A geothermal ternary diagram of tensions.

There is a lot of discussion about which geothermal technology is best.

Call me a fence sitter, but I always think it depends. Depends on who wants it and what they want to use it for, when.

We can talk about hydrothermal, EGS, closed loop, whatever, but in an unashamed oversimplification, there is one fundamental thing we are interested in for geothermal, whatever we use it for or convert it into - and that is heat recovery from the subsurface.

The first question to think about is how much heat we actually want.  

Lets' say it again. How much heat do we actually want? In terms of heat use, if we are de-icing an engine testing area at Oslo airport, if we are growing tomatoes in a Dutch greenhouse, or if we are supplying power to Addis Ababa, the answers to that question will be very different.  It matters.

First, without getting too hung up on the difference between temperature and enthalpy, it's good to briefly talk to it. Suffice to say that we can get more energy out of things of different phase and pressure than temperature alone would indicate.  For binary power plants and direct heat uses though, we can approximate to temperature. For direct steam and flash power plants, consideration of enthalpy as well as temperature matters more.

The heat we get from a well into the subsurface is really driven, in our enthalpy to temperature approximation, by two simple things. Flow rate and temperature. Some uses become practically & commercially less feasible below certain thresholds of temperature - but it is not a hard boundary - lower temperatures can sometimes work for excellent higher flow rates.  That’s because heat delivery is a function of mass (of fluid accessed) as much as it is temperature. You can have great temperatures, but you are totally stymied if the flow rate is duff.

Usually about 140 deg C is a commercial sort of rule of thumb for power generation, but it’s not impossible you could tweak it a bit lower for really exceptional flow rates. Other uses that are not power might have certain temperature requirements, but if you have a good flow rate, for a bit of extra CAPEX in heat pumps and their OPEX power requirement, you can always lift the temperature to what's needed. So in some ways, for those uses, low flow rate is the one that is harder to deal with than lower temperature. Not without some compromise of cost though.

We can "access" the heat technically through a number of routes.

1)      Through pressurised fluid that circulates in the rocks naturally, or circulation in natural porosity and permeability which we enhance with injection to sustain pressure.

2)      Through closed systems that rely only on fluid in the well systems, and extract heat through conduction from the rock.

3)      Or we can artificially crack the rocks (Enhanced Geothermal Systems - EGS) to create or enhance circulations that would otherwise not be present/sufficient.

But whatever we do, we want to be able to sustain those temperatures over a project lifetime. It's no use having nice temperature and flow rate for 6 months in a 30-year life facility. That matters, because if we inject fluid in either closed or open systems, we cool the rocks down.

For closed systems this is more delicate, because they are much more restricted in where that cool goes, and by definition always have "injection".  That is to say, we have no choice in the matter, - unless there is some measure of natural background flow in the rock the closed loop accesses, we necessarily cool down the same small volume we are extracting heat from.  Whereas open systems utilise heat from much bigger volumes of porous rock collecting heat from a much areally larger heat “catchment”.  It’s like the difference between collecting rain in a bucket and a test tube.

The cooling from any injection can be taken further away in open systems. In some cases, if distally sourced and replenishing overpressures are providing a natural sustenance of pressure, injection might not even be required. All this means means closed systems might have an inherent ceiling on the flow rate, to limit cooling, that “kicks in”, far quicker than for an open system – limiting the flow rate in - and out - and hence the heat recovery. 

Whichever, open or closed, we seek that our sustained heat-need is such that the cooling from injection does not exceed - or bypass - the heating from the deep earth. For closed systems that constrains rate, because production rate = injection rate. It’s closed. You get out hot what you put in cold - by virtue of cooling rocks, which are reheated at some finite rate from below. Simply, this might not matter too much if we can get enough heat out without cooling things down too much, - “enough” to pay for our drilling, facility, and operations costs and make competitive profits on top.

For open systems, there is less constraint – the specific injection wells are not directly connected in the subsurface to the production.   What can sometimes happen though, in either natural or artificially engineered open systems, is that the injected cool water finds a permeability pathway we don’t want it to, and the cool gets a fast highway to the producer.  That is not good.  It might not happen instantly, but sometimes someway into the life of a project after a lot of investment has been made. There are increasingly ways of monitoring and averting that, but it remains a risk, and it’s tough to deal with once it happens. 

More to the point perhaps, for open systems, there is always the risk of investing in an expensive system which relies on inherent rock permeability and finding it doesn’t match expectation, hitting our flow rates. How serious that is depends on how much the hit is outside expected "ranges".  That risk is to an extent dependent on how well we know an area, and/or our rock mechanics to fracture intelligently where EGS is an option, but there is always an element of unknowable well scale permeability or rock rheology variability that we might not know in advance and which might throw “spanners” in our works.  The good news though, is that if we get such systems to work, we get access to a much bigger supply of heat than a closed loop system could furnish.   That can be by an order of magnitude – so it’s not a trivial factor in the economics.  

Talking of economics, superimposed on all of this is how deep we have to go to get the temperature and flow rate.   If we are closed loop, we are less worried about the rock other than its thermal conductivity character, but to a first approximation, we just want to go deep enough to get a decent temperature.  For open systems, we are constrained to the depths that the reservoirs are at.  Note though, that if we find a good one, we might be able to live with lower temperature and spend the CAPEX on heat pumps rather than deep drilling.   That is not as cut and dried as it might sound, since big heat pumps can be expensive, and advances in cheaper, faster, leaner, meaner, not-over-engineered, fit-for-purpose smart drilling is happening all the time. 

What we also have to admit though, is that, given the smaller volume supplying the heat for closed loop than open loop, we need to chase higher temperatures and greater depths to get sufficient sustainable heat and flow rates than we do with open systems, if we can find them. If naturally permeable reservoirs are rare in our area, EGS and closed loop are the only real geothermal options, but where good naturally permeable ones are present - and that isn't so rare, especially in the top 2-2.5 km, they potentially carry a much heftier heat "punch". Clearly, for both open and closed systems, having a high geothermal gradient (how quickly temperature increases with depth) is a factor. The higher the better - it makes the shallower, cheaper to access "reservoir" (in a combined heat and/or permeability sense), more useful in both cases.

Yet we come back to the original question of what we need it for.  If we want it just to do a bit of de-icing in the middle of winter, closed-loop might be fine.  GSHP are a variant of closed loop that has worked for many decades, fulfilling the relatively modest heat demands of space heating for homes and offices.   Some of them admittedly also make use of solar heating of the ground in the top 20m rather than the purely geothermal sensu-stricto below that, but the principle is there – if our heat requirement is relatively modest, closed-loop can sometimes offer it with far less risk.   But the deeper you have to do it, to meet bigger customer heat requirements, the trickier it becomes. Drilling gets more expensive the deeper you go, quickly, in a non-linear fashion. So while there is tied up in that “accessibility” corner of the triangle a few things – the depth and drilling spend (DRILLEX) we need to go to access the heat required - multilateral or not - is a big and important chunk of it.   

Another aspect of accessibility is related to some other regulations and risks, and what is acceptable to communities in which operations take place.  EGS for example might have a lot of technical attraction in some places.   I for example would love to think more about what it can do with radiothermal granite resource in Galicia, NW Spain, but simply – like it or loathe it - some countries don’t like EGS because of risks of induced seismicity – which whatever we might say, are never zero.   However keen Google or Microsoft might be on dabbling in remoter areas of the US for facilities, where the nature of the application and demand can easily follow the resource without sacrificing too much usefulness - and hence without shaking up too many nervous citizens in the event of an odd low-level jiggle.   

Different cultures and places have different acceptance levels of seismic risk, and monitoring and control is always improving,  but it will never improve to a point where the risk is zero.   Drilling EGS projects under cities in Spain, Sweden, Switzerland, or Korea, will then take on a very different “accessibility” complexion to drilling them in the remoter countryside of Utah, Nevada, New Mexico or Arizona.  Bearing in mind, that for many uses, geothermal resource is necessarily constrained to be proximal to its users – so cities and towns necessarily loom larger in geothermal resource exploitation.  That is more so for heat than power, but existing power grids still cost a lot to connect to, so it is not a trivial thing for either heat or power.  There are big industries which can be happy with an independent resource in the Styx, but most applications of geothermal have lots of neighbours on top to also please. And as we have seen over and over again, it only takes one big incident for perceptions – and regulations - to change very quickly and durably. So accessibility is not just about the drill-bit blow ground - it’s about what is regulated above ground too.

Might that aversion to induced seismicity change, with further examples and further demonstrations of reliability? Yes, I think that is in fact likely as we get better at it, but it illustrates the point that accessibility is about more than what we can engineer underground.  

Accessibility can also encompass other things. For dry steam and flash geothermal power plants, CO2 emissions can occasionally be an issue, though the intricacies of that topic are still evolving.   It may for example be at least sometimes the case that they tend to focus them in one place for a geothermal field rather than actually increase them anew.  For some projects, very adverse chemistries, affecting plant operation costs, might also impact “accessibility”.   That might for instance, be a more important feature for very hot aggressive and/or high total-dissolved solid fluids of the kinds being proposed in supercritical and brine mineral applications. So the “accessibility” vertex is most fundamentally driven by depth and well cost, but there are other things which feature there.

The sustainable flow rate though is always the elephant in the room.  The risk of it not being there for project life. Hit by either permeability, or cooling, that constrains it to less than we'd hope.  Would we rather have an 80% chance of a nice big juicy amount of heat being accessible by open systems, but some risk of disappointment, or an assured chance of significantly lesser amounts of heat, perhaps with higher drillex (drilling expenditure), and more questions over the sustainable rates? 

There is no right or wrong answer to that, it depends on the use, and the customer appetites for [which] risks.  None of them are without risks, but different aspects of the project are at risk. Also, to a degree, the choice depends on what various surface-based options can deliver instead.  If for example we are thinking of competing with wind and solar and some kind of energy storage for power supply, then hydrothermal might compete well in a success case – but it will carry that reservoir risk initially.  Closed loops can do it without worrying about reservoir - but might have to go significantly deeper (and costlier) to get a volume of rock that can stay hot enough long enough to supply a competitive amount of heat. It might be trickier for them to compete. There is no hard and fast answer, but what we can see, is that it is this trade-off of sustainable temperature, sustainable flow rate, and accessibility, which we are looking at and balancing for a particular need. 

So a bit of a ramble, but I think the message is hopefully emerging.  We can talk around the houses for a long-long time over what is “best” - but fundamentally all geothermal is jockeying for position on this ternary diagram of tensions between different aspects - to fulfil a particular customer’s requirement for heat.  All sorts of options can be considered, but they all involve trade-offs and compromises on the certainty of these three core things – temperature/enthalpy; flow rate; accessibility.  There are other nuances, but these rule the roost. 

Which geothermal option is most competitive, or is “best” is therefore not a hard-and-fast defined answer – but a function of what the customer wants it for, when; what the surface competition in their locality can offer; and which compromises of the things in this triangle deliver that best. Including, just like our own pension management, individual customer tastes for risk. The answer to the "which" is only answerable when conditioned by the "who".