 Good morning ladies and gentlemen, I'm very glad to be here in Lisbon and to have the opportunity to let's say to contribute to the project and especially we have here a case study which is not the best case study for Lisbon. There is no snow and there are stadiums of course but you are working in a highway authority and you are more interested in bridges or tunnels. So this is another type of case study but I hope it will be interesting for you and it's a case study. We have agreed upon how to present the case study so that we be first the decision scenario, what is important to decide and what is the asset description and what is the monitoring strategy. Then the methods which have been applied but very briefly, I don't go to 30 days, I just mentioned them, some results and of course what is the critical thing is the value of information for the owner and open questions because we are very interested into a kind of feedback from the industry and discussion. So the decision scenario, if you are in Central Europe or also in Northern Europe you have a lot of snow and there have been damages in industrial buildings and the snow load, ice arenas also and we have here a stadium roof which does not comply with the requirements in the Eurocodes. So if you try to verify the basic components of the structure they cannot fulfill the new Eurocodes because the snow load is higher than the originally snow load which has been used in the design. So the snow load dominates the reliability, it's a light roof and if you have a light roof the snow load is the highest load. So what you can do here it's a different part. Usually if you think of monitoring you monitor a bridge and you monitor a parameter which influences the resistance of your structure. We heard about fatigue, we heard about corrosion, you can do also other things, major deflections and so on. So you are more thinking in the resistance but the safety, the reliability of your structure is of course depending also on the actions and here you can directly monitor the snow load or the parameters affecting the snow load. So the difference here is we have, we can monitor a parameter which is directly affecting the reliability of the structure. When a specified limit value of the monitor parameter is exceeded either that you can have actions, either you can remove the snow load which is also expensive or you can close the starting for the event and wait. So there have been some damages on the snow load as I said. So the asset description first we have to, if we are, let's say if we want to take some decision we have to take all the information as I said before from the design state. So we have a starting which is constructed in the beginning of the 90s. It is 190 meters above sea level in northern Italy. It has, it is for 4,000 spectators so you can classify the Euro codes provide some classifications. They are similar in other standards. So CC3 means consequence class 3 is the highest classification with the highest consequences in case of failure. This is the simple system. It's a cantilever beam and we have spacing between the beams every five meters with connected stiffened members so you can also do some system analysis, not member analysis but this is not so important for now. We have done that. So you see here the whole code was 0.9 km per square meter and the new code has about 30% higher design, higher characteristic value for the snow load. If you look at the design requirements if you check the structure you might see that the resistance is about 90% of the resistance provided in the Euro code and if you try to verify these structures you will see that you do not fulfill the current standard the Euro code. So you can do, you can measure the snow load and there are different possibilities. You don't even need to measure on the roof. You have stations close to your starting and you can take the snow load on ground however you have to relay the snow load on the ground or the snow depth and the snow load they measure both. They measure also the density, density so you should imagine you have the depth, the snow depth multiplied with the density and the density has a very high variability. It can be very small to close to water so it's too high so it's can be very very very heavy and the uncertainty of course is high if you just use the meteorological station. There are also systems where you can measure the snow depth and you can have also some cost to buy this system and also to maintain the system for some years and then you can have also their systems the most advanced where you can directly measure the snow load on the roof and you can have an alert if it is higher than a certain snow load which is let's say close to your acceptable risk level you get an alert and you have to decide what to do. So the methods which have been applied here are uncertainty modeling you have the uncertainty snow load, you have predictions for a couple of days so you can use them also you can assign also some uncertainty there. You have uncertainty also in your resistance parameters but they are not as high as in the road parameters as I mentioned in the beginning. You have reliability analysis so you can compute the safety of your component or of your system and the safety is reflected in the so-called reliability index. The higher the reliability index the safer the structure so a typical value can be four or five. Five means it is safer or more reliable your component compared to four. Then you can this is an important part what can happen if you have failure of course it can happen where nobody is there but it could also happen where people are there so you have to look you have economical costs you have also human consequences and then you can also apply to combine this a kind of risk assessment to combine probability that something will happen together with associated consequences and as I mentioned before we have a kind of measure of reliability which is the reliability index and if you assign a so-called target reliability index which reflects the acceptability criteria of risk and I said values typical are from three to five. This would be in practice depend on the costs on the investment into cost of safety the higher the investments into cost of safety and also the higher the costs are then the lower the the targeting ability would be and this is in practice done also in the highway where you are working for example imagine a tunnel a big tunnel then you would not construct in the tunnel an emergency lane because it's very expensive you have to construct a larger tunnel so it is a typical reflection of this principle which combines the investments into safety with the consequences so if the investments these are very small numbers because the failure consequences are really really high so if the investment into safety gets higher the target reliability would get lower. The eurocodes currently they say for CCC structures you should use 5.2 which is a real a really high value and you can go lower if you apply other codes or standards or guidelines so there are more let's say more than guidelines like the ISO 2394 which reflects these aspects cost related to investments into safety and cost of failure which is something you are applying in practice we are applying it in bridges we are applying it in tunnels but you don't have a direct reflection on how you apply it so this is another parameter here which is important what to do with existing structures which do not fulfill current codes and how to deal with cost problem so this is in case of cost of exceedance of the limit so we have a parameter which is here the performance parameter it is the snow load but it is directly related to the reliability so therefore is this case study very simple you have the snow load everybody can understand the higher the snow load the worst is the situation for the safety of the study and you can directly reflect it or relay it to the inspector to the reliability target reliability and you have two costs the one is the cost of safety measures clean the study don't do the event which of course is you lose money and failure cost and here you have to relate failure cost to some monetary value which is possible to be done because you have human you have fatalities injuries and you have also economical economic damage here business losses and as I mentioned in the beginning you have three monitoring strategies with three different uncertainties related to these monitoring strategies monitoring strategy number one you have you are monitoring which is actually done without any costs because you have meteorological stations close to the standard so you are monitoring the snow on the ground you can monitor it then this is a one you can monitor this is very cheap you can monitor the snow depth but you have the uncertainty on the snow density you can monitor the snow depth on the roof and you also you can monitor it directly with modern devices there's no load directly which has a very good precision of course there is always some uncertainty and now you have to look at the costs how how what are your costs of the monitoring systems so you have the acquisition cost you have to buy you have the operational cost and costs when the limit and cost of safety measures when the limit is exceeded so you have also to see how many times during the residual lifetime of this study or of your structure the limit will be exceeded so this will be one or two times in the next 20 years for example because you have experience and we did that for a reference period of 20 years here so you can see for your the results which are obtained again you see the the reliability which is covered depending on the measured ground snow load here the snow depth and the snow load on the roof so the higher the load is of course the lower the reliability which is covered by your structure or existing structure so the monitor values versus target reliability is here a typical result you can apply them also here the framework which was explained before you have the plan investigations which are here done you have the results of the investigation which is the measured parameter there you have the decisions close the study clean the study do nothing three possible decisions and then you can see after all what are related in each three cases costs which are involved in these possibilities here is a typical now we we have a typical result sorry this is a typical previous measurements which you have so you see as soon as you get to the limit you have to clean the study and then you will have to clean it maybe a second time in a couple of years so what is interesting here is what is the benefit of this let's say somehow simple structural health monitoring which is done through that load it provides limit values for loads and corresponding which are corresponding to some risk acceptability criteria which you have to put in your structure it allows for real-time evaluation so you have a kind of reliability online it supports decision regarding to safety measures so you can apply all these ideas investment into safety failure costs and so on it can lead to a more economical solution compared to an expensive upgrade because it will be very expensive more is a structure to upgrade to fulfill the miracle it indicates also the optimal strategy so it has been found that direct let's say the direct measuring of the snow load is the best option the most economical one and also what is important for me it increases the research knowledge so you try you you develop new standards you have a better reputation also research it is reflected also in future applications so this is also one benefit which cannot be directly quantified into money but it it is of an essential issue for me and also as I said to the reputation if you do this you get as a company or as an or a better reputation so these are you have let's say some hard benefits and some soft benefits which are indirect benefits cannot be easily quantified into money but they are there and assist development of standards as I said before so the open questions I'm finishing here for as we discussed we have to my open question our open questions here is what to do with existing structures which are not fulfilling current standards it's a typical problem in many countries how can we use value of information in practice and of course very open question is how to estimate the cost of failure in case something is happening in case we have a collapse of the structure thank you very much for your attention nobody else wants to say anything then let me start so normally when we when we when we are trying to manage the safety of structures we are addressing two types of criteria and and the same also applies for the criteria provided in ISO 2394 namely criteria directly derived from cost efficiency of structures and criteria related to safety for people exposed to the performance of structures so in this study here I think you started out with some criteria from the euro quotes with very high levels of of features and then you you were pointing at other quotes and guidelines so with respect to the actual type of criteria life safety or cost efficiency what what did you choose here in your study yeah we we choose the criteria of optimization of economy so the relation between costs of failure to cost of investment so cost of failure you have to quantify so we have let's say we have to put some monetary value value to the cost of human life and we did that in terms of the LQI principle so life quality so we have optimization and human safety which have to be fulfilled here at the end the result was that as a reliability of 3.7 would color both and it is lower compared to the 5.2 which you have to fulfill in the euro code yeah similar problems and maybe another type of of approach to a question now we have we have decision makers here so in the case let's say what was the starting point for this this case study or this study all together because you could imagine that there would be many structures out there that have been designed for according to an old code and then things have changed in the code developments and the structures are not fully complying the most recent codes now and and all these structures are still standing there right until professor Diaventidis comes and say that maybe there's something to look at or somebody else gets the idea that maybe we need we need to consider whether the structures are still up to present standards the initiator of such a study is this kind of important and the regulation around it also because if there's no duty or responsibility for for for the building owners specifically to keep track on such changes then it may actually be better for the building owner for the decision maker not to do anything because the only thing that will come out of that is expenses right did you think about that uh to the first question which is very interesting is how did we choose this example got all the examples were bridges so I was and I'm very glad that Sebastian and Michael involved us into this project uh we thought about something which is understandable I was listening to many it's very complicated for bridges you have these performance indicators there was a whole session on performance indicators so I was a little bit thinking what is a performance so the first thing we tried to present some uh let's say case study in which the reliability is directly related to which the value which you are measuring so you don't have some complicated transformation from stiffness of bridges to the safety of the bridge and so on and the second thing is this problem is in all countries we have a lot of existing structures which do not fulfill the codes and the situation is that according to my opinion standardization committees and they are pretending you have to fulfill the euro codes and so on even if you don't fulfill so you need some clear some clear guidelines what to do in that case because 50% of the structures are existing structures of the budget standard take for example the tunnels after that's that we had in the tunnels the directive changed the 2004 directive and you have to prevent much more uh stricter criteria so half of the tunnels do not fulfill this criteria the european community gives time until 2020 to fulfill the criteria or to put other safety measures which are somehow equivalent to those new safety measures so there are guidelines which are trying to reflect this but not as strong as the reality is according to my experience so there is in fact needed to use this information yeah we need to stop this interesting discussion unfortunately uh but but i think uh this small discussion after the presentation points that uh two two things in my opinion so responsible building owners responsible structural system owners is a very very important point here and the other point is the regulations regulations actually specifying that some minimum monitoring has to be undertaken is necessary so imagine the case that your structure actually fulfills the most recent code requirements right the design loads were exactly according to what the code is prescribing but then we had a super heavy storm fall and it just exceeds the let's say the assumed conditions at the point of design does that mean we shouldn't do anything if it's completely obvious you had a roof the four meters of snow uh but we can no but but there has to be some sort of legal fix uh in the codes which just implement sound reasoning right that when you see things which are above clearly above design criteria then you need to do something so responsible owners but also codes which directly and explicitly incorporate structural health monitoring at at least a minimum level