 Good afternoon and good morning. Thank you for joining us today again to discuss the end use, detailed energy consumption by end uses. I'm Thomas Logosi from the IEA Energy Data Center and we'll now explore the why and how we can understand the demand side of the energy system. Okay, so here's the agenda note. You'll see that some of the data presented here might not be the most up-to-date. It is often because COVID has such an impact on consumption patterns. So to give you the general idea, best was to do the pre-COVID data. And again, please feel free to ask any question on the chat or else. Okay, so let's start with the why. Why do we need to look at the demand side of the energy system beyond the sectoral level? First, as for any budget, the energy system is a complex set of intertwined flows. Here we have incoming energy flows on the left like revenues in the budget and outgoing flows on the right, like expenditure. And to manage a budget, you need to study both the input and the output. So here we'll need to look at the details of how we consume energy, the output side. Here at the IEA's Energy Data Center, we like to split energy consumption into four main sectors as we've seen before in the other presentations. So we have industry that is manufacturing and other industries like mining. We have services, we have residential buildings and we have transport. In the IEA countries, transport is the most consuming one with about one third of consumption. Then comes industry with about a fourth. And then buildings, residential buildings with about 20% and services with about 15%. So as you can see, no sector can be left behind they all weigh a significant part of the output. Starting with transport, we can split it further already into the four main modes. We have road which takes up more, we take up more than 90% of the consumption but also rail, air and water transport. Currently, oil products still account for the overwhelming majority of the consumption as you can see on the right with 98%. And even within the oil products too, namely gasoline and diesel represent a huge part. But what about the split between passenger and freight? That is how much is consumed by cars versus trucks? And also what is the typical distance covered by passenger or goods depending on the vehicle type? Or what is the load factor? How does it differ from other countries? These are all the type of questions we want to look at in more details. Looking at the industry by fuel, we see a variety of cases. So in some countries, fossil fuels are still abundantly used about three times as much as in the least fossil-dependent country of the IA, while in other countries, electricity is the most used. Again, factor three between the largest heat and electricity use and the lowest. To understand such data, we would need subsectoral details to know how these fuel consumptions are distributed between subsectors, allowing to track technology deployment, assess and reduce security risks, for instance, or even sensitivity to energy prices. Then the residential buildings. Here again, multiple factors influence which fuels are used the most. The need for cooling or heating, the availability of some energy products, policy measures and lifestyles and habits. In some cases, in the cases pictured here, we can see electricity is central in Japan, but almost no biofuels and waste are used, only 0.02%. While in Chile, it provides a large share of residential consumption. More data will enlighten us about the most consuming end uses, their dependence on some fuels, cultural habits in each country and technological penetration in households. Finally, we recall these fuel mixes and other factors will impact the emissions of each sector. Let's look at the time series of energy consumption and emissions for the services sector, which is often very electricity dependent. We see that the decarbonation of the electricity production leads to decrease in emissions in light blue, while overall the consumption increases in the meantime by about 6%. Still, if you look carefully, you'll see that the post-COVID energy consumption rebound is smaller than the emissions one. This is the kind of analysis that can be performed when we have enough data to be able to drive policy decisions, is to look at why we have such a carbon-intensive update. Interestingly, the sector can be looked at using two angles, the economic side, splitting consumption by subsectors, such as health or retail, and the building side, splitting consumption by end uses. These would allow us to decompose the drivers of the trends that we see here. For each sector, we can glimpse why it is key to have detailed and reliable energy and use data. For country comparison, as you can see here between the various countries, we have the final energy consumption by sector. For sensitivity and security analysis, with the fuel split here, for perspective analysis and impact tracking, why are some subsectors having their consumption decreasing a lot, or for climate goals? On this last point, climate roadmaps than other environmental targets can be achieved only with control over our energy spending. Here, you see a graph from the new update of the IA net zero 2050 roadmap, a major analysis from the IA. We can see how much each mitigation major contributes to the net zero target. But if you look carefully, you'll see they almost all depend on the demand side of the system. We have lifestyle impact, that is demand, then efficiency, which is the main side again, then a number of fuel switches, that is how the tools that we use and which consume one fuel will be adapted to consume something else, and then CCUS. So the fine understanding of energy consumption is a key aspect to manage the energy system. So this is why the IA decided at the 2009 governing board meeting at ministerial level to collect a detailed and use consumption data annually. The action plan that was approved at that meeting included the objectives to promote energy efficiency. So any data and statistics need to be collected annually by member countries to develop and track consumption patterns and efficiency indicators. The IA in concert with international experts developed a questionnaire template that we're gonna have a look at a bit later based on a robust and harmonized methodology. This includes also internal consistency and comparison checks that would perform hand in hand with our counterparts in the institutions providing us with the data to ensure data quality and assist when issues arise. So indeed, let's have a look at the demand side data collected and which is essential for analysis by looking at the questionnaire. So let me know if you don't see well or if it's too small. Let me actually quickly check this. It looks all right. So the question looks like this. You have a content page and then a set of tabs to collect the data. Six tabs here. So we also have the micro-commit data that collects data which are a bit cross-central. We have populations. We have data about the dwellings or household occupancy. We have data about employment and services for ARIA. And then we have economic activity such as exchange rate of PPPs. We have GDP obviously and value added by a sub-sector. Then in the commodity time, we collect data on physical production output from the various sub-sectors of the manufacturing and industry. Then going into each sector, we have for each of them a split and we see here all the different fuels for which we collect data. And these fuels are repeated for all the different sub-sectors. So for our industry, we have a few other industry sub-sectors and we have a lot of manufacturing sub-sectors following the Isaac revision for classification such as manufacturing of textiles or wood and wood products and so on. The services, as I said before, you have two types of split. You have the split by NUCs with space heating, space cooling, lighting and all the building. And then you have the split by services categories that is sub-sectors such as retail and retail trade or accommodation services, health, education and so on. And again, for each you have a set of fuels to report. In the residential, we only have end-use splits. So we have space heating, space cooling and so on and we also have a certain number of appliances types for which we want the energy detailed. And at the end, you also have some specific residential activity data, especially appliances stock from which we take appliances diffusion or unit energy consumption here. And finally, in the transport sector, we start with some specific activity data. So we have for the passenger segment, we have passenger kilometers split by various vehicle types like cars or motorcycles or buses. Then you have the same for the freight segment with ton of kilometers. Then you have a certain number of data for both segments. So vehicle kilometers split by vehicle types and vehicle stocks as well with a bit more details because that's often a bit easier to collect. And then you get into the energy consumption. And again, we split in various fuels and all these fuels are reported for a certain number of vehicles types such as cars, motorcycles and buses and passenger trains and so on in the passenger segment and the same in the freight segment with for instance trucks and then a certain number of specific type of trucks and so on. So this was the question and then you have a number of times which allow to give remarks and communicate with our country parts. So now back to the presentation. So now that we have seen what we collect, let's look at what we can learn from this type of data that we collect in the questionnaire. So the transport sector as one could expect passenger cars are the most energy consuming vehicle type. For instance, they can for half of Korea transport energy consumption and nearly 2,000 in the USA. So what do we need to know in addition to understand why this happens? While looking at activity data to start with, we can see that indeed cars account for a higher share of passenger activity in the USA about 80% than in Korea with 50%. And the same data will be needed in the freight segment. Moving to industry, industrial database sectors allow you to put focus on key parts of manufacturing for each country. For New Zealand to understand the major trends, the six most consuming subsectors are representative as they consume about 95% of the manufacturing consumption. Well, they cover just less than 60% in Mexico, for instance. Now, if we split these six main subsectors by fuels, we can see general trends and we can see also country specificities. So for instance, we can see that chemicals often depend on gas. This is free with countries and we can see that wood or paper printing can depend a lot on biofuel and waste like we see in New Zealand. Basic metals here depends a lot on electricity. While in Mexico, it depends a lot more on fossil fuels. This depends on technology choices and technology penetration. So tracking these splits allow to anticipate the impact of policies and regulations. The same exercise can be done in the residential consumption patterns by end uses. This shows the influence, for instance, of climate. In Brazil, there's no space heating at all. And we can also see the share and fuel split for cooking is very dependent on lifestyles, such as habit of eating out, cooking methods and types of stoves. The fuel split of most of the consuming end uses are central to understand the reliance on imports, the impact of price variations or the design of emissions and efficiency programs, but also health related issues, such as indoor pollution. I'll quickly go further and skip the services sector to go directly to this type of analysis that we can do when we have enough data. This is called a decomposition analysis that can be performed once we have the activity and energy consumption split following the same methodology on both the activity and the energy. This is very important. And we can split the energy consumption trend into three main factors or three main drivers. We have activity, larger consumption due to more output, for instance. Structure, which is from the same consumption, some sectors generate more or less wealth. So the split between sectors influence the energy consumption trends. And we also have efficiency, which is how much energy is needed to produce one ton of goods. This analysis here helps understand the driving force of the energy consumption and design efficient policies. We can do the same in the transport sector. Interestingly, we have to split between freight and passenger because the activity below is a ton of kilometers for the freight and passenger kilometers for the passenger segment. So for transfers, we can do that. And here you have it for the United States. We see the data for energy and activity. Sorry, we need data for energy and activity across all modes, road, rail, air and water. And ideally all vehicle types. For instance, for road transport, we will need cars versus buses at least and possibly motorcycles as well. And again, methodological alignment is key in this type of analysis. So if we look a bit closely at these graphs, we can see how we can disentangle the different drivers. Activity, which is the amount of goods or the number of passengers, but also the distance traveled. The structure, which is a modal shift, for instance, if you move from car to buses or to trains. And efficiency. Efficiency here takes into account the load factor, the engine efficiency, the driving behavior and all those things like that. So in the example here in the passenger segment, efficiency gains in dark green have compensated for increased activity overall in blue. And it leads to more or less stable consumption in kind of gold. On the freight side, efficiency deteriorated. Here it went up. And it's probably through lower load factors to accommodate maybe with more stringent deadlines, for instance. These are the kind of things that we can track here. So from all this data that we collect, the IEA data center compiles a large database available in our web page. The main publication in Excel provides all this data as well as a set of interactive graphs and visualization tools to assist the analysis. This full database is also available in IVT format. We also provide two free samples, the highlights and the demo and availability files in Excel too. Another way to browse our data is the data export on our website with a web interface to produce graphs again. And finally, we maintain a documentation with methodology and detailed country notes as well as two manuals on statistics and policymaking. Currently the EI database covers 61 economies that is almost 2 billion people and we wish to include as many countries as possible and we're always working bilaterally with multiple institutions in countries or international institutions to make this database as representative and complete as possible. And we invite obviously any of you to reach out to discuss how we can assist you on collecting and use data. We're going to see that later as well. In order to assist the user, we publish our database in a hybrid format with data tabs. I'll skip this for more time. And now let's have a look for a few minutes for each sector how to collect data because this is no simple task. So as I said before, we have four data collection methods, mainly the administrative sources, the surveys, measuring and finally modeling to complete these. For commercial activities, that is on the industry and so this is a great basis of information can be found through economic authorities such as Chamber of Commerce, Regulation Authorities or tax reports. Complimented with energy companies data and sales report, sectors are often already collecting a significant amount of data for themselves. So the main trick is to gather them and ensure methodological consistency. Buildings are often requiring the most varied data sources, whether residential buildings or services from an end-use point of view. With basic activity data, such as population, employment, dwelling or for area covered by administrative sources, most energy consumption data relies on the accommodation of service and measuring base cases. Modeling all allows to avoid repeating coffee surveys too often. And finally in the transport sector, many data exists from all consumption to car sales and registration, but these data need some manipulation, if not modeling, to obtain data split by modes and vehicle types. So surveys are also needed to provide the basis for model calibration or to complement where data is lacking. So these were the main ideas, for many countries, every sector, every situation might appear unique. The truth is other countries' experiences are incredibly valuable resources when one needs to collect new data, which is why the IEA gathered online numerous data collection practices, along with some documentation, survey examples and reports from many countries around the world covering a huge diversity of situations. For sure you'll find some insights by browsing this tool. We can put you in contact with the institution in charge of this data collection, if they be. So let's have a quick look at some examples. So typical feedback includes some tips about the tools, whatever the method, better start by carefully designing the tools you use before anything else, based on your needs and resources. They start with looking at pre-existing data collections to expand them and ensure methodological consistency and then carefully designing questionnaire or platforms, including accessibility, assistance and various languages, for instance. And then to ensure the maintenance cost would be limited and the data quality would be preserved, taking time to accurately document the methodology is key. To keep the data flowing in, we need long-term institutional arrangements, we need trained staff and foster relationships with colleagues and counterparts. This is true within institutions, but also to third-party sources, whether academia or the private sector, whether they are respondents of the service or not. I quickly finish to compliment and strengthen this framework the IEA developed a guide to build national world maps. It aims first at helping countries locating their departing point and to identify the targets according to national interests and priorities. In addition, building on countries' experiences, it assists professionals and decision-makers when willing to develop and use data collections and efficiency indicators, this guide provides a generic method acknowledging the contextual differences. International collaborations are crucial to speed up the learning process and to keep up with the pace of technological, economic and geopolitical changes. So with this in mind, thank you very much for your attention and please feel free to comment or question any part of this presentation. Thanks.