one is to drill into this supercritical fluid, there will be a very high temperature resource to exploit. The sticking point is the technology.  Talk is the casings would be ceramic as all standard ones or even specials like high chromium steel or titanium won’t work. The current equipment has failed.  A whole new drilling equipment system with exotic materials would be needed to drill and complete the wells. Even a conventional well in a standard geothermal field to the depths discussed would be very expensive. 

Some wells have intercepted this deep fluid at a relatively shallow depth. It was found to be heavily mineralised and very acidic. One well in Mexico even discharged hydrogen chloride gas. The linked article also details other major problems that have occurred.  There are no commonly used (and not prohibitively expensive) materials that could contain this fluid for power station use. At the predicted temperature, the water will dissolve gold.

It is yet another example of where the theoretical value is there but the materials to handle it haven’t been invented.

Acknowledgements       JC for providing the impetus and challenge to do this article. Planning Engineer for forcing me to distil and simplify my writing. Rutherford’s barmaid dictum should be a guiding principle for all.  Most of all, I thank my work colleagues present and past for taking the time to explain the intricacies of their specialities. Things are the way they are because that is the way they have been proven to work best within the real-world physical and economic limits.

Afterword      Note that when describing the general geothermal power production industry, terminology and phrases, efforts at simplification may in some cases result in statements that are not 100% accurate in all situations. There can also be new developments that haven’t yet made the trade papers or even smoko discussions. The article’s broad scope and length limitations mean that all minor exceptions can’t be covered. For this overview, speaking generally is preferable to littering the post with distracting mealy-mouthed qualifiers. Please accept that the article is overwhelmingly correct as to the operation of geothermal plants at this time.