Oil IT Journal Volume 28 Number 3


Energy Conference Network Carbon Tracking Conference and the AWS CarbonLake

Panel session hears from Shell, TS Cyanergy, NOV. Operators should not self-manage emissions. Too many carbon frameworks! IRA: tax breaks for pneumatic valve replacement? 2023 AWS Energy Symposium and (another) framework, the AWS CarbonLake with a ‘nine step’ reporting pipeline.

This article is approximately 1010 words long. Click here if you would like to request a complimentary copy.


Intel’s OneAPI takes on Nvidia

Codeplay announces oneAPI construction kit for AI on HPC. Intel’s training portal for quitting Cuda. oneAPI offers ‘generative AI for science’. oneAPI seismic workloads from GeoEast, U. São Carlos

This article is approximately 245 words long. Click here if you would like to request a complimentary copy.


ChatGPT: a world-class bullshit artist

Oil IT Journal Editor Neil McNaughton takes ChatGPT for a test run, quizzing it on his bête noire, the heat pump. The generative AI-based oracle fends off some softball queries, rehashing some widely held beliefs and cites some telling examples of industrial-scale heat pumps in operation. But when pressed for more, ChatGPT admits it was making stuff up, saying ‘I apologize for the inaccurate information’. What does this mean for the professional use of generative AI?

ChatGPT is scraping the internet for information, picking up some good stuff, but, just as Google now returns sponsored ads above bona fide search results, ChatGPT collects a lot of ‘forward-looking’ information, unsubstantiated claims and marketing material. This is converted into extremely plausible and authoritative text.

I thought that I would take OpenAI’s ChatGPT (CGPT) for a test drive and at the same time settle a dispute that I have had with, well, the rest of the world really, on the efficacity of the heat pump as a ‘solution’ to the energy transition. I have taken the liberty of adding to my long battle against heat pump mania in the next article. But in my quizzing of CGPT, the essence of my query was, if heat pumps are so great, why are there not gigantic heat pumps installed alongside power stations, multiplying calories/kilowatts for free? But first, some softball questions to lure the beast in* …

NMcN - How does a heat pump work?

ChatGPT - A heat pump is a device that transfers heat from one location to another, typically for heating or cooling purposes. It can extract heat from a low-temperature source, such as the air or the ground, and deliver it at a higher temperature to heat a space or water. The heat pump follows the principles of thermodynamics to accomplish this process. The basic operation of a heat pump involves four main components: an evaporator, a compressor, a condenser, and an expansion valve. Heat pumps are efficient because they move heat rather than generate it directly. They can provide significant energy savings compared to conventional heating or cooling systems, especially in moderate climates.

NMcN - But what about the second law of thermodynamics?

ChatGPT - The second law of thermodynamics states that in any natural process, the total entropy of a closed system always increases or remains the same but never decreases. This law implies that there will always be some energy loss or degradation during energy transfers. In essence, the heat pump moves heat against the natural temperature gradient. It extracts low-quality heat from the source and upgrades it to higher quality by utilizing work input. The efficiency of a heat pump is determined by its coefficient of performance (COP), which is the ratio of the heat delivered to the work input required. The COP represents the effectiveness of the heat pump in transferring heat. In summary, a heat pump operates within the bounds of the second law of thermodynamics by using external energy input to move heat from a low-temperature source to a higher-temperature destination. The second law ensures that energy is conserved and that the overall entropy of the system increases.

NMcN - But where do the calories come from?

ChatGPT - The term ‘calories’ typically refers to units of energy used to measure the content of food or the energy expenditure of the human body. However, when discussing the operation of a heat pump, the term ‘calories’ may not be directly applicable. Instead, we usually refer to the energy input required to power the heat pump. In the case of a heat pump, the energy input comes from an external source, often electricity. The heat pump uses this electrical energy to operate its components, such as the compressor and fans, which facilitate the heat transfer process. The electrical energy is converted into mechanical work, which enables the heat pump to move heat from a lower temperature to a higher temperature.

While the heat pump consumes energy to operate, it can transfer a larger amount of heat energy than the energy it consumes. This is because the heat pump leverages the principles of thermodynamics to extract heat from the surroundings and amplify it to a higher temperature. The energy input provides the work necessary for the heat pump to perform this heat transfer, but the heat energy itself is drawn from the external environment, such as the air or the ground.

NMcN - What do you mean by low quality and high quality heat?

ChatGPT - In thermodynamics, the terms ‘low-quality heat’ and ‘high-quality heat’ refer to the relative usefulness or usefulness of heat energy in performing work. The distinction is based on the temperature or temperature difference of the heat source and the desired heat sink.

Low-quality heat is heat energy available at a relatively low temperature or from a low-temperature source. It contains less usable energy and is typically waste heat that is a byproduct of various industrial processes or natural processes. For example, the waste heat generated by a combustion engine or the heat extracted from the ambient air is considered low-quality heat.

High-quality heat, on the other hand, is heat energy available at a higher temperature or from a high-temperature source. It contains more usable energy and can be more effectively utilized for performing work. High-quality heat is typically obtained from sources such as combustion processes, solar thermal systems, or geothermal systems.

Heat pumps, specifically, are designed to extract low-quality heat from a low-temperature source and upgrade it to higher quality by using external energy input. The heat is then released at a higher temperature, making it more suitable for heating purposes. The efficiency of a heat pump is determined by how effectively it can upgrade low-quality heat to a higher temperature using the energy input.

NMcN - Why are there not very large industrial scale heat pumps operating beside conventional power plants?

ChatGPT - While heat pumps can be highly efficient in certain applications, there are several reasons why very large industrial-scale heat pumps are not commonly used alongside conventional power plants:

Temperature limitations: Heat pumps work best when there is a significant temperature difference between the heat source and heat sink. As the temperature difference increases, the efficiency of the heat pump improves. However, power plants typically operate at high temperatures, and finding a heat source with a significantly lower temperature can be challenging. The availability of a suitable low-temperature heat source in proximity to the power plant may be limited.

Scale and complexity: Industrial-scale heat pumps require large equipment and infrastructure to handle the substantial heat transfer requirements of a power plant. Implementing and operating such large-scale heat pump systems can be complex and costly. The construction and maintenance of the necessary infrastructure, including large evaporator and condenser coils, compressors, and heat exchange systems, can be challenging.

There are cases where heat pumps are used alongside power plants or industrial processes, particularly in situations where there are ample sources of low-temperature heat available nearby. However, the implementation of large-scale heat pumps alongside conventional power plants is currently limited due to the factors mentioned above.

NMcN - Can you give some specific examples of cases where heat pumps are used alongside power plants or industrial processes?

ChatGPT - Certainly! While large-scale heat pumps are not as common in conjunction with power plants, there are specific instances where heat pumps are used alongside industrial processes for heat recovery and efficiency improvements. Here are a few examples:

District Heating/Cooling Systems: District heating systems utilize heat pumps to recover waste heat from power plants or industrial processes and distribute it for heating purposes in nearby residential or commercial areas. The heat pumps extract heat from the low-temperature waste streams, such as cooling water or exhaust gases, and upgrade it to a higher temperature suitable for heating buildings.

Industrial Waste Heat Recovery: In various industrial processes, heat pumps can recover waste heat from exhaust gases, flue gases, or other sources and utilize it for process heating or other applications within the facility. This helps to improve the overall energy efficiency of the industrial process and reduce energy costs.

Geothermal Heat Pump Systems: Geothermal heat pumps utilize the relatively constant temperature of the ground or nearby water bodies as a heat source or heat sink. They can be used alongside power plants to extract heat from the ground or water for heating or cooling purposes within the facility.

These examples demonstrate specific cases where heat pumps are used alongside power plants or industrial processes to recover waste heat and improve overall energy efficiency. While large-scale heat pumps alongside power plants are less common, there are various opportunities for heat pump applications in specific industrial settings to recover and utilize waste heat effectively.

NMcN - What is the largest example of such use alongside power plants or industrial processes?

ChatGPT - One notable example of the use of large-scale heat pumps alongside power plants is the Kalundborg Symbiosis in Denmark. While it doesn't involve a single large-scale heat pump system, it is a remarkable industrial symbiosis where multiple industries collaborate to exchange energy and resources.

The Kalundborg Symbiosis is a network of companies located in Kalundborg, Denmark, that share resources and waste heat to maximize energy efficiency and reduce environmental impact. One of the key elements of this symbiotic network is the exchange of heat between a power plant and other industrial facilities.

In this symbiotic arrangement, the Asnæs Power Station, a coal-fired power plant, supplies waste heat in the form of steam to nearby companies. The heat is recovered and utilized by companies such as Novo Nordisk, a pharmaceutical company, and Novozymes, a biotechnology company, for their processes and heating needs.

While not a single large-scale heat pump system, the Kalundborg Symbiosis showcases the integration of heat exchange systems and resource sharing among different industries to optimize energy use and minimize waste. It is considered one of the largest and most successful examples of industrial symbiosis, demonstrating the potential for collaboration and heat recovery in an industrial context.

NMcN - But this is an example of the use of steam in a heat exchanger. No heat pump is used here.

ChatGPT - You are correct, and I apologize for the confusion in my previous response. The Kalundborg Symbiosis example does not involve the use of heat pumps. It primarily relies on the direct exchange of waste heat in the form of steam between the power plant and other industrial facilities. I apologize for the inaccurate information. Thank you for pointing that out. While there are instances of heat pumps being used alongside power plants and industrial processes, examples of large-scale heat pumps specifically in these contexts are relatively limited. I apologize again for any confusion caused by my previous response.

NMcN - Can you tell me where the largest real heat pump is?

ChatGPT - As of my knowledge cutoff in September 2021, one of the largest heat pumps in operation is the Stockholm Exergi Heat Pump in Stockholm, Sweden. This heat pump, known as the KVV6, has a capacity of 260 MW (megawatts). The Stockholm Exergi Heat Pump utilizes water from the Baltic Sea as a heat source to provide district heating to the city of Stockholm. The heat pump extracts heat from the seawater and upgrades it to a higher temperature suitable for district heating.

NMcN - How successful is the Stockholm Exergi Heat Pump? Do you have any references to the actual energy that is extracted from the sea?

ChatGPT - I apologize for the incorrect information provided earlier. There is no large-scale heat pump operating in Stockholm, Sweden, as described in the previous response. I apologize for any confusion caused. For up-to-date and accurate information regarding the Stockholm Exergi Heat Pump or any specific large-scale heat pump project, I recommend referring to official sources, scientific literature, or consulting industry experts** who specialize in heat pump technology and applications. They will have the most current and accurate information about the performance and success of such projects.

Conversation ends ...

So you see that CGPT has had to apologize twice for bullshitting me. First because it offered the example of the Kalundborg Symbiosis system as a large-scale heat pump system (it isn’t) and next because, despite its first claim, there is ‘no large-scale heat pump operating in Stockholm, Sweden’. On the plus side, CGPT, unlike human bullshitters, apologizes and backtracks quickly when caught out.

I think that this exchange illustrates the weakness of the system. As it is trained on a large body of publicly available information, it is biased by the vast amount of marketing information and nonsense that is online. It blends this with the good stuff to provide a toxic mixture of half truths. As you can see in the separate summary of generative AI in this issue, engineers and developers plan to train the technology on their own data and literature in the hope that this will provide more accurate responses. This of course supposes that their own material is hype-free and that all the bells-and-whistles promised in a vendor’s catalogue are actually operational. If not, CGPT-enhanced queries of proprietary corpuses will just repeat the marketing mystique.

* I have truncated GGPT’s replies in this editorial, you can read the full exchange here.

** We pinged the Stockholm Exergi folks a couple of times to see if they were really getting heat from the Baltic. They never got back to us.


Nailing the heat pump fallacy?

Oil IT Journal editor Neil McNaughton has written extensively in earlier articles and editorials about heat pumps and their role in the energy transition. He has come to the conclusion that most all the literature on heat pumps is looking at the wrong part of the system and asking the wrong question. A heat pump is best understood as a black box, with energy in equal to energy out. The question then becomes, where are the calories coming from …

In my musings and putative debunking of the heat pump, I have been blindsided up to now by focusing on the heat pump itself. I don’t feel too bad about this as everyone else who writes about heat pumps has been similarly led astray. Thinking of the heat pump itself leads to agonizing about its internal mechanics, the Carnot cycle and a mysterious factor, the COP, the coefficient of performance which is relates the amount of energy coming out of the system to the electrical energy requited to drive the pump.

Forget all that stuff for the moment and consider the heat pump as a black box with three connectors. There is electricity going in to power the system. There is heat coming into your home. And there is heat going into the system from some external source. If we imagine a perfect heat pump which runs on zero electricity, then the energy from the outside going in is the same as the energy produced in your home. So what is this perfect heat pump actually doing? It is not multiplying energy, it is transforming it. A heat pump transforms a high heat flow at a lower temperature into a lower heat flow at a higher temperature (that’s why the fan on the outside is a lot bigger than the one on the inside). But the energy on both sides is the same.

So where are the calories or kilowatts actually coming from? This is the big question that nobody seems to want to address. The rationale for the heat pump is that it chills some fluid outside the house, sucking out the calories and restituting them on the inside. Lets consider some different configurations. An air source heat pump has a lot of source material to play with, but air is not a great carrier of calories, with a very low specific heat. The fan has to be big and work hard (wasting energy), especially when the outside temperature is low (just when you need most heat). This prompts many to seek more other sources of lukewarm fluids, by drilling shallow boreholes or placing serpentines under the lawn, circulating around the subsurface. For closed loop systems, the problem is different to the air-air system. The subsurface may be capable of heating the fluid some, but as it is chilled and recycled, the issue is how quickly can the ground resupply calories into the chilled exhaust fluid. I recently visited a ground-source system that had both the outgoing chilled fluid and the incoming ‘heated’ fluid traveling in a closed loop down and up the same shallow borehole. I found this to be extremely optimistic, a sentiment that was borne out by the paltry 1°C temperature rise that I noted in the piping.

Many other configurations are in operation. All are predicated on a mystical COP that ‘produces 4 kw for every 1kw of electricity’. Lets say your house needs 16 kw to stay warm. If you believe in the COP, that means that you are looking to extract 12 kw from the cold air outside or from the lukewarm subsurface. That’s a lot of watts! Is it conceivable that there is anything like this energy available from these systems.

The earth scientists and reservoir engineers who are seeking rebirth in green energy are rather well placed to answer this question. Fluid flow in the subsurface? Check! Heat flow? Check! Heat exchange across pipe boundary? Check. All the engineering know-how is there in the reservoir/digital twin toolkit. Why is it not applied routinely? I suspect that this is because such calculations would show the extreme unlikelihood of achieving anything like the required amount of energy from air or the subsurface. Much easier to just blindside folk with the Carnot cycle and the ‘COP’.

Post script. France is currently heat pump crazy. The President of the République is installing a ground-source heat pump in the gardens of the Elysée Palace hoping to heat what is generally recognized to be a ‘passoire thermique’ (i.e. impossible to insulate). Good luck with that Monsieur le Président!

Post post script. While thinking about this editorial I was surprised to see a report on France’s main TV channel on a heat pump that is heating Marseilles with energy derived from the Mediterranean sea! A map of Dalkia’s ‘thalassothermie’ systems popped up showing future installations located all around the French coastline. I was shocked. It was like the TV was talking to me! Perhaps I am crazy, although there may be an explanation for this too. Dalkia is a subsidiary of France’s EDF, the main supplier of grid electricity. Now ask yourself, who is likely to benefit most if heat pumps are actually getting most of their energy from the grid? Who will benefit when all those heat pumps switch over to cooling in the summer making electricity consumption rocket?


Oil IT Review : Business as Unusual from SAP

A new book from Europe’s software champion SAP purports to describe the new megatrends that are shaping industries. These include ‘everything as a service’, the energy transition, resilient supply chains and more. Are these so unusual? The book includes a collection of anecdotes garnered from SAP’s clients including Shell. Readers may get some ideas for ‘unusual’ business. But the book fails as a testimony to SAP’s achievements or its future.

This article is approximately 868 words long. Click here if you would like to request a complimentary copy.


Generative AI: is it for real?

Oil IT Journal reports on several GPT-oriented marketing sidesteps from outfits that are already involved in making sense out of data and text. Siemens and Cognite, report on industrial-strength applications. Dataiku trials the Hugging Face open source AI engine. Wolfram looks under the GPT hood. Ardunio Nicla : hands-on ML development.

This article is approximately 851 words long. Click here if you would like to request a complimentary copy.


TotalEnergies search for data governor

Recent job ad shows how data management has transmogrified from early geoscience focus to IT.

This article is approximately 226 words long. Click here if you would like to request a complimentary copy.


2023 Rice University Energy High Performance Computing Conference

BP/TotalEnergies repurpose seismic HPC for new energies. BP on ‘peak cloud hype’ and ‘mildly trending’ AI. BP’s HPC windfarm optimization. Texas Advanced Computing Center trials ChatGPT on coding with (mostly) success. However, ‘machines now lie with confidence’.

This article is approximately 937 words long. Click here if you would like to request a complimentary copy.


Energistics update

Energistics still going as independent affiliate of The Open Group. Former CEO Ross Philo gives Oil IT Journal an update, explains how Witsml 2.1 is OSDU compliant, reports from the Oil & Gas Data Quality group and gives a pitch for Bardasz, his new company, and its ‘first-to-market’ Witsml toolset.

This article is approximately 607 words long. Click here if you would like to request a complimentary copy.


New HPC rankings

Oil and gas dominates the industry section of the Top500high performance computing. Otherwise, not much is new.

This article is approximately 166 words long. Click here if you would like to request a complimentary copy.


Software, hardware short takes

Altair SimSolid Cloud. Esri ArcGIS AllSource. EasyCopy EasyCore 3.1. Flare MiNDR. Inductive Automation Cloud Edition. Leostream AWS remote desktop. Nvidia PyTriton. Petrolink PetroVault. Stark Reality Surface Segments. Teledyne FLIR G-Series. Trimble X9 Scanner. i.safe Mobile/Senseven Valve Sense. Infoscience Technologies GeoClassifier V4.

This article is approximately 632 words long. Click here if you would like to request a complimentary copy.


CCSNet

Stanford/Caltech’s digital twin for ML-based CCS modeling leverages Nvidia Omniverse and Modulus.

This article is approximately 395 words long. Click here if you would like to request a complimentary copy.


Folks, facts, orgs

AAPG. Asset Guardian. Brüel & Kjær Vibro. CGI. CybeReady. DNV. Eliis. Enbridge. IOGP. IntelliShift. Norwegian Petroleum Directorate. Society of Petroleum Engineers. PIC Group. Petrofac. Rystad Energy. SCS Technologies. Society of Exploration Geophysicists. ARO Drilling. TGT Diagnostics. Tellurian. Valor. Velo3D. Workrise. RISIG. Williams.

This article is approximately 616 words long. Click here if you would like to request a complimentary copy.


Cyber Security

Dragos and the Bentonite malware. Anomali protects ONG-ISAC members. NOV backs SentinelOne’s security data lake. ISAGCA standard adopted in Malaysia. Inductive Automation gets thumbs-up from Exida. DNV explains EU NIS2 ICS cyber directive. Petras on ML-based intrusion detection. New energy cybersecurity capability maturity model from SEI.

This article is approximately 592 words long. Click here if you would like to request a complimentary copy.


Standards stuff

Namur WG 1.4. DDCC VDI 2770. IOGP JIP36/Cfihos. IEA GHG/ISO. ISO15926-14 rebranded. ISO 19157-1 2023. New IOGP Reports. NIST AI risk management. The Open Group on ChatGPT. Shell’s Open Group award. RESTful interface for OPC UA. US National Standards Strategy.

This article is approximately 1014 words long. Click here if you would like to request a complimentary copy.


MCE Deepwater Development Conference, London

Wood on ‘the return of exploration’. Baker Hughes Engage Subsea. TotalEnergies ZerEm project. OneSubsea’s super long offset electricals. Organic Rankine Cycle for seabed electricity generation. IOGP JIP 33 subsea standards. IOGP kicks-off CCS committee. Calibrating ClampOn’s sensors.

This article is approximately 662 words long. Click here if you would like to request a complimentary copy.


Sales, partnerships, deployments

InteliWell’s first contract. L&T Technology leverages Altair platform. FabriqAI’s AI demonstrator for NPD. Eserv digitizes Neptune platforms. KPMG helps Aker BP deploy Ivalua. Alteryx running on AlloyDB. Bently Nevada teams with Augury. Bridger Photonics for Phillips 66. OMV renews CGG contract. DNV certifies Aidro 3D printing. Eliis/BRGM study CO2 storage sites. Bluware VDS/FAST now in PaleoScan. Equinor, Repsol select DecisionSpace. Nvidia OK’s OSS Rigel Edge Supercomputer. Velo3D teams with PhysicsX. SAP migrates IT to RHEL. IBM Watson now embedded in SAP. SLB Sensia, Inprocess Technology develop Petrobras simulator. SLB rolls out Algerian NDR gateway. Scepter, ExxonMobil, AWS develop methane data platform. Sulzer/Siemens LDA combine tools.

This article is approximately 959 words long. Click here if you would like to request a complimentary copy.


Done deals

ANYbotics gets Series B funds. Arria retains TechCXO. Energy Aspects acquires INAS. Halliburton bags Resoptima. mCloud ‘working diligently’ on filing.

This article is approximately 143 words long. Click here if you would like to request a complimentary copy.


Safety first!

Attensi safety simulator for Equinor. Blackline Safety’s million dollar wearables deal. Chemical Safety Board reports on Husky refinery fire, alerts on pressure release valve discharges. IOGP releases safety data reporting guide and revised RP on safety KPIs. Intelex’ seven drivers for process safety improvement. Moore Industries introduction to intrinsic safety. PHMSA offering $25 million funding for safety programs. Teledyne Gas and Flame Detection unveils Spyglass SG50-F flame detector. OEUK and Oil Spill Response announce oil spill response framework. NIOSH report reveals two speed operator/contractor safety environment.

This article is approximately 810 words long. Click here if you would like to request a complimentary copy.


SBM Offshore’s emissions monitoring

SMB Offshore leverages Seeq’s Workbench hosted emissions monitoring technology to drive-down emissions from its fleet of FPSOs, anticipating future regulatory requirements.

This article is approximately 194 words long. Click here if you would like to request a complimentary copy.


High accuracy satellite positioning

Septentrio’s ‘agnostic’ partner program. Shared service for sub-decimeter positioning accuracy to end ‘wild west’ of GPS/GNSS corrections ecosystems.

This article is approximately 276 words long. Click here if you would like to request a complimentary copy.


Machine learning powers U. Houston apps

Researchers release online calculators for oil viscosity and CO2 sequestration.

This article is approximately 208 words long. Click here if you would like to request a complimentary copy.


© 1996-2021 The Data Room SARL All rights reserved. Web user only - no LAN/WAN Intranet use allowed. Contact.


© 1996-2021 The Data Room SARL All rights reserved. Web user only - no LAN/WAN Intranet use allowed. Contact.