CO2-reduced cement and sensor technology

[responsivevoice_button  buttontext=”Play Transcript”]

Johannes: Welcome to Concretely!

I am your host, Johannes Lohner, and I talk with experts from various fields about the maintenance of our structures. Today, I am speaking with Professor Ueli Angst, one of the world’s most renowned researchers in the field of the durability of building materials. We will discuss CO2-reduced cements and the monitoring of corrosion processes in concrete.

Before we talk to Ueli, first a small digression: As mentioned in the last episode, cement is responsible for 8% of global CO2 emissions. Why is that? Approximately 80 to 90% of the raw materials in cement production is limestone, chemically CaCO2. At temperatures over 1400 degrees, temperatures that are not easily achieved with renewable energies, CaCO2 becomes CaO and CO2, which is then released into the atmosphere. Therefore, there are many efforts to reduce the CO2 emissions from cement. This can be:

– reducing the cement content in concrete, which usually means weaker concrete

– concreting less or building more filigree

– one could use CO2 capturing, but the problem here is the amount of CO2 produced and that only a part of it can be extracted from the atmosphere

– and most efforts are aimed at replacing the cement: this can be done with slag from steel or iron production or fly ash from coal production. Both are extremely fine materials that ultimately change the properties of the concrete. For example, the electrical resistance, frost-thaw salt resistance, and also the pH value. This can influence durability. By the way, there are also other approaches like the LC3 cement, which we will also discuss in another episode.

Let’s talk a bit about the durability of concrete. Last time I also mentioned that corrosion is the main cause of concrete decay. According to a study by the British Cement Association from 1997, it was found that in 3/4 of the cases, corrosion was the cause of premature decay in bridges. In concrete, there are reinforcing bars that take over the tensile force but also ensure crack distribution. Due to the high pH value of 12.5 in concrete, which is created by the cement, an oxide layer forms around the reinforcing steel and protects it from corrosion. It is also said that the reinforcing steel is passivated. There are mainly two factors that can lift this oxide layer and lead to corrosion.

– one is chloride ingress: this is when chloride ions, mostly from road salt in winter service, enter the concrete through liquid water or through mist from traffic. As soon as they reach the reinforcing bars, the chloride ions, the passive layer is locally lifted, and rapid and often invisible corrosion occurs.

– the other factor is carbonation: this is about the reduction of the pH value of the concrete by CO2 that enters in gaseous form from the atmosphere. This is especially the case in warmer regions or industrial areas.

But when exactly corrosion occurs is much more complex and depends on the pH value and the steel potential, as well as the transition zone between the reinforcing steel and the concrete. A central role here is played by the pore solution of the concrete. That means which and how many ions are present. A topic on which Professor Ueli Angst and his research group are working, among others. His research could help one day to detect the indication of reinforcement corrosion earlier.

Johannes: Warm welcome, Ueli

Ueli: Yes, thank you.

Johannes: The first question I have: these more sustainable, CO2-reduced concretes usually perform worse regarding carbonation resistance. Why is that?

Ueli: Yes, that’s correct. The modern CO2-reduced cements have a reduced content of a mineral called portlandite. This mineral is also responsible for the high CO2 emissions in the production of cement, and therefore many of these CO2-poor cements simply have a lower content of this portlandite. And this leads to the concrete made from these cements having a lower capacity to buffer the pH value in the alkaline range. Thus, there is a faster pH value reduction when this concrete is exposed to CO2, leading to faster carbonation.

Johannes: So that means we now have a problem regarding durability. Can something be done about this, or should something be examined more closely when introducing it?

Ueli: Yes, it is perhaps also important to emphasize that durability is not necessarily bad. Carbonation itself is not directly a durability problem. Carbonation is feared because it can trigger the corrosion of the reinforcement. Because the corrosion of the reinforcement occurs at a lowered pH value rather than in an alkaline environment. Of course, that is the case, it has been seen in various cases, but carbonation is not the only parameter that favors corrosion, and therefore it is central that we shift the focus a bit from carbonation to corrosion. Up to now, it has been simply assumed: carbonation leads to corrosion. However, there are countless cases from practice that show that corrosion in carbonated concrete does not necessarily occur. Therefore, we must also focus on what actually promotes corrosion in the end, and how it can be prevented even in carbonated concrete. Factors such as moisture in the concrete, the concrete structure, etc., play central roles. In this regard, there is still much to be done in the field of science. Particularly, the understanding of moisture transport in concrete under certain, e.g., cyclic exposure conditions and how this moisture influences corrosion are questions that still need to be clarified. This can certainly be done in the lab or with numerical models, but one option that could be very informative here is the use of sensors that can be installed in actual structures. For example, structures made with these new low-CO2 cements. Because with these sensors, it is relatively quick to see how these materials actually behave in reality, which may not always be the same as rapid tests in the lab suggest. Therefore, I see opportunities here also with the use of sensors to get a better handle on the behavior of these new materials.

Johannes: I would like to go into more detail about the sensors. But just before that, that means we need to delve deeper into carbonation in relation to corrosion and change the focus. So, more towards moisture, pore distribution, etc. Does this mean that the current standards regarding carbonation resistance need to be adjusted?

Ueli: Yes, undoubtedly. Otherwise, we have a conflict of goals. You cannot simultaneously try to reduce CO2 emissions and then also maintain high carbonation resistance. That is simply contradictory. Therefore, we will have to adjust the standards. However, we currently still lack enough scientific basis to do this, and therefore more scientific research efforts are urgently needed.

Johannes: Okay, then as I said, let’s continue with the sensors. A classic example is their use at the end of the lifespan of structures. For instance, if the condition deteriorates unexpectedly, so that we can extend the lifespan by placing the bridge or the component into the next maintenance phase. Or if immediate measures are necessary, to more precisely review or monitor their effectiveness. From various sources, I have heard that it is difficult to make statements about the entire component or structure from sensor data, which are very locally collected. Is this point of criticism justified, and what should be considered when reading the data?

Ueli: Yes, this point of criticism is of course justified. However, it should be considered that the issue of very local data is a general one in engineering. And not only with sensors but also with conventional established testing methods. Consider, for example, the extraction of core samples from a structure. Here too, one gets extremely local information. For example, when determining the compressive strength, chloride content, or carbonation depth at such core samples. Even then, you have at best 5 to 10 local individual values from a structure, and the question of how to extrapolate this or transfer it to the entire structure arises just as much with these conventional methods as with the sensors. For transferring these local data to the entire structure, you need contextual knowledge and also sound material technological knowledge. An understanding of the behavior of the structural framework. And with this knowledge, professionals must choose the location of sample collection or the location of sensor deployment very carefully. Because only in this way can one obtain goal-oriented information. And this applies to both sensors and conventional measuring methods. So, this is actually nothing new at all, and this point of criticism does not only concern sensors. There may be a few methods that can indeed provide area-wide information, such as potential measurement as a non-destructive testing method for detecting corrosion, or of course, the conventional visual, image-based inspection, which provides information that can cover larger areas. But of course, these methods also have their disadvantages. They are either very complex, for example, potential measurement, or limited in their informative value, like visual inspection, which is not able to detect damage early. So, the quintessence: there is not one solution that can solve the problem entirely, but in the end, one must naturally combine various methods sensibly. And this includes local methods as well as those that can read area-wide information.

Johannes: This means that engineers and maintenance planners are more challenged. They must understand the deployment correctly.

Ueli: Exactly.

Johannes: One more question I have is about periodic data collection. To make an overall assessment of a structure, many influencing factors come into play. And each structure is unique. Do you think in the future it will be impossible to make a more accurate prediction about the durability of a structure using technical and digital tools, and if so, how could this be done?

Ueli: Yes, that is correct, the overall assessment of structures is naturally extremely complex. There will certainly be even better possibilities in the future, and we are also dependent on them,

 which must be said quite clearly. We urgently need better, sharper diagnoses of the state of the structures of which we have so many. We have countless aging structures and we need to better diagnose when and to what extent maintenance measures will be necessary.

So, what can we expect in the future? How can we become better and more precise in diagnostics? On the one hand, there are constant developments in the field of non-destructive testing methods, abbreviated NDT. Especially in combination with robotics and drones, one can automate the application of NDT methods more efficiently and also improve the reproducibility compared to manual application. Which could, of course, also reduce this “human bias”, that is, the subjectivity of professionals to some extent. So, in the context of the digitalization of condition assessment, there is natural potential, opportunities to get better, more precise, but here too, there is still much to do about how to arrive at a statement from all these measurement data in the end.

Johannes: If you were to periodically capture data and see from the past how the component deteriorates. Are there works or studies to make statistical predictions about the future?

Ueli: Yes, that’s exactly what’s still a bit missing. Time series of inspection data are of course available. The problem is that these measurement data often lie in completely different data formats. There are not always standardized procedures for how to do the reporting of condition capture/assessment. It is usually presented in the form of reports, e.g., text-based or plans, and partly also data. And bringing this together so that it is also comparable over time is one of the challenges. So, I think in the end, we need workflows for how to arrive at a statement, an assessment from these measurement data. And how to then also compare them over time and make an extrapolation or a prediction into the future from that. We recently had a dissertation at ETH Zurich that took a first step in this direction, specifically the integration of data from non-destructive testing methods that are taken with a drone at the object. The integration of such data also with destructive testing methods and then also with physically-chemically based models that allow predicting damage processes into the future. But these are the first steps, as I said, there is still much to do in this respect, and I believe the international research community is also strongly challenged.

Johannes: Very interesting. Thank you for being here, and I hope I can find similar conversations in the future.

Ueli: Of course, very happy to do so.

Johannes: Then, see you soon.

Ueli: Thanks, goodbye.

Johannes: Thank you for listening. Please share your opinion in the comment section of the episode on my website www.concrete-ly.com. Feel free to also leave personal experiences or complaints. Please “like” and share the podcast with your girlfriend and colleagues, also via Spotify or Apple Music. By the way, the episodes are also translated into English on the website. [/responsivevoice]