Apartment Complex Electrification in a Vault With a Chef

Greetings from the "Vault" in Fishwick, Canberra, for the "Apartment Complex Electrification - Information Session". The vault is a windowless concrete structure, on a secured industrial site, looming like something from an action movie. Inside the Vault is an industrial kitchen, complete with a chef. As I walked in someone said "High Tom good to hear you on radio". 

This all distracted me from the public servants presenting on the Apartment Complex Electrification project. Some time ago I was on a working group advising the Chief Minister how to reduce energy use in the ACT. We focused on apartments as there were already programs for houses. But it turns out this is difficult for technical, legal and social reasons, thus the pilot program. 

After a few minutes it was explained the chef is demonstrating cooking on a low energy induction cooktop. The food was excellent.

The call for applications from apartment buildings to participate in the pilot will be released Monday.

E-waste not a major challenge with generativeAI

Wang, Chen, Zhang, Tzachor (2024) suggest that the demand for generative AI will create an e-waste problem of 1.2 to 5.0 million tons  for 2020 to 2030. They also suggest this could be made worse by political restriction on access to more advanced  efficient chips and rapid replacement of old hardware. I suggest the problem is not that large and energy use will remain a larger problem than e-waste.

One aspect the authors do not mention is the lack of price signals between server providers and the end user with current generative AI services. This is likely to be self correcting. Currently demand for generative AI is being generated by offering of free services to the public. As the user is not paying for the service and there is therefore no built in fee for responsible disposal of created e-waste, there may be a later problem. Some speculative AI ventures are likely to become bankrupt leaving a toxic legacy (similar to the mountains of scraped e-bikes left by failed startups). However, as users come to rely on Generative AI services, vendors will introduce charges, which can cover e-waste costs.

Currently generative AI server farms are using generic Graphic Processor Unit chips. These are the same chips used for cryptocurrency server farms. The environmental issues are similar with both. There have been articles about AI consuming as much power as small countries (just as there were for crypto). However, there is more of a mainstream use for AI, which will allow for better long term regulation of environmental effects. With its abundance of renewable energy sources and a stable regulatory environment, Australia could provide a popular location for AI centers. This would allow a small query to be sent across the world and answer sent back, effectively embedding the renewable energy in the answer. 

There may also be scope for reuse of older, slower, more energy using AI chips in locations with abundant renewable energy. As more efficient chips were installed close to the user in high energy cost countries, the old hips would be installed further away. Rather than store energy in batteries to run these chips, it may be cheaper to shut them down when the sun isn't shining & the wind isn't blowing. How to do this is something engineers and computer professionals can learn to optimize with specialist training (Worthington, 2012). 

Reference

Wang, P., Zhang, LY., Tzachor, A. et al. E-waste challenges of generative artificial intelligence. Nat Comput Sci (2024). https://doi.org/10.1038/s43588-024-00712-6

Worthington, T. (2012, July). A Green computing professional education course online: Designing and delivering a course in ICT sustainability using Internet and eBooks. In 2012 7th International Conference on Computer Science & Education (ICCSE) (pp. 263-266). IEEE. https://doi.org/10.1109/ICCSE.2012.6295070

What country does Generative AI use as much power and water as?

Generative AI is the new bitcoin. Pundits used to compete for which country they could compare Bitcoin's power use with. Some nominated were: Greece, Ireland, * Netherlands, and New Zealand. I put a section on this in my Sustainable Computing course.

I tried asking Copilot:

"What country does Microsoft Copilot use as much power and water as?"

It replied:

"Microsoft Copilot, like all datacenters, consumes energy and water. Let's delve into the specifics:

1. **Power Usage Effectiveness (PUE)**: This metric gauges how efficiently a datacenter utilizes energy. The closer the PUE number is to "1," the more efficient the energy use. Microsoft tracks PUE globally and across operating geographies. While newer datacenters and those operating at peak utilization tend to be more efficient, the goal is to get as close to "1" as feasible¹.

2. **Water Usage Effectiveness (WUE)**: WUE measures the efficiency of water consumption for humidification and cooling in datacenters. It's calculated by dividing the liters of water used by the total annual power needed to operate datacenter IT equipment (measured in kWh)¹.

Now, let's address your question. Microsoft's water consumption for its cloud data centers increased by **34%** in 2022, totaling **6.4 million cubic meters** of water. Generative AI workloads are partly responsible for this surge³. However, it's essential to note that this data pertains to Microsoft's overall operations, not specifically to Microsoft Copilot.

In summary, Microsoft's datacenters, including those supporting Copilot, have a significant environmental impact. However, pinpointing a specific country for comparison would require more detailed data. 🌐💧🔌

Source: Conversation with Bing, 22/02/2024
(1) How Microsoft measures datacenter water and energy use to improve Azure .... https://azure.microsoft.com/en-us/blog/how-microsoft-measures-datacenter-water-and-energy-use-to-improve-azure-cloud-sustainability/.
(2) Microsoft's water consumption jumps 34 percent amid AI boom. https://www.datacenterdynamics.com/en/news/microsofts-water-consumption-jumps-34-percent-amid-ai-boom/.
(3) Microsoft 365 admin center Microsoft 365 Copilot usage - Microsoft 365 .... https://learn.microsoft.com/en-us/microsoft-365/admin/activity-reports/microsoft-365-copilot-usage?view=o365-worldwide.
(4) Measuring usage, adoption, and the impact of Microsoft Copilot. https://techcommunity.microsoft.com/t5/copilot-for-microsoft-365/measuring-usage-adoption-and-the-impact-of-microsoft-copilot/ev-p/4051091."

Cryptomining Consumes More Energy Than Gold Mining

Krause and Tolaymat (2018), estimate that "mining" of the four commonly used cryptocurrencies (Bitcoin, Ethereum, Litecoin and Monero), from 1 January 2016 to 30 June 2018, produced between 3 and 15 million tonnes of carbon dioxide emissions. They also estimate this is more energy intensive, in terms of value produced, than conventional mining of copper, gold, platinum or rare earth oxides.  Only production of aluminum was more energy demanding than cryptocurrency. One worrying trend is that as the cryptocurrencies drop in value the energy needed to mine them increases. The good news is there are alternatives, such as Ripple, which does not use mining.

ps: Last semester I asked my ICT Sustainability graduate students at the Australian National University: "Are Bitcoin and Blockchain Bad for the Environment?". 

Reference

Krause, M. J., & Tolaymat, T. (2018). Quantification of energy and carbon costs for mining cryptocurrencies. Nature Sustainability, 1. url

https://doi.org/10.1038/s41893-018-0152-7

Zero-Carbon Energy for the Asia-Pacific Wins ANU Grand Challenge

The project "Zero-Carbon Energy for the Asia-Pacific" has won the Australian National University Grand Challenge.  This project will be funded to research how to capture solar and wind renewable energy on a large scale in north-west Australia, then export it to Asia.

Zero-Carbon Energy for the Asia-Pacific

Greetings from the ANU Grand Challenges pitch night in Canberra.  Four teams are presenting their ideas. One will receive funding. The last is "Zero-Carbon Energy for the Asia-Pacific". This project proposes to capture solar and wind renewable energy on a large scale in north-west Australia, then export it to Asia.

The energy will be exported by undersea cables, as synthetic fuel in supertankers and as refined metal. The project has to work out how to negotiate with traditional owners of the land on which the energy will be collected, and with the nations the cables and tankers will need to transit.

One issue is that cables and supertankers of fuel make attractive targets for terrorists and nations. The tradition approach to such defence would involve UAVs and rapid reaction air and seaborne forces. However, this could also include the local community as part of the early warning system.

Decreasing Campus Energy Use With Flexible Classrooms and e-Learning

The Australian National University is preparing an Energy Master Plan. The aim is "... to create an Australian world class energy efficient, low-carbon, least-cost campus ...". One target in the existing ANU Environmental Management Plan 2017-2021 is to "Reduce energy use per person by 20 per cent by 2021". New forms of education can help, by using the buildings more efficiently and moving some education online. I suggest these techniques can be used to reduce energy use per person. In particular, teaching staff can be trained to teach in flexible and online mods. Also, a system can be used for students to book a seat.

Use Classrooms More Intensively

One way to reduce energy use per person is by increasing the intensity of use of the buildings. Empty rooms use almost as much energy as full ones, so the more use of each room in each building, the lower per person energy use.

New ANU Buildings
(artists' impression).

One way to increase the use of classrooms is to make them more flexible. The ANU Union Court redevelopment is a prime example of this. Two new buildings will allow forms of education beyond conventional lectures and tutorials.

ANU Culture and Events Building

The ANU Culture and Events Building will have rooms with retractable seating. At the press of a button, the room can be changed from a flat floor for a conference, to theater format, with tiered seating for a lecture.

ANU Collaborative Learning Building

The ANU Collaborative Learning Building will have flat floor rooms with movable furniture and walls. This enables the same space to be used for different size and format classes. Walls can be folded back for a large class or moved in form more smaller ones. Desks can be in straight rows for a more conventional eyes-front classroom, or in circles for a more collaborative approach.

Flat floor classrooms are conventionally used for small groups of up to a few dozen students. However, with technology and new teaching techniques, such rooms can accommodate hundreds of students. 

Training to help make use of classrooms

Training can help make flexible classrooms more efficient, in terms of teaching and energy use. Academics familiar only with conventional lectures and tutorials will tend to use the new flexible teaching spaces only for old fixed teaching formats. To use the rooms flexibly requires the staff to be training in flexible formats.

As an example, an instructor used to conventional lecture driven teaching will find the idea of hundreds of students in a room working in teams a recipe for chaos. The instructor needs to be taken through the theory and practice of how this works, and ideally act as an assistant in such a class before being required to be the lead instructor.

 The ANU TechLauncher events for Team Formation and Bootcamp run in the ANU's large flat floor teaching rooms, at 7-11 Barry Drive, have demonstrated that group exercises with 300 students can work. By introducing academics to these techniques, they will be more confident to use them in courses.

Use e-Learning to Supplement Campus

E-learning can also be used to increase the intensity of campus use. Most courses at Australian universities are now, to some extent, "blended", with part of the tuition on-line. However, to make the best use of the technology course designers and instructors need to be trained in on-line techniques.

Academics who have only training and experience in face-to-face lectures and tutorials will tend to continue to rely on those techniques and be reluctant to try e-learning. There will be concern the students will not do the work, or will cheat, which can be the case if materials are not well designed and trained staff are not available to run on-line courses.

As an example, the ANU course COMP7310 "ICT Sustainability" is run entirely on-line, with no lectures and no examinations. This course runs alongside ANU's face-to-face courses, with the same status and meeting the same quality standards. However, design of this award winning course required the assistance of specialist e-leaning professionals. It also helped to have undertaken graduate education training at ANU and other institutions, focused on in e-learning.


With the appropriate level of training it is feasible to aim for the typical university program to be a 20/80 blend: 20% in a formal classroom setting and 80% on-line outside the classroom. An example of this approach is the ANU Techlauncher program, where students undertake a group project building software for a real client, in government or industry. Students are expected to undertake 10 hours of study per week, but are only required to attend a 2 hour formal session in a typical week.

Ideally students should be on campus no more than the equivalent of one day a week. Students should be out in the real world learning and practicing their skills.

Double University Per Person Intensity of Use

Less than half of students attend a typical university class. A survey at ANU found "... attendance declines over semester to around 30% of original signup ...". This is not confined to ANU and is not a new feature of universities. However, universities still tend to allocate teaching space based on the number of students enrolled in the class at the beginning of semester. As a result the classrooms tend to be less than half full after the first few weeks. This is a waste of space and also a waste of energy needed to maintain these spaces.

One way to engage students, and thus attract them to class, is with group activities, as is done with ANU Techlauncher. This ensures almost full attendance at interactive group activities. Blends of on-line and face to face activities (so called flipped classrooms) can keep students coming to class. However, not all students need to attend every class and many classes will have one third to one half attendance. This can be incorporated into space use planning.

Students can be given the opportunity to book a seat in class a few days in advance. Classrooms can then be allocated, making use of different size rooms and flexible walls and formats, to suit. A reasonable aim would be to double the intensity of use of classrooms, thus halving the per person energy use.

Reduce Staff Campus Use

Persons use of the campus includes staff as well as students. Most teaching undertaken at a university is not by full time, permanent academics, but by graduate students and part time staff from industry. E-learning is a particularly useful way to make use of part time staff skills. Industry professionals can teach students without having to leave their workplace. Researchers in the field can teach students. However, this will require the staff to be trained in techniques for e-learning. The ideal way to conduct this training is on-line.

More on:

1. Classroom design 
2. Digital Teaching 

ICT Reduces National Carbon Emissions?

Dolores Añón Higón

Hign, Gholami and Shirazi (2017) found an inverted U shaped relationship between ICT and carbon emissions between countries. The researchers  used an ICT Index, made up of telephone subscribers (both fixed and mobile), PCs and Internet use. They mapped this against carbon emissions for difference countries. At first emissions increase with ICT, but this then plateaus and then reduces. The authors conclude that "... sustainable environment may be linked to greater development of ICT ...". Of course it may be that use of ICT is just a side-effect and this is not a causal relationship. That is, it may be that as people get richer they can afford to use ICT more. On the other hand it may be that the reduction in the cost of ICT, with widespread access to broadband and smart phones in developing nations can reduce energy use.

Wagner's masters thesis (2017) explores the concept of a "Green Global Golden Age", where ICT decouples development from resource use. The author poits out that, not surprisingly, it is not that simple:  ICT can address social issues, as well as simply increasing material needs. ICT may not be able to simply substitute for material goods and in some cases social means may be a better option for improving wellbeing.

Reference

An Hign, D., Gholami, R., & Shirazi, F. (2017). ICT and environmental sustainability. Telematics and Informatics, 34(4), 85-95. URL https://doi.org/10.1016/j.tele.2017.01.001

Wagner, K. C. (2017). Towards a Green Global Golden Age?: ICT enabled cornucopian sustainability and a suggestion for its reform. Master Thesis Series in Environmental Studies and Sustainability Science. URL http://lup.lub.lu.se/student-papers/record/8912373

Online Education Reduces Carbon Emissions

Marieke Versteijlen

Versteijlen, Salgado, Groesbeek and Counotte (2017) looked at if online education reduces carbon emissions. It may seem obvious that if students don't have to travel to a campus and don't use a classroom, then their energy use and so carbon emissions would be reduced. But as the researchers point out, energy and emissions at home may increase. Also if students still have to attend a few courses a long way from campus (in the extreme case international students fly from country to country) this can offset daily travel savings. However the analysis concluded that student travel was between 40 and 90 percent of all emissions for Dutch universities.

The authors point out that online education is not the only alternative and students can be encouraged to use public transport. Surprisingly, despite it reputation for the use of public transport, emissions from student travel in the Netherlands were much the same as the US. Another point made was that staff need training in how to provide development of online education.

Australian National University is now accepting enrollments in my course "ICT Sustainability" (COMP7310), commencing in February. Already I have a small segment to the course on "Are Bitcoin and Blockchain Bad for the Environment?". But I though I should have a look at what else has happened in the field since the course was last run.This is the first paper I looked at. More to come.

References

Versteijlen, M., Salgado, F. P., Groesbeek, M. J., & Counotte, A. (2017). Pros and cons of online education as a measure to reduce carbon emissions in higher education in the Netherlands. Current Opinion in Environmental Sustainability, 28, 80-89. URL https://doi.org/10.1016/j.cosust.2017.09.004

Social Cost of Carbon

Greetings from the Australian National University in Canberra, where Professor Christopher Kellett, University of Newcastle, is speaking on "The Social Cost of Carbon - Mitigating Global Warming Whilst Avoiding Economic Collapse via Optimal Control for Carbon Pricing". He went through a mathematical model of the climate and the economy. Such models have a checkered history, going back to "Limits to Growth" (1972). Professor Kellett made some useful comments as to how reasonable how assumptions underlying such models. However, assumptions from engineering and control theory may not be applicable when what is being modeled is human behavior. As an example, Professor Kellett questioned if carbon reduction could be made very rapidly, given the time taken to install new plant and equipment. However, reductions can also be made due to human behavior, where for example, a small increase in price can lead to people using less energy and therefore causing less emissions, within days.

Two factors suggest that  Professor Kellett's work will be more valuable than previous efforts: he has a healthy skepticism as to the accuracy of such models and the model produces plausible prices for carbon (US$22.90 to US$65.10 tonnes of carbon dioxide).

There is a paper:

Weller, S. R., Hafeez, S., & Kellett, C. M. (2015). Estimates of the social cost of carbon using climate models derived from the CMIP3 ensemble. Retrieved from https://www.researchgate.net/profile/Steven_Weller/publication/274392763_Estimates_of_the_social_cost_of_carbon_using_climate_models_derived_from_the_CMIP3_ensemble/links/55656c2a08aec4b0f4859bcf.pdf

ANU Students Cut City Air Pollution Using ICT

Air pollution has been exceeding recommended levels, not just in Beijing and New Delhi, but also in Sydney. Solving this problem might make a suitable project for my ICT Sustainability students, who start 15 February 2016 at the Australian National University. Normally they estimate carbon dioxide pollution from a single company and suggest how to reduce it. But the same techniques are applicable to other pollutants and on a city-wide scale. Students might like to research this in more depth as well so I created it as a project topic:

Cut City Air Pollution Using ICT

Student research topic at the ANU College of Engineering & Computer Science

Air pollution has been exceeding recommended levels, not just in Beijing and New Delhi, but also in Sydney. This project will investigate using the ICT Sustainability techniques usually used for carbon dioxide pollution from a single company and apply them to other pollutants on a city-wide scale. There are many ways to use ICT to reduce pollution in theory, but the problem is in how to get people to use them in practice. This project therefore emphasizes the Human Computer Interaction (HCI) aspects: applying knowledge of human behavior so people can use, and want to use, technology which cuts pollution.

Some techniques, discussed in Worthington (2011) to reduce pollution are:

1. Dematerialisation: use computers instead of physical goods. For example, replace traveling to meetings with teleworking.
2. Smart Motor Systems: More efficient computer controlled motors to reduce energy use.
3. Smart Logistics: More efficient good delivery to reduce truck use. These techniques might be used for example, for a parcel service using hybrid or electric trucks. Also an Uber-type service with electric vehicles could reduce pollution from private cars.
4. Smart Buildings: Computer control of buildings to reduce energy use. This is normally thought of being applied to large corporate buildings and new apartment blocks, but increasingly can be applied to smaller, older, buildings.
5. Smart Grids: Computer controlled electricity system to more efficiently use power, especially renewable energy. An example is Dr Lachlan Blackhall's Reposit Power in Fishwick. 

Reference

Worthington, T. (2011). ICT Sustainability: Assessment and strategies for a low carbon future. Retrieved from http://www.tomw.net.au/ict_sustainability/introduction.shtml  

Reducing Australian Electricity Demand

Dr Hugh Saddler will speak on "Electricity demand and Australia’s renewable energy targets: where to?" at the Australian National University in Canberra, 12.30pm, 26 August 2015.

"Average annual electricity consumption per residential consumer has been falling steadily in every state since around 2009. The presentation will examine the factors which may explain this reduction and speculate on where electricity consumption may go in the next few years. It will also include some preliminary results from an analysis of how low income households use electricity."

E-learning an ICT Solution for the 21st Century

The report "#SMARTer2030: ICT Solutions for 21 st Century Challenges" (GeSI 2015) suggests that computers and telecommunications can provide a "triple win" delivering "significant environmental, social and economic benefits":

1. Environment – Decreasing emissions and resource consumption whilst allowing for growth
2. Economic – ICT is good for business, creating new revenue opportunities and reducing cost
3. Social – Boosting incomes, cutting costs and improving lives

The areas the report suggests will benifit most are: Health, Learning, Building, Food, Mobility, Energy, Work/business and  Manufacturing. The authors suggest three groups who need to act are: Policy makers, Business and Consumers. Curiously ICT professionals, educators and researchers are not included.

The #SMARTer2030 report is incorrect in suggesting that in the future e-learning will be used for collaborative, location-independent education. This revolution has already happened, it is just that most people (including most university professors) have not noticed. In 2008 the ACS commissioned the design of an on-line course on how to reduce carbon emissions by and with ICT. The ACS approach treats it students as working professionals who collaborate on-line and produce real proposals for their employers. The versions of the "ICT Sustainability" course are now also run by ANU in Australia and in North America. This form of education is also in routine use for other courses for millions of students world wide.

Environmental impact e-learning much lower than campus

In looking for some papers for my ICT Sustainability students to read, I came across "Designing low carbon higher education systems: Environmental impacts of campus and distance learning systems" by Roy, Potter and Yarrow at Open University UK (2007). They, not surprisingly, found that Distance Education uses about 65% less energy (and correspondingly lower carbon emissions) than campus-based courses. But they found e-learning only produced a modest 20% energy reduction over print-based courses, as students tended to print out the e-learning materials. It would be interesting to see if this still applies in 2015, with student more used to electronic documents and having mode potable devices to read them on.