The Frac’ing Dividend
June 29, 2025 § Leave a comment
Few will dispute the fact that the US was a net importer of natural gas in 2007. Cheniere Energy was gearing up to import liquefied natural gas (LNG) and deliver it to the country. Then natural gas from shale deposits became economic and scalable. How Cheniere reinvented their business model by pivoting to convert their re-gas terminals to become LNG producers is a story for another time. As is the tale of shale gas singlehandedly lifting the US out of recession. Low-cost energy is a tide that lifts all boats of economic prosperity, and this one sure did.
The story for today is that the technologies that enabled shale gas production are showing up as key enablers for geothermal energy, the leading carbon- free energy source which can operate 24/7/365, unlike solar and wind, which are intermittent, with capacity factors (roughly defined as the portion of time spent delivering electricity) of up to 25% and 40%, respectively. The short duration intermittency (frequently defined as under 10 hours) is covered by batteries. But for longer periods, the most promising gap fillers are geothermal energy, small modular (nuclear) reactors and innovative storage means. Of these, the furthest along are geothermal, as represented by advanced geothermal systems (AGS), some battery systems, and hydrogen from electrolysis of water using power when not needed by the grid.
The most commercially advanced AGS variant is that of Fervo Energy. It employs two parallel horizontal wells intended to be in fluid communication. Hydraulic pressure is used to create and propagate fractures from the injector well towards the induced fractures at the producer well. Fluid is pumped into the injector well and flows through the fracture network into the producer well. Along the way, it heats up while traversing hot rock. This hot fluid is used to generate electricity. It may also be used for district or industrial heating. The heat in the rock is replenished by heat transfer from near the center of the earth, where it is created by the decay of radioactive substances. The center of the earth is at temperatures close to that of the sun.
The two key enabling technologies for accomplishing the process described above are those of horizontal drilling (and drilling a pair reasonably parallel to each other) and hydraulic fracturing, known in industrial parlance as “frac’ing”. Clever modeling (from a co-founder’s PhD thesis at Stanford) dictates many operational parameters such as optimal well separation, but the guts of the operation is that employed in shale oil and gas production. With one difference. The temperatures are greater than in most conventional shale operations. This affects the drilling portion more than the fracturing one. Temperatures will be greater than 180 C in AGS operations, and in excess of 350 C in the so-called closed loop systems (which have no fracturing involved, except for one minor variant). Even in AGS systems, higher temperatures are preferred, and Fervo recently demonstrated operations at 250 C. While this stresses industry capability, it still falls firmly in oil and gas competency and the geothermal industry can rely upon this rather than attempt to invent in that space.
The frac’ing dividend mentioned in the title of this piece refers to advances in frac’ing operations which accrue directly to the benefit of AGS operations. A key one is the ability to use “slick water”, which is fracturing fluid with little to no chemicals. Opponents of shale gas operations have often cited the possibility of surface release of these chemicals as a concern. Similarly, accidental surface release of hydrocarbons in fracturing fluid has been a concern but is irrelevant here because of the complete absence of hydrocarbons in the rock being drilled. AGS operations do have the possibility of induced seismicity. This is where the pressure wave from the hydraulic pressure potentially energizes an active fault, if present in proximity, and the resulting slip (movement between rock segments) causes a sound wave in the seismic range. However, the risk is small, especially if the activity is in a naturally fractured zone, in which fractures can be propagated at lower induced pressures than in unfractured rock. In any event, competent operators such as Fervo are placing observation wells and to date induced seismicity has not been a concern. The sound heard is that of (relative) silence*.
Finally, the shale oil and gas industry devised economies of scale by placing multiple wells on each “pad”. These techniques, including elements such as the rigs moving swiftly on rails, are directly applicable to AGS systems using frac’ing. So, any reported economics of single well pairs could, in my opinion, be improved by up to 40% when tens of well pairs are executed on a single pad.
This is the big frac’ing dividend.
Vikram Rao
June 29, 2025
*And no one dared disturb the sound of silence, from The Sound of Silence, performed by Simon and Garfunkel, 1965, written by Paul Simon.
The Bulls Are Running in Natural Gas
January 2, 2025 § Leave a comment
Ukraine shut down the natural gas pipeline from Russia to southern Europe yesterday. While not unexpected, yet another red rag for natural gas bulls. And ascendency for liquefied natural gas (LNG) futures and associated increase in US influence on Europe because, according to the Energy Information Administration, most of new LNG supply in the world will be from the US. Of course, that means more fodder for the debate in the US on whether LNG exports will increase domestic prices more than mere price elasticity with demand. I see greater demand impact from a different source, but more on that below.
Natural gas usage will not be in decline anytime soon. In fact, usage will steadily increase for the next couple of decades. Much of the incremental usage will be for electricity production, with a business model twist: expect a trend to captive production “behind the meter”. Not having to deal with utilities will speed introduction. Of course, the entire production will have guaranteed offtake, but that will not be much of a hurdle for some of the deep pocketed applications owners.
So, what has changed? Why are fossil fuels not in decline in preference to carbon-free alternatives? Much of the answer is that all fossil fuels are not created equal. Ironically, oil and gas are created by precisely the same mechanism, but their usage and the associated emissions are horses of different colors. At some risk of oversimplification, oil is mostly about transportation and natural gas is mostly about electricity and space heating.
At a first cut, oil usage will reduce when carbon-free transport fuel alternatives take a hold. Think electric vehicles, methanol powered boats, biofuels for aviation and so forth. Similarly, natural gas usage reduction relies upon rate of growth of carbon-free electricity (note my use of carbon-free instead of renewable), which today is almost all solar and wind based. Advanced geothermal is nascent and nuclear is static except for some rumblings among small modular reactors (SMRs).
In the case of oil in transportation, when the switch to battery or hydrogen power arrives, oil will be fully displaced from that vehicle. Not counting hybrids in this discussion, nor lubricating oils. In electricity, however, solar and wind power being intermittent, some other means are needed to fill the gaps. Those means are dominantly natural gas powered today for longer duration (greater than 10 hours). For short durations, and diurnal variations, batteries get the job done for around 2 US cents per kWh. Very affordable, and unlikely to change. To underline the point, solar and wind need natural gas for continues supply. Until alternatives such as long duration storage, geothermal or SMRs make their presence felt, every installation of solar or wind increases natural gas usage.
As if that were not bad enough, a recent complication is increasing electricity demand. Artificial Intelligence, or AI, and to a greater extent the Generative AI variant, has increased electricity demand dramatically. For example, a search query uses 10 times the energy when employing Gen AI, compared to a similar conventional search. The information is presumably more useful, but the mere fact is that these searches and other applications such as in language, are power hogs. The data center folks are trolling for power in geothermal and nuclear. Microsoft went so far as to commission the de-mothballing of the 3 Mile Island conventional nuclear facility. Control room picture shows its age. The operators will face what an F18 pilot would, if asked to fly an F14 (apologies, just saw Top Gun Maverick film).
Almost all the big cloud folks want to use carbon-free power, 24/7/365. Good luck getting that from a utility. Some, such as Google with Fervo Energy geothermal, are enabling supply to the grid and capturing the credit. Others are going “behind the meter”, meaning captive supply not intended for the grid. The menu is geothermal, SMRs and innovative storage systems. All have extended times to get to scale. Is there an option that is more scalable sooner to suit the growth pattern of AI?
Natural gas. A combined cycle plant (electricity both from a gas turbine and from a later in cycle steam turbine) could be constructed in less than 18 months. Carbon capture is feasible, even though the lower CO2 concentrations of 3 to 5% (as compared to 12 to 15% for coal plants) makes it costlier (a reason I am bearish on direct air capture, with 0.04% concentration). At the current state of technology, I estimate that will add 3 to 4 cents per kWh. This technology will keep improving, but that number is already worth the price of admission, at least in the US, where the base natural gas price is low. Not renewable, but nor is nuclear. You see why I prefer the carbon-free language?
To be behind the meter, the plant would need to be proximal to the data center. Data centers prefer cool weather siting for ambient heat discharge. Reduces power usage. Since natural gas pipelines serve a wide area, this ought not to be a major constraint. However, it could favor producers in the northern latitudes, especially if a rich deposit is currently unconnected to a major pipeline. Favorable deals could be struck especially with long term offtake contracts in part because the gas operator will eliminate the markup by the midstream operator. These conditions could be met in Wyoming, Alberta (Canada) and, of course, Alaska.
We used to refer to natural gas as a bridge to renewables, until that phraseology fell out of vogue. The thinking was that gas could replace coal to provide some CO2 emissions relief (and it did that for the US), and eventually be replaced by renewable energy. The model suggested above does not fit that definition. Those plants will likely not be replaced because they would be essentially carbon neutral. And they would enable a powerful new technology the foundations of which already have been awarded the 2024 Nobel Prize in Physics. Gen AI may well lose some of its luster, but the machine learning underpinnings will survive and continue to deliver. All that will need data crunching. More data centers are firmly in our future.
Those that are still inclined to teeth gnashing on emissions from natural gas production ought to ponder nuclear spent fuel disposal, mining for silica for solar panels and the problem with disposal of disused wind sails, to name just a few. Every form of energy has warts*. We simply need to minimize them.
Vikram Rao
January 2, 2025
*Every rose has its thorn, from Every Rose Has Its Thorn, 1988, performed by Poison, written by Brett Michaels et al.
Drill Baby Drill, Drill Hot Rocks
December 5, 2024 § Leave a comment
“Drill baby drill” is being bandied around, especially post-election, reflecting the views of the president-elect. Thing is, though, baby’s already been drilling up a storm. World oil consumption was at an all-time high in 2023, breaking the 100 million barrel per day (MMbpd) barrier. And the International Energy Association (IEA) projects further demand growth, to about 106 MM bpd by 2028. The IEA also projects the US as the largest contributor to the supply, provided the sanctions on Russia and Iran continue.
Courtesy the International Energy Association
To execute the stated intent to stimulate US production, all that the new White House needs to do is not mess with the sanctions. For ideological reasons they may be tempted to open the Alaskan National Wildlife Refuge to leases. But none of the majors will come, and not even the larger independents. Easier pickings in shale oil and in wondrous new opportunities such as in Guyana. Is it still a party if nobody comes?
Note in the figure above that the projection by the IEA has roughly the same slope as the pre-pandemic period, with a bit of a dip in the out years ascribed to electric vehicles. And if that were not enough, world coal consumption hit a historic annual high of 8.7 billion tonnes in 2023, despite Britain, which invented the use of coal, closing its last mine this year. The largest increases were in Indonesia, India and China, in that order. Let me underline, both oil and coal hit all-time highs in usage last year. So much for the great energy transition.
So, what gives? China and India, two with the greatest uptick in coal usage, need energy for economic uplift, and for now that means coal for them, since they are net importers of oil and gas. Consider though that the same countries are numbers 1 and 3 in rate of adoption of solar energy. What this means is that solar and wind cannot scale fast enough to keep up with the demand. Making matters worse is the ever-increasing demand created by data centers.
One reason for not keeping up with demand is land mass required. Numbers vary by conditions, especially for wind, but solar energy needs about 5 acres per MW, while wind on flat land typically needs about 30 acres per MW. Compare that to a coal generating plant, which is 0.7 acres per MW (without carbon capture). Wind also tends to be far from populated areas, so transmission lines are needed, and much wind energy is curtailed due to those not being readily constructed. To add to the complication, both solar and wind plants have low capacity factors, under 40%. So, nameplate capacity is not achieved continuously, and augmentation is needed with batteries or other storage means. Finally, governments would like the communities with retired coal plants to benefit from the replacements. This is hard at many levels, not the least being availability of land mass, and because the land area required is many times that which was occupied by the coal plant being replaced. All this holds back scale.
Geothermal Energy. Two types of firm (high capacity factors) carbon-free energy that fit the bill in terms of land mass, are geothermal energy and small modular reactors. Here we will discuss just the former, which involves drilling wells into hot rock, pumping water in and recovering the hot fluid to drive turbines. Fervo Energy, in my opinion the leading enhanced geothermal (EGS) company (disclosure: I advise Fervo, and anything disclosed here is public information or my conjecture), has been approved for a 2 GW plant in Utah, which has a surface footprint of 633 acres. This calculates to about 0.3 acres per MW. The footprint of Sage Geosystems is also similar. Sage also has an innovative variant which takes advantage of the poroelasticity in rock, and which could provide load following backup storage for intermittency in solar and wind, thus enabling scale in a different way.
Aside from the favorable footprint of Fervo emplacements (incidentally, the underground footprint is significant because each of the over 300 wells is about a mile long), the technology is highly scalable for the following reasons. All unit operations are performed by oilfield personnel with no additional training, and therefore, readily available. Certainly, the technology is underpinned by unique modeling (developed in large part in the Stanford PhD thesis of a founder), but the key is that when oil and gas production eventually diminishes, the same personnel can be used here. In fact, an oil and gas company could have geothermal assets in addition to their oil and gas ones, and simply mix and match personnel as dictated by demand.
The shale oil and gas industry found that when multiple wells were operated on “pads”, cost per well came down significantly. Those learnings would apply directly to EGS. Accordingly, I would expect EGS systems at scale to deliver carbon free power, 24/7/365, at very favorable costs.
Governments and investors ought to take note that EGS variants are possibly the fastest means for economically displacing coal, and eventually oil. In the case of the latter, even that displacement does not eliminate jobs.
As the title revealed, the refrain now changes a bit to: Drill baby drill, drill hot rocks*.
Vikram Rao
* Lookin’ for some hot stuff, baby, in Hot Stuff by Donna Summer, 1979, Casablanca Records
Research Triangle Energy Consortium 