 So that's the first decay, so I will try to follow the same items as others. So that's the name of the Wharf, so there is no interest about that. So I'm from the US government and I made this presentation with some colleagues. Just to remind you, to give you a global overview of the stakes, so as we know, of course, in Portugal, the world of Wharf and the societal value is very high because we have in France, in the world, 80% of the overseas trade that is by sea and 98% in the USA. There is a key role in the European defence too and 3 million of people are employed in the maritime transport sector in Europe. So for Magnus, what are the problems? So as you can see here and you have, I think, example in Portugal and we have example in other countries of Europe too. There are large stakes because, for instance, in France, we have more than 100 kilometers of Wharf. So what does it mean, the kilometers of Wharf, if you look to this one, a concrete platform of 350 meters that means that you have 1.6 kilometers of beams. So that's a very, very large. 60, 64 kilometers are built with real-force concrete platforms. So that's the main focus of the talk. And we spend quite a lot of money per year for curative maintenance mainly. So that's the case here. And USA will find that in this report. There is large money that is spent in this area too. So Sebastian spoke about wind turbines here. So they are not far from the Wharf here because it was built due to this new activity. So that's offshore wind turbines that are for the moment developed for the US market. But it will be after for the French market. So that's Naval energy here. And so the Wharf is built here, but it's now finished since six months. Here we have the Loire River. And so we had a project with several partners, industrial partners mainly. You probably know this one that's built. And there is, of course, port authorities and so forth. So how it is built. So again, there is a river, there's another bridge here. And so you have pipes, it's still. And we will focus on these platforms that is in concrete. Because the main problem is here when we have feedback from existing structures. So we are monitoring these structures. So I don't go into detail about that. And we are monitoring this area. So with two beams. So it was only to compare several sensors. And to be able to have a more precise view about the cost, about the implementation, where implementation is structured and so on. What is very typical, if you compare to the bridge, is the access. So the access for inspection, for maintenance is very, very difficult when you speak about Wharf. And of course, it's closely due to this access. So here you have typically a be-sure for inserting that. So I told already about that. In France, we make some studies in you to better know the patrimonial. And so 60% of switchers are more than 53 years old. So that's why that's a huge problem in France. The maintenance of this type of structure. You can find just to have more information about technical aspects. So there was an interactive project five years ago. That was edited by the NEC in Portugal. And so we gathered several forces here and we developed a technical guide on the web. And so it is still open about concrete, still, and about structure management, how to manage the structure. So that's a way to have more details about technical aspects. So I will not read all of this, but the main topics are the decision scenario, as you have seen today, this morning, the methods, the results, and the value of a session. So about the decision scenario. So what decision we have to do when we want to maintain a given point? So we have to inspect. So that's generally some administrative testing tools. We can't do it all. So that's SHM. So we will speak about the limited lifetime of SHM because that's not the lifetime of the structure, typically. The maintenance. So by concrete removing, here you have a typical, so it was in a project. So that's the beam. It's not in the good position. It should be the opposite. But it was removed from the port. And after, so there was an algorithm addition. And so from this little algorithm addition, we recover, replace the old concrete by a new one. So typically you have three types of metal. So the weight short grid, the forms concrete, and the manual worker. So that's preventive maintenance because you don't change the ribbons. So what we call a curative repair here is when we change the steel ribbons. So that's all the decision you can make. Of course there are all. There are all those. But that's the main decision you can have. What we have to think about, and that's why structuralism monitoring can help us, is that when you remove the concrete, you will have steel chloride inside this part of the concrete. So the concrete will go from the outside and from the inside. So after repair, the problem is more and more complex than it was at the beginning. And it's very difficult to model. And if you have only small inspection or free inspection this time, it's impossible to model. And you are hearing an illustration of how the concrete, how the chloride is evolving with the madness. So madness every 20 years at 7 centimeters depth from the surface. You can see that it's always increasing because of the chloride that is already in the concrete. So the repair is not perfect. And you have a more complex problem after repair. There are material changes and resilient chloride content. So that could be interesting to monitor. So about the methods now, we have to simulate an hidden power process. So there are some methods to simulate the chloride inside the concrete, from the type of difficulties, the temperature, the humidity, the concentration of chloride on the surface of the concrete and so on. So we can use several types of processes. What is interesting here is that you have several steps. So first step is the diffusion of chloride near the river. After you have the corrosion of the river and after you have the deterioration with large cracks. So that's what you already know probably. And so we are speaking about here on this presentation and this 2D case. We speak about the first part, the diffusion and about the primitive madness with this type of information. So when the chloride comes to reach the steel river. So we have to simulate the action after all the methods. We have to simulate the actions with their uncertainties. So that's very, very important because you see that several actions are not perfect. For instance here you have the chloride content on a given wharf. So that's the same type of beams but on the wharf. And you have that with that here. So you can see that you have a large uncertainty of the assessment of the chloride. And part of this uncertainty is due to the method of assessment. So it's due to the subjective testing, not to the interesting uncertainty in the concrete. And so on you have here an uncertainty of the threshold. So there is a threshold between diffusion and corrosion. So that's this threshold. And so there is an uncertainty. And of course if you have a rubber here at 50 millimeters depth. If you are more chloride than this threshold you will have probably a decision of corrosion. So you have to repair. If you want to have a primitive repair you have to repair before that. Okay so that's typically a model of degradation. So a model of chloride ingress in the concrete. So that's done here like a gamma process. I don't speak about that. But what is interesting when you look to supply for metoane is that you have a lot of points. And something that is interesting is that the uncertainty of the measurement has not the same probabilistic structure at the degradation process itself. So we can remove the uncertainty of the monitoring. And so the predictive state is very close after this from moving from the hidden state. Okay and the third method that I wanted to speak about is improvement of the prediction with accelerated tests. So as you have seen it's very difficult to forecast something when you have a very complex process. So what is feasible is to increase the cryopropagation in the concrete that you use during the repair or during the build. But of course you accelerate the process with something that has nothing to do with the natural process of degradation. So the acceleration rate is very difficult to compute because the process is not the same. And so what we did is to use the Bayesian network to use this type of acceleration test and the test that you have on site because we have the metoane on site. So we have the cryopropagation on site. And when you compute the factor of acceleration you are able to use this test in you to forecast the future behavior of the concrete. So that could be added as a method on the world SSEM added value. So after that about the reasons. So of course we need some assumptions. So we computed the cost for the several actions. So I just focus here on the cost of SSEM. So you have the cost of the implementation of SSEM. And after you have the cost of data treatment each year. So that's why you have the initial cost and after the cost each year if you want to treat it each year. So why we do that is to remove the uncertainty of measurement. So at least once a year we can remove better the uncertainty of measurement. Yes. Typically a type of result that we can obtain. Of course we will optimize the cost, the expected cost, with semi-districtive testing and with SSEM. And we will try to compare these two costs. So what we have to find is the time interval of inspection that reduces the total expected cost. So that's a very classical approach. So the total expected cost is the cost of inspection, repair, maintenance and failure. And what we can add is, that's due to my finger, is to find the repair threshold that reduces the time in the failure zone. So that's typically here the preventive threshold that is changing with time. And that we can optimize. So that's a repair threshold that we can optimize in due to reduce the cost. So that's one kind of output. Another kind of output is the comparison of the three types of a repair in terms of emission gas, in terms of cost and in terms of waste. And what is interesting is that depending of the time horizon or the service lifetime of the structure that you want to use. So for instance here, for civil engineering that could be 50, 75 or 100 years. Depending on that, the interest of the techniques will not be the same. Because you will not use the same number of times, given techniques depending of this service lifetime. And about the value of a recession. So that's the first result. So the last one we obtained yesterday. So it's very fresh result. So the objective of course is to compute the minimum maintenance cost with semi-dissuited testing. So that's what you have here. So that's a blue line here. That's a submission of the inspection repair corrosionization cost. And here you can see that if you inspect every seven years, you will obtain the best decision in terms of repair and you will minimize the maintenance cost over the lifetime of the structure. And after, we will use that, the same thing, but with the use of a session. So what is interesting with that session is that you will add information every year of course, and not every seven years. The problem is that the cost, the initial cost of course, is larger. So we use here, the best repair was the wedge shock culprit and five centimeters of concrete removing. And what we obtain is a gain of a reduce of cost of 80% with a session. Knowing that we have limited the service lifetime of a session itself, so without maintenance of a session itself, at 10 years. So that's very important to speak about that because we know that the service lifetime of a session is not the whole lifetime of the structure. So we have to know if we ought to limit that and first and second, if we need to more information with time and in that case, if we need to maintain the sensor of the session. Okay, and we have to ask open questions. So I have some questions for you. Is it acceptable for tech owners to support Excel engine test using the three, five, first year? So is it feasible to do that? Second, is it acceptable to pay for sensors with very limited lifetime? Now we know that sensors in concrete, their lifetime is probably between five or ten years. So do you need certification? And what type of sensor are you able to use? Should we add sensor maintenance and availability of sensors? Is there any feedback of these existing chloride sensors? So I don't know if you use some of these type of sensors in some of your circular. And is it feasible to introduce monitoring early in the design process because it's very intrusive when you do that very late in the design process. And so it costs a lot of money if you do that very late because you have to change a lot of things in the design. So is it possible for you or do you do that already to introduce monitoring early in the design process? And thank you. Anybody here from the audience, from Blizzard to Woodlife, any bridge engineer? Or I don't know. Yeah, we would be very glad. Hello, I'm from a bridge design company. So this last question is very interesting and we tried to input the structural monitoring system when the bridge is important. The scale of the bridge is quite important in the decision of the operation. But our experience is somehow we have some cases where some of these structural monitoring systems were developed in some universities. And what we can criticize a little bit what is happening sometimes is sometimes the students that are doing their thesis or so on and they are involved in these monitoring systems and then they go away from universities and nobody knows about that. There's still a long way we have to track here in Portugal and maybe in Europe to involve a little bit more of the designers in this process and try to have some feedback. Because the design of the bridge is maybe the one who knows better the behavior of the bridge or what he expected to be the behavior of the bridge and he is very interested in getting back the numbers and the behavior of the materials of the bridge. So maybe I think that I can have the areas try to involve a little bit more the designers in this process and the involvement is difficult because I know you have to have a contract and it's a long-term contract, it's always a problem otherwise it's too theoretical and a little critical. Very good, thank you very much, you created a nice piscale. Hello, my name is Carlos Piscale. I was the project manager of the Ponto of Ziria which was presented this study by Antonio in the beginning of the session. At the time I had the privilege of working with the designer, Antonio Peri Legander present here and also with the elder which is the one that implemented the munitari system. In my opinion this is your presentation is very interesting and very well formulated because I believe that things must be born at designer it's crucial that designer defines what he intends to measure and where in order to formulate because this must marry with the models of behavior at each time the measures should be matched with the theoretical models otherwise we cannot learn and we cannot act. In my opinion three things must be done. The first one is to define what to measure and when the second one is to follow always and this following means not only by the one who installed the instrumentation but specially by the designer. Design is the only one can measure and matching with the theoretical model and third that should lead to thresholders and to measures for actions, triggers for actions and which is the reason I believe we are here. We should have should read the signs of instrumentation we should interpretate, we should match with the model and that should trigger actions and in a way that can save money to us. All this must be designed in the very very early phases. Of course if we don't have another way doing this after is always better than having nothing but the right ways to start with the very first phases of the project. Thank you very much. Thank you very much for a very good comment. The three points which are very essential and we continue. Please your name. Hello my name is Vikram Akrashi and I am trying to Frank answer your question around what is the is it feasible to introduce it in the design process. I'm trying to give you the feedback from what we do in Ireland wearing my cap as a chartered engineer and not as an academic. So I've been the like design engineer in some of these projects in Ireland the method is that yes monitoring is introduced in the design process but on a case by case basis and when the consequences are high usually either when there's a large rehabilitation involved and when there was a borderline case or sometimes there's an immediate rehabilitation or repair needed. In each case the method is to employ the design engineer led by a chartered engineer which is a must and then the design engineer exactly like my like previous commentator on this and the design engineer decides where to measure what to measure etc. But does not take the decision on sensors what he said. So the maintainability sensors they even their calibration they have to be ratified by the lab. So it's never given to a university. The university is too expensive and not cost effective. So we get a lot of money and so what we do is as engineers we hire companies either from Northern Ireland or from Ireland whose job is to only guarantee that yes this is the right sample is correct or this say half cell furniture thing that we're doing the process is correct but these subcontractors only guarantee those results. They take no responsibility for their interpretation where they will measure it and why they're going to do it. That entire responsibility then lies on the engineer side. So this is how it's done in Ireland. Thank you very much for your comment. Yes I think proportionality is very important. For most important thing you do more work. For less important thing you do less work. Proportionality as a criterion for effort is a very important one. Any other we have already three discussions from the audience and we appreciate it very much from the local people. Are there any other questions or discussion points? We can yes. I was just wondering if there's any support by the regulations in Ireland actually to take benefit of the results to support let's say the case of continued operation or change of operation. So there hasn't been any existing project where continued operations has been done except for in one case I remember I think for Ferry Carrick Bridge there was continued monitoring and things like that. So there's no regulation that prevents it but there's no case right now. But how do you argue in that case that monitoring can provide the basis for let's say continued operation or for operation despite observed damages or whatever. So how can you actually argue for your case of monitoring if it's not a part of the regulation? Because in some cases I'll give you a concrete example here. So one of the cases where the county council did decide to go for monitoring was when there was significant uncertainty about what has happened to the system. So there was this bridge and a heavy truck was going under the bridge and I was an engineer of that project and one of these hydraulic arm it came up and hit the bridge upside down like this and destroyed the entire clean stress peak. Now the thing is if you think that complete redistribution of stress has taken place nothing has actually happened to the bridge. You can ignore that. If you think that no redistribution of stress has taken place then that bridge has collapsed as it stands. So the reality is somewhere in between and our models considered all these situations, whatever situations but the county council prescribed monitoring throughout the process while we were designing the recommendations on what to do while the repair was done and while further strength gain was done on the bridge for 24 or 14 days each minute so that there were limits that were decided on if it goes beyond this then we take this model if you go beyond this model and there was also one case where it goes beyond this then we're probably like we're done for and called insurance but these are some of the situations where they consider this. That's a positive story. A negative story is there were health monitoring devices for the Malahide bridge in Ireland where despite monitoring there was very little movement that was observed on the bridge and a day later that bridge fell into the sea just after a train went over from Dublin to Belfast and so that is one of the examples that's sometimes taken as a criticism in Ireland. Well despite this monitoring there has been very little movement the train went past, the whole thing went into the sea but I think there we have to be mindful of what is the mechanism of failure. In our case elasticity took a big part or let's say deformable media took part. In the case of Malahide bridge essentially what happened was the bridge was like this the water washes away one of the piers and the thing falls down. I don't expect any health monitoring system to pick up very early that kind of washing away. Well if it's monitoring the skull then it might be able to do it. Yeah but not on the bridge. It depends on what you're monitoring. But I'm just wondering because in Ireland you're also using the Eurocode system right? Now yes. And in the Eurocode system there's no probation opening for taking benefit of non-stop fees. No this was the assault. This was the heart of the safety context. This was before Eurocode. This is a little bit an issue in my opinion. This is a big issue. This is a very very big issue because like I was an engineer in that company in this project in 2009 that was just when the Eurocodes were coming in and we were doing our national like essentially like adapting the Eurocodes for Ireland's numbers and things like that. But for assessment we were still using BS codes. And after that let's say like a full scale like monitoring hasn't been done in Ireland. So we are really talking about handful. I could count the number of projects in one hand. But at least there is a legal framework through which this can be done and the Council can accept it. With the Eurocode it's a big challenge that the whole assessment monitoring site it's not there. So what do we do? I think the actual the solemn opportunity with the Eurocode system is the let's say the opening through design by testing. That is probably the only opening we have at the present time but we are undergoing the visions of course right now on the Eurocode system. And we are from the very beginning pressing and suggesting clauses on how to take benefit of non-structural means including specifically structural health monitoring. But we are experiencing a significant reluctance to that idea in the more important Eurocode committees. We are still fighting. The battle is not lost. But I think that the organizations they also have a role to play in coal emissions and they should make their voices heard. So this is just to flag the possibility to encourage you owners to make your voices heard in a way with that. It's really really important. Yeah I think very very interesting. Yes yes everybody. But it's getting very interesting so we continue with your remark and either we can also have an event. We have one hour. So just follow up question to Mike really and the theme seems to be because from your previous comments about this reluctance and so the situation with the codes say the situation where tons of structures that do exist do not satisfy current codes. It's a reluctance to allow the provisions what say SHM could be explicitly acknowledged in the code. And you say you have problems with sort of people from the Eurocode committees telling about being reluctant but what exactly do they say? Do they give any argument or is this design code to not include anything but structural means? Yeah but why? Is it just sort of ideology or what? I think this is very general. Very important I have to say. And I also agree that you have to put new things which you have to discuss with designers and you have to make things like that that they should be applicable. And you have to educate. It's very important and we are doing it. We will keep this comment if you allow me because there are some others. But I will come back. Sebastian we will come back to this comment what you do in the instandars and we have a session tomorrow. And I would like the second presenter to come because we exceeded already