Geological Thermal Energy Storage (GeoTES): Combined with Concentrated Solar or Heat Pumps: Premier Resource Management’s GeoTES Project in California and Potentially Other Areas in the Southwest U.S.

 

     California company Premier Resource Management expected to drill and operate oil wells when they bought wells and leases in California’s Central Valley in 2018. This area in Kern County near Bakersfield is one of California’s largest legacy oil fields. They could not get permits to drill due to California’s regulatory environment which tends to be hostile to fossil fuel producers. Instead, beginning in 2020, they are focusing on geological thermal energy storage (GeoTES). This means utilizing favorable geology to store geothermal energy in the form of pre-heated water that is injected through one well and produced through another well when needed to power steam turbines to produce electricity.  Geological favorability also includes a reservoir temperature that will keep the water hot. Areas where the geothermal gradient, or temperature increase with depth, is higher than normal, are most amenable to geothermal energy storage. The company thinks that they can store the water in a usable state, heated at the surface with solar collectors utilizing parabolic mirrors to concentrate the energy enough to heat the water, for a month or more. They plan to use the brackish water already in the reservoirs. They plan to pump it to the surface, then heat it to 700 degrees F, run it through heat exchangers, and then inject it back underground. This project involves drilling new geothermal wells, but other projects may utilize existing wells.

     The company is partnering with the National Renewable Energy Laboratory (NREL), Berkeley National Lab, Idaho National Lab, and industry partner Ramsgate Engineering. This first pilot is still in the planning and permitting stage. It is hoped that the demonstration plant will begin construction in 2026 or 2027. NREL is also working with a company in Texas to store energy in existing wells, but in reservoirs that have not produced hydrocarbons, perhaps in deep saline aquifers, but maybe in shallow freshwater or brackish water aquifers as well.

     Geological thermal energy storage utilizing existing oil and gas wells being combined with solar heating requires certain conditions to be most successful and economical: adequate sun, suitable geology, distance from homes or sensitive areas where water contamination could be an issue, and proximity to power transmission lines.

     Premier’s CEO Mike Umbro thinks that eventually, California’s San Joachin Valley alone can support 60 GW of GeoTES. He noted some other advantages such as the project uses existing oilfield land with no new land disturbed and the components of the system will all be low to the ground and so less of an eyesore than oilfield equipment. They also tout jobs. They think their project will support 200-400 construction jobs lasting 10 years or more and 100 ongoing jobs. According to a 2023 report the pilot demo project is expected to produce 10 MW of electrical power for five hours every night. According to the same article in ThinkGeoEnergy:

 

“The company is planning to construct 60-acre solar arrays and a series of tanks for separating and cleaning the water. Energy will be stored in 37 geothermal wells. The system would then be connected to a nearby substation and power transmission lines owned by Pacific Gas and Electric Co.”

    

Umbro stated then:

“We believe the oil fields could meet roughly half California’s 2045 long duration energy storage goals — with 45 gigawatts of potential on the west side (of Kern) alone.”

     If the pilot project works as designed the company plans to expand the project to 400MW of energy storage at a cost of about $2 billion. A project that size could power Bakersfield as needed. GeoTES provides long-duration energy storage that can provide needed support for seasonal low output of solar in the winter. When this happens California turns to natural gas. As a state, California is one of the biggest consumers of natural gas. Gas is especially needed on hot summer days and in general through the winter. Winter natural gas price spikes are common. Such systems could mitigate these issues. Of course, the potential avoided carbon emissions are desirable as well and should help to speed up permitting, although that does not seem likely.

Schematic of Premier Resource Management's Project

     NREL describes the project as follows in terms of its components: reservoir circulation, solar heat collection, and power generation:

Reservoir Circulation:  The project will be equipped with multiple producing and injecting wells in a “seven-spot” arrangement.  Seven spots typically possess improved reservoir contact and increased lifting capacity where reduced, flow-related pressure drop in the reservoir is desired, when compared to five-spot geometry.  The reservoir circulation loop will operate with varying circulation rates depending on demands made by the two other, interacting loops.

Solar Heat Collection:  Solar heat will be collected using helio-dynamic, parabolic trough-style solar concentrators.  Heat will be absorbed into a circulating working fluid, it being heated to roughly 700F.  As heat is collected this loop will command the Reservoir Circulation loop to provide sufficient fluids to absorb the collected solar heat.

Power Generation:  The Reservoir Circulation loop will provide heated fluids sufficient to boil and superheat a power-producing working fluid, which will be circulated through a power turbine.  When power is demanded by the power grid this loop will command the Reservoir Circulation loop to deliver sufficient heat for power production purposes.

The pilot project will consist of seven, 2½ acre seven spot patterns.  Roughly 40 acres of solar collectors will be installed to support the process heating requirement and a 10MW peaking turbine/generator will be installed to generate pilot project sales-power.

Cost estimates and a timeline for the project is shown below:

Source: NREL

 

Different Configurations of Geological Thermal Energy Storage Being Explored by NREL and Others

     The above project utilizes concentrated solar to heat the water. The other main way to heat the water is via a heat pump system which uses electricity. This is known as a Carnot Battery. The heat can be used for industrial processes as well. NREL compares costs, as levelized cost of storage (LCOS), for GeoTES vs. other types of energy storage:

 “… a GeoTES charged with solar thermal energy and calculated it to have a levelized cost of storage (LCOS) of 0.12 $/kWhe for 700 hours of capacity. This value was low compared to other comparable technologies at the same scale, such as hydrogen (0.5 $/kWhe), compressed air energy storage (2.8 $/kWhe), and pumped hydro-electric storage (1.6 $/kWhe) (Sharan et al., 2020). These low costs derive from the fact that – unlike other storage systems – the GeoTES storage volume has little-to-no cost. Wells provide access to the reservoir and determine the rate that energy can be extracted (and therefore the cost of power), but the marginal cost of adding energy capacity is effectively zero as long as the reservoir volume is large enough.”

This analysis suggests that GeoTES will be quite competitive with existing long-duration energy storage, including pumped hydro, by far the most common form of it. GeoTES can be used to provide an array of energy storage services including “load-shifting, arbitrage, grid reliability, energy capacity, and seasonal storage.”

     Systems heated with concentrated solar thermal (CST) utilize a parabolic trough collector (PTC) system where the mirrors concentrate the solar energy, and a piping system utilizes mineral oil as a working fluid in the heat exchange system.

Source: NREL

     Carnot batteries use heat pumps. There are a wide range of working fluids and configurations that have been explored, including different power cycles and thermal storage materials. Carnot batteries typically use thermal energy stored at the surface in tanks of water, molten salt, or fluidized particles. GeoTES Carnot batteries use the heat underground to insulate the fluid to keep it warmer longer. Heat can also be utilized for heating and cooling through exchangers. The cost for this heat as a levelized cost of heat (LCOH) is comparable to the price of natural gas. Working fluids such as supercritical CO2 (sCO2) can be among the most efficient.

Source: NREL

The results for a concentrated solar (CST) system and a Carnot Battery (CB) system are shown below:

Source: NREL

     NREL came up with a techno-economic model for GeoTES heated with concentrated solar or heat pumps utilizing depleted oil & gas reservoirs or suitable shallow reservoirs. Like in enhanced geothermal systems, these reservoirs may be naturally porous or fractured, may have been previously hydraulically fractured, or may be hydraulically fractured within the scope of the project. Adequate permeability, typically fracture permeability, is required to get the needed flow rates. Suitable geology also includes an adequate seal both above and below the reservoir to keep fluids contained, an aquifer that is reasonably confined. NREL notes that while upfront costs can be high for GeoTES, the levelized cost of storage is much lower than both molten salt and battery storage as the following graph shows.  

 

Source: NREL

NREL also shows a single well configuration with a hot well and a cold well. With this configuration, the hot well is the only well charging and discharging the system. The cold well keeps the reservoir adequately pressured and keeps the supply going to the heat exchanger.

Source: NREL

 

NREL also describes the main geological suitability requirements:

 

1)        Reservoir temperature – a minimum temperature of 91 deg C (195 deg F) is required for binary systems. Higher temps increase the efficiency of power cycles.  

2)        Reservoir pressure – a depleted reservoir is often a pressure-depleted reservoir, which means that the current reservoir pressure is much less than the original pressure. In this case the reservoir needs to be re-pressured so that it can reestablish geopressured where the reservoir pressure exceeds the hydrostatic pressure and the fluid will naturally flow to the surface when given a borehole.

3)        Porosity and permeability of the formation – as mentioned this is all about getting adequate flow rates in geothermal and GeoTES. These properties vary considerably among different rock formations and fluid reservoirs.

4)        Potential for scaling and clogging – this too is a risk for both geothermal and GeoTES wells. Properties like pH, formation water composition, mineralogy, temperature, pressure, injection rates, and presence of salt, affect scaling and clogging likelihood.

5)        “Permeability of caprock/seal: Low permeability seals/caprocks act as a barrier for heat and mass flow and also stops inflow and outflow of gasses such as methane, CO2, and sulphur oxides.”

6)        Presence of oil remaining – this can be beneficial as an enhanced oil recovery operation could be done simultaneously, adding to project revenue. I would expect more pilots to utilize a hybrid system like this.

7)        Formation depth – it costs more to drill deeper, but temperature increases with depth so both costs and benefits change with depth.

8)        Formation damage from geothermal extraction -  changes in reservoir temperature after prolonged production and injection can result in the plugging of clay particles, reducing permeability.

9)        Steeply dipping beds in the formation – this can cause updip movement of hot water away from production wells due to buoyancy.

 

 

References:

These abandoned oil wells near Bakersfield could store enough solar power for 300,000 homes. Adele Peters. Fast Company. June 7, 2024. These abandoned oil wells near Bakersfield could store enough solar power for 300,000 homes (msn.com)

Geological Thermal Energy Storage Using Solar Thermal and Carnot Batteries: Techno-Economic Analysis. Preprint. Joshua D. McTigue, Guangdong Zhu, Dayo Akindipe, and Daniel Wendt. NREL. 87000.pdf (nrel.gov)

Techno-Economic Analysis and Market Potential of Geological Thermal Energy Storage (GeoTES) Charged With Solar Thermal and Heat Pumps into Depleted Oil/Gas Reservoirs and Shallow Reservoirs: A Technology Overview. Preprint. Guangdong Zhu, Dayo Akindipe, Joshua McTigue, Erik Witter, Trevor Atkinson, Travis McLing, Ram Kumar, Pat Dobson, Mike Umbro, Jim Lederhos, and Derek Adams. NREL. September 2023. https://www.nrel.gov/docs/fy23osti/86609.pdf

Premier Resource Management LLC is working on a hybrid solar power and geothermal energy storage project in Antelope Hills in Kern, California. Carlo Cariaga. ThinkGeoEnergy.  June 5, 2023. Geothermal energy storage project proposed in Kern County, California (thinkgeoenergy.com)

AI/Machine Learning is the New Impending Energy Consumer: How Much Will It Affect Electricity Demand?

     Artificial Intelligence and machine learning are here to stay, whether we like it or not, simply because they are useful in many ways and more uses are still being found. Machine learning can find hidden trends in data quickly, pointing scientists, engineers, and others to important relationships and discoveries. AI and ML can diagnose diseases, decipher and translate ancient languages, optimize energy and mineral exploration, optimize manufacturing and industry, and much more. AI/ML can make art, animation, and precision devices. It can edit text and summarize articles. It has a 99.9% chance of destroying humanity says Sam Altman, CEO of OpenAI. I’ve got my doubts about that!  

 

 

A 2022 Study of Google’s Data Center Machine Learning Training Energy Use and Emissions

 

     Around 2020 or a bit before, researchers began estimating the energy use and carbon footprints of AI and ML in earnest. We already knew that the data centers that would be required were big energy users and emitters. A 2022 paper in Computer, ‘The Carbon Footprint of Machine Learning Training Will Plateau, Then Shrink’, researchers set out to model the carbon footprint of machine learning training and find ways to keep it minimized. Training the models, a necessary and key feature of them uses the most energy. The paper classifies ML emissions as follows:

 

● Operational, the energy cost of operating the ML hardware including data center overheads, or

● Lifecycle, which additionally includes the embedded carbon emitted during the manufacturing of all components involved, from chips to data center buildings.

 

The paper focuses on operational emissions. In particular, the study focuses on Google’s machine-learning data centers. In a larger scope. AI/ML emissions are a part of Information and Communications Technology (ICT) emissions, which I wrote about in 2016 here. The paper identified four best practices that contribute to lower emissions, which they call the four M’s: model, machine, mechanization, and maps, which they explain as follows:

 

1. Model. Selecting efficient ML model architectures while advancing ML quality, such as sparse models versus dense modes, can reduce computation by factors of ~5–10.

 

 2. Machine. Using processors optimized for ML training such as TPUs or recent GPUs (e.g., V100 or A100), versus general-purpose processors, can improve performance/Watt by factors of 2–5.

 

 3. Mechanization. Computing in the Cloud rather than on premise improves datacenter energy efficiency, reducing energy costs by a factor of 1.4–2.

 

 4. Map. Moreover, Cloud computing lets ML practitioners pick the location with the cleanest energy, further reducing the gross carbon footprint by factors of 5–105.

 

The following graph from the paper shows how Google was able to apply these best practices to reduce energy consumption by 83 times and CO2 emissions by 747 times!

 

 

     The standard metric of data center efficiency is the Power Usage Effectiveness (PUE) which is “a ratio that describes how efficiently a computer data center uses energy; specifically, how much energy is used by the computing equipment (in contrast to cooling and other overhead that supports the equipment).” PUE became a global standard in 2016. There are some issues with PUE that can complicate comparing the PUE of one facility to another, such as local climate and completeness of all energy sources (i.e. not omitting something like lighting). An ideal PUE is 1. The basic formula to determine PUE is as follows:

 

 

 

     “The average industry datacenter PUE in 2020 was 1.58 (58% overhead) while cloud providers have PUEs of ~1.10”

     “The average datacenter carbon emissions in 2020 was 0.429 tCO2e per MWh but the gross CO2e per MWh can be 5x lower in some Google datacenters.”

The number of processors running and the time they are running to perform the training tasks make up the bulk of energy use. This is calculated as follows:

MWh = Hours to train x Number of Processors x Average Power per Processor

MWh = Hours to train x Number of Processors x Average Power per Processor x PUE

tCO2e = MWh x tCO2e per MWh

The development of machine learning is ongoing and energy efficiencies are being improved in many areas. Thus, it is reasonable to assume that overall energy use per a given unit of work will continue to drop and efficiency will continue to increase. Big Tech companies that use the cloud and data centers like Google, Amazon, Microsoft, and Meta are the main players. The paper notes that global data center energy use only increased by 6% between 2010 and 2018, while the number of data centers and processors in them increased by a vastly larger amount. Predictions of a 70% increase over this period did not occur. Thus, these improvements are having an effect. These big data companies are also known for using renewable energy at their data centers as much as possible. Some of them use too much energy for the land around them to support enough solar panels.

     The paper also notes two other concerns about ML energy use and emissions: “the impact of Neural Architecture Search (NAS), which may run t.housands of training runs as part of a single search—potentially exploding overall energy consumption—and ML’s impact on client-side energy usage.” NAS uses computer power to search for and find models with higher quality or efficiency than humans can find. Client-side energy use involves the use of mobile phones that have ML accelerators built in for processes like bar code reading, OCR, face recognition, etc. However, this energy use is estimated to be about 5% of total phone energy use, if that. With billions of phones around the world this usage adds up, with global client-side usage from phones estimated at 0.4 TWh or less in 2021. The authors give several recommendations to reduce energy consumption and emissions. They note in their conclusion:

 

“Machine Learning (ML) workloads have rapidly grown in importance, raising legitimate concerns about their energy usage. Fortunately, the real-world energy usage trend of ML is fairly boring. While overall energy use at Google grows annually with greater usage, the percentage for ML has held steady for the past three years, representing <15% of total energy usage. Inference represents about ⅗ of total ML energy usage at Google, owing to the many billion-user services that use ML. GLaM, the largest natural language model trained in 2021, improved model quality yet produced 14x less CO2e than training the previous state-of-the art model from 2020 (GPT-3) and used only 0.004% of Google’s annual energy.”

 

“Furthermore, we illustrated that in large scale production ML deployments, minimizing emissions from training is not the ultimate goal. Instead, the combined emissions of training and serving need to be minimized. Approaches like neural architecture search increase emissions but lead to more efficient serving and a strong overall reduction of the carbon footprint of ML. Another perspective is that some consider the carbon footprint to be erased entirely if the cloud provider matches 100% of their energy consumption with renewable energy, as Google and Facebook have done and as Microsoft will soon do.”

 

     The location of data centers also matters for emissions. A 2022 paper in the Proceedings of the ACM Conference on Fairness, Accountability, and Transparency, showed this in the following graph which compared energy use for the same function in 16 different regions.

 

That paper also showed that the time of day of the functions also matters. This is likely due to availability to the grid of solar and wind at certain times of the day.  

     Not all AI processes are equally energy-intensive. Creating an image is far more energy-intensive than generating text. But AI does threaten to delay the aggressive climate goals of the Big Tech companies. Just the additional materials like concrete and steel to build the additional data centers required will add significantly to those carbon footprints.

 

AI/ML Can Also Lead to Better Energy Efficiency in Many Industries

     While AI/ML is set to use more power and put more CO2 in the atmosphere, it can also be used to find ways to increase energy efficiency. According to a 2023 article by Microsoft:

“The World Economic Forum underscores the role AI plays in the energy transition and estimates that every 1 percent additional efficiency in demand creates USD1.3 trillion in value between 2020 and 2050 due to reduced investment needs.”

It is no easy task to calculate and account for the energy use and avoided energy use provided by AI and ML. They also improve safety. The article gives several examples where Microsoft Azure data ML, and AI services are helping power generators, mining companies, telecommunications companies, and oil & gas companies optimize their processes for efficiency, improve inspection capabilities, and improve safety. Machine learning models like digital twins have resulted in many of these improvements.

 

The Coming AI Boom

     Remote severs processing away in data centers is the bulk of AI/ML energy use. The International Energy Agency reported that data centers and data transmission networks were responsible for 1% of energy-related GHG emissions in 2023. Data centers as a whole account for 1-1.5% of global electricity use. That usage is set to grow as AI booms. Elizabeth Kolbert, in her article in the New Yorker noted that U.S. data centers now account for about four percent of electricity consumption and is expected to climb to six percent by 2026. It is estimated that NVIDIA will ship 1.5 million AI server units per year by 2027, resulting in 85.4 TWh per year of power consumption. Data scientist Alex DeVries, a Ph.D. candidate in the Netherlands has developed ways to calculate and keep track of data center energy use. He did the same for cryptocurrency energy use, coming up with the Bitcoin Energy Consumption Index. He has also been tracking cryptocurrency water use and e-waste. I should point out here that AI/ML is a useful and net beneficial technology much more important to society than cryptocurrencies, which have many problems in addition to energy use. But both use up energy through the same method: processing power. DeVries points out that simply changing Google’s search engine to ChatGPT involves a massive increase in energy use.

     The two phases of AI/ML energy use are training and inference. Training is simply the initial training of the models. Inference refers to when the model goes live, is fed prompts, and gives responses. With Google the ratio was 60% inference and 40% training but that could vary among companies and task focuses. ChatGPT is powered by large language models that utilize huge datasets with billions of parameters. Cooling of the servers is expected to add 10-50% to total energy usage. It is also true that efficiency improvements enabled by AI/ML will offset a portion of energy use, but those numbers are not easy to predict. Of course, efficiency improvements can also increase demand for the service so that has to be accounted for as well. The 2022 paper above suggests that AI/ML energy use will grow, plateau, and then shrink, but they did not give a time frame. It is estimated that ChatGPT is already responding to about 200 million requests per day, the energy generated by about 17,000 households. DeVries noted that it took a long time to require cryptocurrency companies to disclose their energy use and he is disappointed AI/ML energy use has not developed disclosure requirements faster, considering that we already know they are needed.

 

What Will Produce the Electricity that Powers the AI Boom?

     It is true that the Big Data companies have been exemplary in using as much renewable energy as they can to power their data centers. However, there are limitations that they cannot overcome. Land availability is a big issue with solar and wind. Some could utilize or sell the waste heat that is generated in their operations to offset energy use elsewhere. Cryptocurrencies quested for and opted for cheap energy to mine the coins. That is less likely to be the case with AI/ML since Big Tech has already shown commitments to sustainable energy use. It has been estimated that a new data center is built every three days. It has also been acknowledged that renewable energy alone will not be enough to power the AI/ML boom. Once again, good old natural gas is emerging as a less than ideal, but pragmatic energy source. New data centers mean more electricity demand wherever they are built, and power authorities and grid operators need to plan for that. This new demand is resulting in some older coal and gas plants delaying their scheduled retirements in the name of maintaining grid reliability.  

     Governments are actively working to develop reporting standards for AI/ML impacts. The International Organization for Standardization is involved as a Yale Environment 360 article points out:

 

“Those will include standards for measuring energy efficiency, raw material use, transportation, and water consumption, as well as practices for reducing A.I. impacts throughout its life cycle, from the process of mining materials and making computer components to the electricity consumed by its calculations. The ISO wants to enable A.I. users to make informed decisions about their A.I. consumption.”

 

As DeVries suggested, this should have been done a few years ago, so we are a bit behind.

     While energy demand in the U.S. has been flat for a decade that is expected to change. A 20% increase is expected by 2030 due to ICT and AI demand, electrification, and EV demand. AI data centers alone are expected to add about 323 TWh by 2030. All data centers could represent up to 8% of U.S. energy demand by 2030. That is quite a lot. As mentioned, natural gas is emerging as a solution. While it is more carbon-intensive than renewables, it is both more dispatchable and cheaper. Estimates are that natural gas use could climb by 28% by 2030, adding about 10BCF/day of new natural gas demand. The U.S. as the largest producer of natural gas is well situated to provide it. The U.S. Southeast is expected to be a major AI data center hub. Pipelines can bring it from the south or from the north. Obama’s energy secretary Ernest Moniz noted that renewables will not be able to keep up:

 

“We’re not going to build 100 gigawatts of new renewables in a few years,” Moniz said.

 

Indications are that tech companies and natural gas companies are consulting with each other on these matters. The inability of tech companies to power their facilities with renewables due to things like land constraints has also led to them using more and more carbon offsets to meet their goals. The competition between tech companies to develop AI is leading them to move quickly as well.

 

 

What Can We Do to Mitigate These New Emissions Sources?

 

     As noted, the tech companies continue to work on process efficiencies for decreasing AI/ML energy consumption. The IEA lists eight recommendations going forward:

 

1)        Improve data collection and transparency

 

2)        Enact policies to encourage energy efficiency, demand response and clean energy procurement

 

3)        Support the utilisation of waste heat from data centres

 

4)        Collect and report energy use and other sustainability data

 

5)        Commit to efficiency and climate targets and implement measures to achieve them

 

6)        Increase the purchase and use of clean electricity and other clean energy technologies

 

7)        Invest in RD&D for efficient next-generation computing and communications technologies

 

8)        Reduce life cycle environmental impacts

 

References:

Power usage effectiveness. Wikipedia. Power usage effectiveness - Wikipedia

Natural Gas: A Natural Bridge to Fuel AI’s Electric Demand. Paul Hoffman. TipRanks. MSN. Money Markets. Natural Gas: A Natural Bridge to Fuel AI’s Electric Demand (msn.com)

AI is an energy hog. This is what it means for climate change. Casey Crownhart. MIT Technology Review. May 23, 2024. AI is an energy hog. This is what it means for climate change. | MIT Technology Review

The AI Boom Could Use a Shocking Amount of Electricity. Lauren Leffer. Scientific American. October 13, 2023. The AI Boom Could Use a Shocking Amount of Electricity | Scientific American

The growing energy footprint of artificial intelligence. Alex de Vries. Joule. Volume 7, Issue 10, 18 October 2023, Pages 2191-2194. The growing energy footprint of artificial intelligence - ScienceDirect

The Carbon Footprint of Machine Learning Training Will Plateau, Then Shrink. David Patterson, Joseph Gonzalez, Urs Hölzle, Quoc Le, Chen Liang, Lluis-Miquel Munguia, Daniel Rothchild, David So, Maud Texier, Jeff Dean. Computer. 2022. 2204.05149 (arxiv.org)

The Obscene Energy Demands of A.I. How can the world reach net zero if it keeps inventing new ways to consume energy? Elizabeth Kolbert, The New Yorker. March 9, 2024. The Obscene Energy Demands of A.I. | The New Yorker

As Use of A.I. Soars, So Does the Energy and Water It Requires. David Berreby. February 6, 2024. Yale Environment 360. As Use of A.I. Soars, So Does the Energy and Water It Requires - Yale E360

The era of AI: Transformative AI solutions powering the energy and resources industry. Darryl Willis, Corporate Vice President, Worldwide Energy and Resources Industry. Microsoft. September 28, 2023. The era of AI: Transformative AI solutions powering the energy and resources industry - Microsoft Industry Blogs

Data Centres and Data Transmission Networks. International Energy Agency. Data centres & networks - IEA

Measuring the Carbon Intensity of AI in Cloud Instances. Dodge, etal. FAccT '22: Proceedings of the 2022 ACM Conference on Fairness, Accountability, and Transparency. June 2022Pages 1877–1894. 3531146.3533234 (acm.org)

AI companies eye fossil fuels to meet booming energy demand. Mack DeGeurin. Popular Science. March 25, 2024. AI companies eye fossil fuels to meet booming energy demand | Popular Science (popsci.com)

 

Nitrogen Oxide (NOx) Emissions: Sources, Global Warming Effects, and Mitigation Strategies

   This is a repost of a March 2016 post on my previous blog. At the end I add some new information about N2O emissions from a new and quite detailed report by the Global Carbon Project. I also added some info about promising research to develop and propagate N2O-consuming bacteria.

   Nitrous oxide (N2O) made up about 5% of U.S. anthropogenic greenhouse gas emissions in 2013. This makes it the third most abundant greenhouse gas after CO2 and methane. The source of 74% of those emissions was “agriculture and soil management” according to the EPA. 5% of the emissions were sourced by “manure management.” That makes nearly 80% sourced from the agriculture sector. Industry, transportation, chemical production, and stationary combustion make up about 16%. The EPA also notes that N2O has an avg. staying time in the atmosphere of 114 years (compared to about 10 years for methane). This gives it 300 times the warming power in weight equivalence to CO2. About 40% of global N2O emissions are thought to derive from human activities. Since the Industrial Revolution nitrous oxide concentrations in the atmosphere have risen by about 15%. Variations in naturally emitted N2O were not addressed in the EPA report.

 

     N2O is the same gas used as a dental anesthetic (so-called laughing gas), an oxidation agent, and a food additive. Nitrous oxide is distinct from nitric oxide (NO) and nitrogen dioxide (NO2), but all three are produced during reactions from combustion. NO, NO2 and N2O react to form smog, acid rain, and tropospheric ozone, or ground-level ozone, none of which are desirable. These Nitrogen oxide emissions are usually referred together as NOx emissions.

 

     One issue I found annoying in the EPA report was the section on – Emissions and Trends – where they stated that there was an 8% increase in emissions since 1990 (from the graph it looked like it increased about 8% from 1990 to 1991). Technically this is true but emissions since 1991 have been close to flat overall. Presentation of data and statistics should avoid being misleading, if possible. Emissions of N2O are projected to rise 5% by 2020.

 

     The transportation sector makes up about two-thirds of non-agricultural N2O emissions. Stationary combustion from coal and gas power plants makes up a much smaller amount of N2O emissions as does biomass burning. An even smaller amount is released during nitrogen fertilizer manufacture. Domestic wastewater treatment is another minor source.

 

 

Mitigation Strategies

 

     Under-utilized nitrogen-based synthetic fertilizer is the biggest source of atmospheric N2O. Mitigation strategies such as organic farming could theoretically help but crop yields would be reduced and the use of manure-based fertilizer would increase, also increasing N2O emissions from manure management. More land use would also be required to make up for the decrease in crop yields resulting in reduced carbon sink potential. Better management and more efficient use of synthetic nitrogen-based fertilizer is perhaps a better mitigation strategy. This could also decrease fertilizer runoff which is a serious problem around the world as nitrogen and particularly phosphorous runoff into bodies of water is the main source of dangerous algae blooms, red tides, and de-oxygenated dead zones where rivers meet seas.

 

 

In Crop Farming

 

     Nitrogen (N) from fertilizer, whether synthetic or organic (typically manure) is often mobile. Synthetic fertilizer often has N in inorganic form which is more readily available to plants. Organic fertilizer contains organic N that converts to inorganic N over time. N can be lost as nitrate to groundwater or in gaseous form as nitrous oxide (N2O), dinitrogen (N2), or ammonia (NH4). It is typical that about half of the applied fertilizer is taken up by the crops for which it is destined. Soil microbes produce the N2O from the N during both aerobic nitrification and anaerobic de-nitrification. The anaerobic process is thought to make the most N2O. Thus one important mitigation strategy is simply to try to reduce the amount of waterlogged soils where anaerobic microbial functions can occur. Strategies to reduce N2O formation involve avoiding the formation of inorganic N by basically using the N by increasing the NUE, or N use efficiency. By tweaking the application rate, fertilizer formulation, timing of application, and placement, the N2O produced can be reduced. The rate of application depends on the crop as different crops take up fertilizer at different rates. Formulation can also depend on crops – whether to use anhydrous ammonia or urea ammonium nitrate. Additives can also reduce some N2O emissions by inhibiting nitrification.  The timing of application can be tweaked to when it is most readily taken up by the plants. Adding fertilizer in the fall or spreading manure on frozen fields can lead to big nitrate and N2O losses. Placement may involve concentrating the fertilizer near the plant roots where it is needed rather than spreading it across the fields. Carbon reduction credits as incentives are also a potential reward for targeting fertilizer to reduce N2O emissions.

 

 

In Automobiles

 

     In automobiles N2O emissions can be reduced by lowering the operating temperature of the engine through exhaust heat recirculation which employs the exhaust gas recirculation (EGR) valve to recirculate part of the hot exhaust gases to perform other functions, several of which can help power the hybrid batteries, keep the engine and fuel warm, help warm the interior, and improve gas mileage, all while reducing N2O emissions. This technology is used extensively in hybrid vehicles to help charge the lithium batteries.

 

     Simply increasing MPG in vehicles to reduce overall fuel consumption will decrease N2O emissions. Catalytic converters and other pollution control technologies can also reduce N2O emissions.

 

 

In Dairy Farming

 

Cows fed on grass release more urea in urine than in dung so mitigation strategies can involve helping cows to have more efficient digestion. Applying nitrification inhibitors as a spray to fields where cows pee can reduce nitrous oxide emissions from urine patches by 60-90%. The sprays also tend to increase nitrogen availability and the subsequent fertility of the soils. Avoiding grazing on wet soils can trigger less anaerobic N2O production. Better soil drainage, improved irrigation management, and effluent management (applying effluent dry rather than wet) are other strategies that can reduce N2O emissions.

 

 

In Industry

 

     In the power generation industry, one simple way to reduce N2O emissions is to switch fuels from coal to natural gas since natural gas produces far less when burned than coal. Natural gas power plants emit 7% of the nitrogen oxides (NO, NO2, N2O) emitted by coal plants so that is a pretty dramatic difference. In the manufacture of nitrogen fertilizer some fiber materials such as nylon, N2O is emitted in the production of nitric acid for fertilizers and adipic acid for making materials. EPA lists “technological upgrades” as a means to decrease emissions in these industries, which may involve capturing and reusing the gas.   

Global Carbon Project's New Report on N2O Emissions 1980-2020

    A new study confirms that nitrous oxide emissions have continued to grow from 1980 through 2020. The study by the Global Carbon Project concluded that 74% of nitrous oxide emissions came from agriculture in the 2010s. Chemical fertilizers and animal wastes on croplands were the main culprits. The study, "Global Nitrous Oxide Budget 2024," developed a global nitrous oxide budget model. Key aspects of the model are shown in the graphic below.

 

Australia's national science agency noted:

"The growth rates of atmospheric nitrous oxide in 2020 and 2021 were higher than any previous observed year and more than 30% higher than the average rate of increase in the previous decade."

Sources and sinks for N2O are shown below:

Other graphs show the increase through time:

Two graphs from the paper shown below indicate the anthropogenic contributions from different sectors and the contributions from four anthropogenic N2O sources over the time period: direct soil emissions, manure left on pasture, manure management, and aquaculture.

     N2O is also emitted naturally from the open ocean, continental margins, and terrestrial environments. The graph below shows the relationship between fertilizer application, manure additions to croplands, and atmospheric nitrogen.

Addendum: July 21, 2024

Norwegian Scientists Test N2O Consuming

Bacterium

     Researchers in Norway aim to tackle the problem of nitrous oxide emissions from fertilizers by adding naturally occurring soil bacteria that consume nitrous oxide. A field study shows quite promising results, reducing N2O emissions by 40-95%. The researchers studied different bacteria species, noting that some both consumed and produced N2O. They eventually found a bacterium that did not have a gene for producing N2O. The researchers developed a wheeled robot that can measure N2O levels in soil. Interestingly, the researchers developed a way to grow the bacteria in organic waste that would then be applied as a fertilizer and soil conditioner. This ensures that there will be enough of the bacterium in the soil to affect emissions. The results were much better than expected and the researchers are moving forward with larger field trials and exploring production of a super-fertilizer that contains these bacteria and other bacteria. They are also continuing to look for more bacteria species that do not produce N2O. Since fertilizer represents the vast majority of anthropogenic N2O emissions, these discoveries can have real-world impacts in reducing greenhouse gas emissions from the agricultural sector.  

References:

 

Overview of Greenhouse Gases: Nitrous Oxide Emissions – U.S. EPA (www3.epa.gov)

Global Mitigation of Non-CO2 Greenhouse Gases, 2010-2030 – U.S. EPA, EPA-430-R-13-011, September 2013

Mitigation of Non-CO2 Greenhouse Gases in the United States: 2010 to 2030 – U.S. EPA, EPA-430-S1-4-002, April 2014Management of Nitrogen Fertilizer to Reduce Nitrous Oxide (N2O) Emissions From Field Crops -  by Neville Millar, Julie E. Doll, and G. Phillip Robertson, Michigan State University Extension Bulletin E3152, November 2014

How Exhaust Heat Recovery and Recirculation Works – by Christopher Lampton – Auto/Hybrid Technology, at howstuffworks.com

Reducing Nitrous Oxide: Options for Reducing Nitrous Oxide Emissions from Dairy Farms, at dairyaustralia.com.au

Emissions of Greenhouse Gases in the U.S. – U.S. Energy Information Administration (EIA), March 31, 2011

What Are the Main Sources of Nitrous Oxide Emissions? -  from whatsyourimpact.org

Switch to Gas Slashed Power-Plant Emissions, Study Finds – article by Douglas Fischer, in the Daily Climate, Jan. 10, 2014

Study finds human-caused nitrous oxide emissions grew 40% from 1980–2020, greatly accelerating climate change. Science X staff. Phys.org.  June 12, 2024. Study finds human-caused nitrous oxide emissions grew 40% from 1980–2020, greatly accelerating climate change (msn.com)

Global Nitrous Oxide Budget 1980-2020. Earth System Science Data. October 9, 2023. Global_Nitrous_Oxide_Budget_1980-2020.pdf

Scientists make game-changing discovery while analyzing toxic byproduct in soil: 'It gives us hope'. Rick Kazmer. The Cool Down. July 20, 2024. Scientists make game-changing discovery while analyzing toxic byproduct in soil: 'It gives us hope' (msn.com)

New approach can reduce laughing gas emissions from agriculture by up to 95%. Tonje Lindrup Robertsen. Norwegian University of Life Sciences. May 3, 2024. New approach can reduce laughing gas emissions from agriculture by up to 95% | NMBU

The Texas Energy Fund: ERCOT Will Invest Heavily in Loans Mostly for Natural Gas Power Plant Projects to Help Provide Power for Expected Load Increases

    Texas power authority ERCOT forecasts a whopping 152 GW of new power load demand by 2030. In order to be able to provide that load in a dispatchable form, most of the load is expected to come from new natural gas power projects. In order to speed up meeting that anticipated demand, voters approved the creation of the $5 billion Texas Energy Fund (TEF) to invest in loans for plant developers. Utility Dive notes: “In April, ERCOT had active generation interconnection requests totaling 346 GW, with solar representing 155 GW of the queue, followed by 141 GW of battery storage. Gas made up just 15 GW.” The anti-natural gas Sierra Club thinks that the solar (much of which probably won’t be built per usual interconnection queue trends and snafus) plus storage could meet that demand. That is, of course, highly unlikely due to the need for baseload and dispatchable energy which intermittent resources cannot provide, and storage is too costly to ramp up enough to fill the void. Skimping on baseload and dispatchable supply would also make the already vulnerable ERCOT system even more vulnerable to extreme weather events. They do need to focus as well on weatherizing their natural gas power systems after the disastrous February 2021 blackout which led to 250 human deaths. It is still uncertain whether the system can weather another bout of extreme cold. Though unlikely to occur, they need to be thoroughly prepared so that such a tragedy never occurs again.

     ERCOT noted in April that the expected load increases are expected to come from artificial intelligence, data centers, industrial electrification including from oil and gas, and hydrogen and electric vehicles. ERCOT’s CEO Pablo Vegas was confident about ERCOT’s ability to accommodate new demand: “We have the ability in our economy to connect dispatchable resources faster than anyplace else in the country.” He cited updated regional transmission planning rules as a major boon for that accommodation. He noted that some renewables projects can get added to the grid faster if they accept curtailment, presumably until transmission and other upgrades are made to resolve congestion issues. This is part of ERCOT’s “connect and manage” strategy. ERCOT is also looking at distributed resources and for opportunities to optimize demand response and energy efficiency. It is noted, however, that 3-6 years is the general time frame for implementing transmission upgrades. That brings these changes up to the 2030 doorstep. ERCOT is looking into establishing “generation hubs” to best locate generation to meet expected load growth. They are also considering increasing their operating voltage level:

 

“We’re now undertaking a study of how a 765-kV system can support longer-term growth projections more efficiently and with lower cost over time than what our current 345-kV planning norms have been,” Vegas said.

 

Asked about the potential of dynamic line rating to enhance the grid and expand infrastructure, ERCOT noted:

 

“Dynamic line rating is something we’ve done here in Texas for years,” Vegas said. “Many of the large transmission service providers and a good majority of the transmission lines in the state of Texas today are dynamically line rated, and we get the benefit of that throughout the year.”

 

     Utility Dive notes that developers, including ENGIE and Vistra, representing more than 41 GW of mostly gas-fired generation projects, have indicated that they will apply for low-interest loans via the TEF. The response to the loans has been very good.

     ENGIE has plans to build a 483 MW gas-fired peaking plant that is expected to switch over to hydrogen when the switch becomes viable in the future. Vistra proposed peaking plants in West Texas to help support Permian Basin oil and gas activity. They also plan to convert a coal plant to gas and to add additional summer and winter gas capacity at other gas plants. Vistra noted that their plans depend on market certainty. Texas regulators have been implementing market reforms to encourage long-term investment. They have been improving grid reliability, including ancillary services. They continue to work on a performance credit mechanism (PCM) to help power generators. They also continue to work on a reliability standard to ensure grid reliability during extreme weather events. This should be prioritized to prevent another tragedy.

 

 

References:

 

Vistra, ENGIE and other developers move to tap $5B Texas Energy Fund for 41 GW of gas-fired projects. Robert Walton. Utility Dive. June 3, 2024. Vistra, ENGIE and other developers move to tap $5B Texas Energy Fund for 41 GW of gas-fired projects | Utility Dive

ERCOT launches new planning efforts as 2030 load growth projections soar 40 GW in a year. Robert Walton. Utility Dive. April 24, 2024. ERCOT launches new planning efforts as 2030 load growth projections soar 40 GW in a year | Utility Dive

Highlights from the Energy Information Administration’s ‘Financial Review of the Global Oil and Natural Gas Industry: 2023’

 

     This report analyzes 175 upstream oil and gas companies and 42 downstream companies, or refiners. The companies in the study were mostly headquartered in the U.S. and Canada. The study utilized publicly available financial statements for the analysis. The key findings given are as follows:

• Brent crude oil daily prices averaged $82.18 per barrel in 2023—17% lower than in 2022.

 • Among the upstream companies, combined petroleum liquids production increased 4% in

2023 from 2022, and natural gas production decreased less than 1%.

 • Cash from operations decreased to $678 billion in 2023—21% lower in real terms than in

2022.

 • Exploration and development spending was 27% higher in real terms in 2023 from 2022.

 • The energy companies reduced net debt by $38 billion and allocated $89 billion to net share

repurchases in 2023.

 • Refiners’ earnings per barrel processed decreased on average in all regions in 2023.

 • Capacity utilization among refiners in Asia Pacific increased substantially in 2023.

 

Upstream Review

 

Several of the comparisons from 2022 to 2023 are a reflection of lower oil prices and especially lower natural gas prices. Companies seem to have done pretty well, especially with the cash generated from high commodities prices in 2022. E&P spending was up significantly in 2023. Going forward, this should keep oil production steady or moving upward. However, some of that spending is a reflection of higher reserve acquisition costs in 2023. As one graph indicates, these costs rose by 43% year over year while E&P spending increased only by 23%, so perhaps E&P activity is lower than it could be since the report notes that reserve acquisition costs made up a higher share of overall upstream costs. Finding and lifting costs also increased a bit. Natural gas production is dependent on prices, including regional prices where pipelines are constrained, as some regional players in Appalachia have shut-in production. The substantial increase in capacity utilization among refiners in Asia Pacific seems likely to be a reflection of Indian and Chinese refiners taking in more cheap Russian oil to refine. Yes, Modi, you too are helping to make Russian sanctions less effective.  

     I am not sure I fully understand the following graph, but the companies added a net of 2 billion barrels of oil equivalent (BOE) in oil and gas proved reserves for the year.

    

 

     The following graph is an interesting one. It shows that Canada had the biggest proved reserves increase of the year. It also shows proved reserves by region/country since 2014. It should be noted that this study may not be a true reflection of all companies combined but I am assuming the EIA considers it to be a reasonable approximation. This graph shows us where our oil and gas are coming from and where it will be coming from in the future. While many people still think oil is all coming from the Middle East, it really isn’t. The U.S. and Canada had combined proved reserves of about 120 billion BOE in 2023, while the Middle East had about 10 billion BOE.  

     Below is a graph of annual proved reserve additions since 2014. The ‘extensions and discoveries’ section shows that 2023 was a slightly lower-than-normal year for those exploration successes.

 

     Another of several interesting graphs from 2014 through 2023 shows cash from operations alongside capital expenditure. This shows that companies are still riding a cash-from-operations wave that began in 2021, although it dropped a bit in 2023 relative to high levels in 2022.  Another graph shows that 2021, 2022, and 2023 were the best years of the past decade for reducing debt by a pretty wide margin. The second graph depicting return on equity shows the same wave that began in 2021. These graphs suggest that most of these companies should be in decent financial health. This graph and the one below it also depicts comparisons of these energy companies to overall U.S. manufacturing. The long-term debt-to-equity ratio confirms that these upstream oil & gas companies with a ratio of about 41% are in better financial shape than all U.S. manufacturing combined which is at a ratio of about 52%.

 

 

Downstream Review

     The graph below shows that global refining capacity continues a very slight decreasing trend since it peaked in 2018-2019. We are now at a similar global refining capacity as in 2014.

     The global utilization of refining capacity increased, led by the Asia Pacific increase and by increases in Latin America. Utilization of refining capacity decreased a little in the U.S., Canada, and Europe.

 

References:

Financial Review of the Global Oil and Natural Gas Industry: 2023. Petroleum and Liquid Fuels Markets Team. Energy Information Administration. May 2024. 2023 Financial Review.pdf (eia.gov)