 Okay, so we've just gone three o'clock, so I would like to welcome everybody back again this week to the next talk in the series that's organized jointly between ICDP and the University of Trento, so of Environmental Meteorology, and this week it's ICDP that's hosting the talk and so I have had the great pleasure of introducing Professor Anna Grady. She obtained her PhD in economics at the Toulouse School of Economics and then she was a post-doctorate researcher at the London School of Economics and she actually came back to Italy for a little while. I remember Anna telling me at one point that she also returned to Italy for a short period of time as a researcher here, but then she moved to Paris again and because I think you were in Paris before you came to Italy, I remember when we met, so Anna and I actually know each other through, we were both on the scientific advisory board of Waskell Project in West Africa, we were just catching up on that now for a number of years together, and now Professor Anna Grady is a professor of economics at Paris Dauphin University, she's also the scientific director of the Climate Economics Chair and she is the scientific director of the Natural Gas Economics Chair and she is going to be presenting to us today a talk about wind farm revenues in Western Europe in the present and future climates. Anna, I will pass the floor to you and thank you very much for joining us and taking the time for this presentation today. Thank you. Thank you very much, Adrienne. It's a pleasure to be with you in my own country, even if at distance, and thank you for the introduction. As you may have understood my research interest, lighting the economics of the energy and climate transition in one of the natural ways to look at this problem is to, in particular, study the role of renewables. So, I'm an economist of course, but this work that I'm presenting today is the result of a joint work. So, the names would have been long to be here in the slide, but you had it in the abstract. So, it's a joint work between economists and climatologists, meteorologists. So, and we have in the team two meteorologists and three economists. So, the twist that I'm presenting today is of course more related to the economic question. So, I beg your pardon in advance if there is something on the climate side and the sense of the scientific part of the climate side that is not clear enough. I'll try to explain that in my own words and in my own understanding. It's a team and actually Adrienne knows also one of my co-authors and perhaps you know it as well, Philippe Dobrinsky. We have a cooperation since a few years and we have studied different, under different angles, the question of renewables, starting with a study that we have published and it was focused in Italy. The project and the paper I'm presenting today is specific on windmills in Western Europe. And if I can just introduce very briefly the question we are going to ask and try to also give some answers, is that the importance of renewables and in particular of windmills is one of the pillar of the decarbonization of the energy industry. And actually as you can read in this slide, when we think about intermittent renewables, wind has a very clear interest and we see that in terms of capacity there was even an increase even last year despite the COVID crisis because of course these investments were planned in advance and they happened to occur. So worldwide we have an established capacity that is important, 65 gigawatts even if of course it is not displacing coal or gas or not as much as we would like to still with additions that are I mean regular and continuous. It's a quite dynamic sector. The one of the reasons why we also find these capacity additions is that it's a sector as many as it is the case for many renewables that is subsidized under different forms. We can have something that I'm going to discuss later on that is feeding tariffs. Those are fixed terms contract that span over from 10 to 20 years depending on the countries and they rely on fixed revenue that is granted to the production of electricity. So now this is the difference that of course this is one thing is the capacity that is installed and given that they are intermittent another one is the production but feeding tariffs are really related to the production of renewables or there are also other forms that may be found for instance in the United States with specific power purchase agreement, public and private but in any case some protection I would say in terms of the recovering the cost that remains quite important when installing windmills capacity. Here you can see just a snapshot of the relative importance of wind and how it has evolved in the last years accounting for a small but substantial part of renewable energy. Which is the question? There is of course this subsidy which is for the moment granting the sustainability the financial sustainability of these investments but there are some problems or some uncertainties when investing in windmills. One is that due to the increasing penetration of renewables into the electricity mix what we can observe is that the price the electricity price decreases. Feeding tariffs are something that is not going to last forever and therefore when you plan today your investment for the next 20 to 50 years at some point you know that this kind of fixed remuneration is going to stop. So this is the first problem this is the idea of are they still profitable and they will be profitable in the future. The second question is that how much they can produce and this is related to the second source of uncertainty which is the uncertainty linked to the availability of the natural resource that is wind. Now in that of course the sites that are the most profitable ones become more and more scarce. Therefore there is an important impact that is given from the due to the variability of wind. So this is really the question that we issue that we take in this paper and we look at the long-term profitability of windmills. Taking into account those two sources of uncertainties on one side the economic uncertainty and the other side the variability of the natural resource. Not to be very long on the on the literature why this paper is interesting just to say that of course there is a bunch of paper on the dynamics of wind power cost. During the last 20 years we have witnessed a decrease of this cost due to different form of learning effects in wind like also in the photovoltaics industry. On another side we have papers on climate data that have already tried to put some methodologies even a little bit I would say plug and play for the investors in order to evaluate the wind potential in the future. But we find a few papers if not to say very few papers that take the same perspective of understanding on one side the profitability of wind on the other side the variability of the resource not only as private operators but also on the public side. So one of the interest points that we are going to discuss is also the cost or the simulation of public support measures needed to make wind energy profitable. And of course in the long term this also means to understand the role of the public intervention that is going to help or that is needed also to meet the objectives of integration of renewables into the electricity mix. Of course there is a target under the latest European directives. So we have targets in terms of percentage of renewable energy in final consumption and we know that to attain these targets actually 27% the investment in wind farm and solar farms has to progress. So the question is really also on the agenda of the not only the energy transition but even more recently of the different forms of green investment or the green deep. So in a nutshell what we do we try to quantify the uncertainty in particular in a very simple indicator of profitability the net present value of standardized wind farm in different European countries. And we evaluate also the level and the total cost of the subsidies needed to guarantee the profitability of the representative wind flits that we are going to model. What is the specificity in terms of methodology? The previous papers have treated separately either the economic questions of profitability or the focusing on the cost decrease not really on the revenues and other papers that have studied the climate question. So what we do is that we have a common framework to do so which means that we have the same methodology or the same methodology at least the same guidelines to build a common methodology on one side to understand the drivers of the actual what we call and I will be more explicit later on the present economic and climate conditions and then we do some scenarios some projections for the future but this is done in a very as from the beginning in I would say two pillars of the same model based on the criteria that can really interconnect the economic side and the climatic side. So just to give a few details we build a localized model so the definition is very precise on the territorial level for wind power output in different countries and we make this coupling with some characteristics of the electricity sector in particular the consideration and the calculation of the present and future electricity demand. We also simulate electricity pricing taking into account some structural characteristics such as in particular the variation of wind the characteristics of the loads and the way in which price are formed in this sector. This is where we had the cooperation with our team the meteorologists and the climatologists something that I've learned during this experience is to use this famous re-analysis data it took me a while to understand this also this kind of terminology and what does it mean to you I'm sure that is perhaps one of the clearest issues of question of the paper. We also use climate projections and taken from integrated assessment models to build scenarios and to make also some robustness well I'm not as an economist when we know that when these scenarios the exercise is interesting on one side but we may miss a lot of questions or a lot of factors that can modify them and therefore we are trying to make scenarios that are not very long term but they are I would say long enough in order to understand the profitability so they are over the lifetime cycle of wind farms. The methodology that we use could be replicated it is generally enough but we had data for France, Germany and Denmark. This is interesting because the electricity mix which of course we take as a starting point for these three countries is quite different France is dominated by as you know nuclear power. Germany is embracing very strong energy transition having stopped nuclear and having invested quite intensively actually already in wind power and Denmark that displays the most decarbonized mix. Denmark really as a mix that is almost 60% based on wind I mean today already and since they are integrated in the in the northern market they also use a lot of hydroelectric electricity so we have also in terms of the impact of penetration of renewables in the future the impact also on the relative impact on prices I mean electricity prices will be different and this is something that is endogenous that is we take into account in our model. The other question that is really an economic one is the cost of the support mechanism. We quantify them precisely in order to understand their impact on different scenarios that we build in order to make wind energy profitable. Overall we can describe the entire interval of the total cost depending on the different countries which varies in the different countries and just as a teaser there will be one country which has the lowest cost in terms of public intervention so we are going to unveil the cost of the energy transition even if this is one possible quantification. I don't know if I can quantify I can say that it is a number estimation or another estimation we try to cover different scenarios but still it is interesting to see how much it is going to cost to decarbonize the electricity sector even in France I mean where the it is something that we we hear quite frequently here in France that saying that since the nuclear power is decarbonized why should we bother for integrating wind this is something that we show it is possible and we show whether the profitability in France differs depending on compared to other countries. Anyway these kind of subsidies both today and in the future seem to be still important to guarantee the profitability of wind farms. So in terms of being now more precise on the characteristics of the model so the model is going to I mean the main interest is to have this idea of generating electricity prices and production from windmills. Today and in the future so little by little what we mean by today and the future will be clarified and Adrienne in the rat me please if there are some questions in the meantime in particular clarification questions if there are aspects that are not clear or not sufficiently explained in in my presentation. So we this is the the main input for the model in order to analyze the variability of wind farm revenues and also the the difference under the the scenarios today and in the future. Those scenarios are the one of the I would say the most important outputs of the model we have different scenarios with different assumptions in terms of not only the economic variability but also the climate variability and here also we see the sensitivity of the scenarios by I would say correcting in different ways for the climate variability and this in turn will have an impact also on the cost of public subsidies. More details on the simulations so we doing this exercise of building future price scenarios is always an approximation right so we had to start we wanted to start from some well known methodologies so we we use electricity demand in renewable penetration scenario projections from integrated assessment models. So then it is something so from existing I should say integrated assessment models that are used in these three different countries in order also to simulate future scenarios by the energy agencies or the system operators. One of the main point is to make these scenarios compatible with the scenarios that are used on the climates dimension so we combine them with data on the wind speed and temperature projections from what you know more than me again the regional climate model inter comparison that is the core data and Filippo is one of the leaders in the in the production of this this data and this corresponds to several representative concentration pathways RCP and in particular as when I was saying that the models are quite sensitive to some corrections that we can make on the climate dimension what we see is that using or not this RCP is going also to have an impact quite strong in terms of the assessment of the profitability of windings. So the the the paper is actually submitted to an economic journal and that is why perhaps the message that I'm going to deliver have a specifically an economic flavor in particular the objective of the of this construction is to really have an idea of the this loop between the penetration of the renewables the impact that they can have on electricity prices and therefore on the net present value which is an indicator the very the simplest one of the profitability of windings. What we do and I'm going to first explain it and then give you a figure so which is a nice I would say is snapshot of what how this can be represented so basically the methodology and compasses three steps knowing that as I was saying hopefully in a clear way from the beginning we have these two pillars one the is the economic and the other is the climatic one and we try to to to have common methodologies in order to come up with the economic scenarios. So in the first step is I would say like the basis in order to also make those the two parts of the model to communicate. So we reduce the bias of the long term long time series of wind speed measured at the surface and the temperature from the ERA 20 series analysis that have a different resolution of the ERA 5 reanalysis so this has been mainly the correction that has been introduced and that was quite interesting I would say is the information from my perspective because I mean these two database or these two reanalysis are quite different so we had to to have this common denominator and one of the statistical analysis that we do is to perform a quantile quantile correction in order to also minimize the variability. In the second step we start introducing the question of the economic questions so in order to to come up with the price model and therefore the long time series of electricity prices we have two compounded tasks one is to generate long time series of national production demand, calibrating the model on the past of the electricity prices in these three countries and on the other side we also downscale the wind speed time series too much with the time horizon at which the electricity prices are set are calculated that is an hourly frequency so we take in particular from the economic side the day ahead prices which are I would say one indicator there are many electricity prices that can be considered but they are the most transparent ones and also the ones on which we can find more reliable data and also the most liquid markets so we try to that is why I was saying we try to combine in a coherent way the data on climate that of course has a frequency that is much more precise than the hourly horizon but this was also compromised to have a common pass or a common time scale with the electricity pricing and finally we combine them in considering production function to electricity and this is going to give us the time series of local wind power production which is going then to generate our prices taking into account an equilibrium analysis that is given the characteristic of the demand that is also called the load and the characteristics of the supply we make therefore the equilibrium at each hour so this is I would say the summary of the two pillars of the model on the right in panel B we see some details on the electricity price dimension on the on the other side we have in the in the panel A the on climatic part and I have to say that behind the model there's an intense treatment of data and combination of data that has taken a lot of time and this has been done by actually Bastien who is working at Métro France and as you can see from the from panel B what we well when we do introduce economics it seems that the question is a little bit more complex but we made also our life a little bit more complex because we wanted to have this really equilibrium perspective that is we didn't just take the calibration of the model as many papers do in the sense that they take the the past data as we are used to do in economics we calibrate the models and we do projections we wanted to keep the model very rich and to have this equilibrium between demand and supply and at each step of modeling production on one side and demand on the other we correct for the important fundamentals in terms of the climate input in particular for the production wind speed and for the demand the surface temperature those are some of the fundamentals but the ones we are interested in and so we see that here we mix up climatic data with the demand in the production data that comes from and so every is the is the European body for the regulation of networks of electricity networks and they have a common framework to to publish data in different European countries so the also the time horizon have been quite a challenge because I mean the time span available for the climate data is much longer than what we have for the electricity production and electricity demand therefore we had also to take into account these differences by therefore simulating the data that we're missing again interrupt me if I can specify more even also our time constraints so now it is time to define what we call present and future very evocative terms to in at the end of the day we we have really quite convergence idea on what it is the present climate they respond in fact to the idea of having this distinction comes not in terms of making or having heterogeneous methodologies is just the focus that is slightly different what we call the present climate dataset it is actually used to study the variability of the wind revenues at the actual climate conditions referring to the ones that we can observe and we take also something on the economic sign as constant as given that is the current market arrangement in the in particular the electricity mix so in terms of wind power production as I would say the resolution is quite precise we look at their health prices and here we then introduce the differences between France, Germany and Denmark France and Germany in fact they belong to a similar price zone but still the the day ahead prices are different and this was interesting to us because there is this link between the of course the electricity prices and the existing electricity mix therefore this part so what it is the present climate is the data from the beginning of the 19 to the 20th century up to 2010 which are therefore some results that they can show so still we have long time trends well long time if I have to adopt an economic vision that is they are here displayed in a 10-year sliding mean of course we have also data more detailed but I wanted just to to see the long trends because they are easier perhaps to compare and to understand so we have here the three countries France, Germany and Denmark and what we have is in in green the electricity price and the national consumption in reds and the penetration of wind so what we can see is that of course there are some evident correlations there are some patterns that are common to the to the three countries this is also one of the results of the current markets design in in the electricity sector in which the prices that we look at the day ahead prices are done through a transparent market mechanism and therefore they are I would say quite neutral in the sense that they reflect some characteristics of the production and and consumption typically the importance of the temperature and the economic activity nevertheless as you can see also we have some divergences that in particular you can see the level that we have represented the the scale is a little bit different with Denmark having the lowest prices this is the combined result on one side of the size of the demand on one part side of the country that is a lower population and therefore a lower demand but also the fact that they use as I was saying a mix that is already relying on a lot of wind and hydroelectric which have very low variable cost well zero for the for the wind wind means and very low variable cost for the hydroelectric power and as soon as you integrate in this mix fossil fuel and thermal generation we have that the price increase Germany is somehow in the middle in the sense that they use over the period that we observe also a lot of coal which is far from being the carbonized but it is not very costly and then we have France where the on average I would say that the electricity price really reflects the long-term cost of the most important technology that is nuclear so it tends to be it tends to be aligned to the long-term cost variable cost of nuclear power what it is then the future and all the long-term so what we describe as the future is our exercise of simulation of scenarios and and it is perhaps conservative in the sense that we do not as was anticipating our scenarios are not very very long term these allow us perhaps to minimize the errors the mistakes the problem of precision of the assessments of this data and on the other side they are coherent with the fact that we in the simulation what it is important to us is to see the productive life of a typical wind farm so those are facilities that depending on the technology that we use a standardized technology can last up to 30 or 40 years so we took the maximum that is 2050 and therefore the future is for us bounded at 2050 we use therefore the future climate data set to study in particular the variability of the wind farm revenues and their value under different characteristics of the future that is different level of demand with energy penetration and of course the evolution of the climate here what we use is a special resolution that changes a little bit in order to have more precision in the in the long term and therefore they correspond to projected scenarios under the 4.5 and 8.5 and we use five different regional climate models this is from the climate and then try to speed up a little bit to give just the important information on the different scenarios for future electricity demand here we use this image scenario that is an impact assessment model quite commonly used for projections in for instance at the energy international energy agency and when we say electrification it means that there are some assumptions on which final users are going to shift our electrification for instance transport so this is what we refer to as a medium low and high electrification two scenarios for wind energy penetration that is again low and high so we have in total six economic scenarios that are completed every time with the climate scenarios so I'm just going here to give some precision or some further data on the wind penetration on one side and the level of electrification of the further electrification it has to be understood in terms of growth rates so when we say zero it means that there is no increase of the demand and therefore so on and so forth for the minimum and the high increase of the demand and therefore we have also those different scenarios and you see that another thing that is interesting in the model forget to see the to be specific before is that we also make the distinction very important between onshore and offshore wind so they of course their productivity is very different their costs also are different and therefore the profitability will be different and in turn also the subsidies will be different so at the end of the day what we have is 60 different price projections for each country they come from the combination of the rcp models the demand scenarios and the penetration scenarios and for each of the six economic scenarios just to again give synthesis of the model this is one of the output that we we can obtain so you see the point at which we start making our simulations so we have the yearly average price projections for the six economic scenarios so what we have in the left is the low penetration for the three countries France Germany and Denmark in the different colors blue black and and red and again we you see that this idea of the common trends but with different levels and also different variability they also depend on the penetration of wind on the other side and the extent of the electrification that is specified in the no trend zero percent increase of demand 28 percent and 43 percent respectively if i consider therefore the different roles there is as you can see also problem in 2020 which is i mean a technical problem of the of the data so in which are the determinants of the wind farm revenues of course we we have we also have extreme variation of the net present value um the over the long term and in particular we can calculate that the variation of the net present value that is the discounted sum of the revenues that the windmills can get given their production that is of course affected by the wind speed and the climate variability um including the subsidies so the order of the variation is one year of revenues um and these the variation can also depend on whether we use or not and which is the level also of the subsidies that we consider uh we consider different form of subsidies because both of them are actually being used and the countries sometimes switch from one to the other this is a problem of consistency of the support that we do not take also for us it is like simulating two different way of supporting windmills one is the feed-in tariffs that i have already explained and the other is the feed-in premium that is an uplift on the electricity price and we take the average values that have been used in the three countries for feeding tariffs on one side and for feeding premium this latter being used actually in Denmark whereas in in France we are still with feeding tariffs and in Germany um there was a shift on premium very recently when projecting the the future value of the wind farm based on historical production record then we take the point of view of an investor and what we show which is quite interesting is that in the paper that i have mentioned at the beginning that um uh want to somehow measure the variability of the net present value of just the representative windmills even considering the future variation of wind speed we see that clearly the precision of this kind of approach is very poor and that's compared to what we do that is having a consistent and a combined economic on one side and climatic model on the other we see that uh it is possible that the the first approach is of course simpler as i was saying plug and play but it may misconduct or may give values that can be in particular overestimated or underestimated it depends on the location of the of the project and therefore they are not very reliable or they do not give the the the the right information to private investors so a few details on on the simulation so we have these data expanding on 111 years we define an 81 virtual wind farm producer for the production points um so this is due to the y81 it depends on the definition the territorial definition of of the data is the grid point uh that we have used the the ones the same that are used in in the uh in the climate data and therefore as i was saying we have 30 years i see you appearing adrian either there are questions or i'm approaching the end uh what do you say right no carry on so i didn't want to interrupt you i was just no questions yet we'll come to those at the end but later okay um it's so at the end of the day we we have finally the calculation of the net present value um and uh since there is a lot of variation both in the price and in the wind speed and the climate of course we also take into account the one of the measure of the risk uh that is commonly used by investors that is the value at risk at the 95 20 um so what we have we we can see the again in a figure that is going to come just afterwards uh the uh the uh mean and the difference between the mean of the value of risk uh over all our project on 30 years and we compare them uh whether they are running with a subsidy under form of feeding tariffs under form of uh feeding premium and we um in this map the red contour the red line is the project that have a net present value equal to zero so um in fact the one that are we we should have perhaps inverted the the color line because the the ones that are there are in blue that appear in blue if they are in fact um not uh so they are displaying a negative uh net present value so and so what what we can see is that depending on the level of the subsidies which is on on the right we can make those those uh project uh value um in particular the uh profitability has to be sustained for offshore wind funds because as I was saying they are much more costly and they um well for instance in France we have uh there is a very limited offshore capacity so everything has to come uh in Germany uh there is um there is already an important installed capacity as well as in in Denmark therefore the additions that we could also um consider are very different uh for onshore wind farms um the profitability is still in France but because France again is at the beginning therefore what we simulate is perhaps some the exploitation of some sites that uh can be still quite profitable whereas the ones that are profitable in Germany and in Denmark have already been taken um in this is regarding the special variability in terms of the time horizon variability the NPV doesn't vary so much in the long term uh the standard deviation is below five percent but as I was saying uh again uh we see the sensitivity of these values to the representation of the uh um the climates in particular uh we have done some exercises by removing the trends on the ERA-20C wind speed and this has a large impact on the variability of the um in particular on the interquantile range of the of the project and if we make this correction um the um I would say that the economic variability uh seems to be um really uh uh questionable uh and they decrease basically the profitability uh all along the time that we consider um we also make for the future therefore we we make another kind of simulation uh what we do is that uh we make an exercise that in economics is especially for the um forecast of the penetration of renewables it is quite common that is uh we start we take the actual um capacity and we decommission it uh therefore we is like building again everything um and uh we start in 2021 so uh we have a new wave of of projects uh we display the net present value over the 10 product simulations and for each of the scenarios uh and for the different hypotheses of electrification uh just to give some some data the NPV is low uh without against subsidies uh it is positive in France but I mean one uh depending on what is still given the cost of this installation one million of euros per megawatt is positive but not extremely high so again there is room for subsidies again this is the uh you can see the projection on the on the future the map is dominated by the blue colors again the low or negative profitability with some uh slots in the red areas that are profitable depending also on the level of penetration uh last point uh the um as I was promising from the beginning the quantification of the support level that would be needed to guarantee the profitability of the wind fleet uh so basically we uh we simulate uh the level of feeding tariffs or feeding premium depending on the different scenarios uh that will uh make the net present value at least equal to zero so um which are the uh the um the uh evaluation that we come up with um we have um an interval given that we have simulation and different scenarios um and here the the the the teasing I was uh announcing uh is now disclosed uh if you are an investor uh you are more secure uh in Denmark in the sense that the premium for onshore and uh offshore sorry sorry between the brackets for in Denmark is uh is very low meaning that at market prices they are almost almost uh viable um in in Germany and in France uh if uh the there are no subsidy in the future then the profitability is not guaranteed uh we have subsidies that rise from uh 33 euros per megawatt hour to 66 depending on the case scenario but they can be as high as 70 euros per megawatt hour recalling that the prices that we had in the past were between 40 and 44 uh it means a lot uh and the the premium that is used in Germany can go up to 102 euros for offshore wind um so again uh the additional capacity will has to be subsidized and if we aggregate over the uh capacity that we consider in the next 30 years so you can read the uh the figures there I mean it is very high um depending on the country still the level of the public intervention uh is uh considerable nevertheless what we say is considerable but still small compared to the efforts that we have to make in order to uh I mean make the energy transition and the decarbonization and even farther now the net zero a reality that is everything that I wanted to to tell you at least before answering your question if I'm able to do so with my angle of an economist and thank you for your attention thank you very much Anna so just to remind people if they want to ask a question they can just type it to me in the chat and then I can invite you to unmute you may have seen my message already but just to start us off um I was going to ask a couple of questions myself um towards the end of your talk you you mentioned about doing some sensitivity tests to the reanalysis for example moving trends and so I was wondering if I could follow up on that in terms of the uncertainty from the modeling uh in terms of where you think the greatest uncertainty lies is it in the kind of the driving climate data or is it in the economics models in particular I mean links to that there was a guardian article uh just a couple of days ago where they were mentioning that in the year 2000 the is the international energy agency they said made a prediction in the year 2000 that in 20 years time by 2020 there would be a grand total of 18 gigawatts of solar energy and and already by 2007 we were installing more than that in one year so it seems that they massively massively underestimated the growth of renewables presumably they didn't see the increases in feed and tariffs and so on so I was just wondering if you could perhaps tell us a little bit more about the uncertainties of the level of modeling and perhaps what's changed in these 20 years that gives us more confidence in these projections now compared to 20 years ago um very rich questions so I'm trying to answer with uh with um my own vision also of the of the energy transition and what I have learned and also including the uh climatic variability in what I was used to think as the future of renewables on the economic side um um the the variability is the key question and disentangling the very the economic variability or uncertainty from the climatic one um is what we have attempted to make even if at the end of the day we have um somehow um proposed a different way to take all these sources of variability on the economic variability we had explicitly the integrated assessment models and the different scenarios for the penetration of renewables in particular for wind taking everything as constant and the level of the demand so we have the low medium and high electrification for the variability due to climate you are right we are played with the different sensitivities of the analysis therefore we have an impact which is indirect so which affects I would say the entire time series instead when we have the different scenario it's like truncating the source of economic variability um this was also useful for us in order to to focus somehow on different evolution of the energy sector uh then if you ask me how much does do I trust to the price series of electricity that we built I had to say that it is an exercise and that I have worked a lot on electricity markets and I'm used to say I do not work to do not want to to to work on it longer because they somehow they are unpredictable um so on the simulations that we do uh and having understood how it works from the climate side and how it works or not on the on the economic side I would say that the heavier assumptions that we make is on the economic side okay um regarding the fact that the reality is always different and that there is a momentum still a momentum for these investments well I think that the something that we do not consider here that that is the decrease in the cost uh of the uh of the uh really on the technological side is very interesting and this learning effect that is the reduction of the cost every time that we double the capacity is still extremely high uh and to me this is the most important driver today for that explains why people invest and the other is the anticipation the uh of what it is going to be the future of the energy industry okay I've got a question on the chat they've asked me to actually just read it out to you they were just asking you talked about using a stochastic model to downscale the daily data to hourly data in time but they were mentioning that when it comes to you know the reanalysis the or even the cortex uh it misses a lot of the complexity of the topography and that wind farms tend to be on hill tops where the wind speed might be a lot higher than it would be as a kind of grid box average in the climate data they were wondering if you accounted for that the fact that the you know you might actually be under optimistic if you weren't accounting for the fact that the individual location within a kind of a grid cell might be much more beneficial yeah this was a huge problem uh how to feel uh the those gaps um so the the the the choice that we have made is to um um to fill those gaps by performing different simulations uh and to to avoid being over optimistic or under optimistic we run at Monte Carlo simulations and we took the uh somehow average values uh of course they we are integrating some uh some well given that those gaps are filled by simulations um there the there is uh an impact on that uh but i would say that at the time scale that we consider and given also the uh the uh on the on the other side the need to keep precision on the uh on the price data that was like a trade-off that we wanted to solve in this way okay i've got there's one other as well which again i've been asked to read out uh this is the last question i've got if anybody else wants to send one they'd still have time to send a quick one they were just asking about the feeding tariffs and they were saying do feeding tariffs ever distort the market to such extent that generation of a certain type is installed where really it's never going to be economical without fits in other words does it can it cause market distortions in terms of choices and they were also asking that when when when feeding tariffs are designed do they account for the fact that they can actually uh can they can they for example to account for the fact that in the future you might actually get a a payback because of the fact that prices come down in time through increased capacity so windmills solar panels and so on the more you produce the lower the price becomes and therefore it might actually become cheaper and therefore in the future cheaper than coal or other sources and you'll actually get a payback over time due to the fact that you're producing a should we say a cheaper form of generation in the future even if you have to invest in it with subsidies in the present time so they are again a lot of questioning in once yeah once so feeding tariffs they are not locational specific even though for economic efficiency they should be but it is impossible since they are public subsidies they cannot be contingent to something that varies within a country so this is a problem a legal problem I would say but in terms of economic efficiency they should be um uh linked to the um for instance the the variability of the wind uh which is territorial so forget about making such a difference the second point is how do they vary in time and how there is a feedback also on the cost um uh the way in which they are set is that uh at the beginning they were really fixed over a long period when I say at the beginning I I'm for the countries that we consider they have started more or less at the beginning of the 20th century so around 2008 what 21 21st century 2008 2009 um due to the introduction of the renewable directive so the first contract were really granting a fixed price to renewables then they have started since they are very costly they are this was creating like more or less a bubble especially in photovoltaics less in wind and then they have been modified so actually in the countries where they still exist uh they the amount of the feeding tariff decreases as long as the capacity is deployed so indirectly they take into account the uh cost effect okay okay and the third point is that is why feeding are progressively being abandoned um and the premium are being introduced because they are not administrative subsidies but they do depend on the uh electricity price so if the electricity price is relatively low so it depends on its country as its own level uh then there is a top up uh if it is high then they are reminderated as the other uh convention production okay so the distortion if there is one was especially at the beginning uh when there was the the initial investment in renewables and if this distortion has led to uh installed capacity well I think that there are distortion that are much heavier uh in the energy system okay kind of like the subsidy to fossil fuels okay brilliant okay so thank you very much