How to monitor water quality?

Welcome to the Urbanista blog where we discuss water management challenges of Nordic cities. From safe drinking water distribution and stormwater collection, to building sustainable urban living environments. The Urbanista blog is based on the Urbanista podcast episodes. This post is based on the interview with Patryk Wójtowicz – Research Manager at Savonia University of Applied Sciences, and Water Network Modelling Expert in the field of environmental engineering. 

What is the Kuopio Water Cluster?

- We formed the KWC together with other founding members. Everything started from the idea that if you want to be serious in your research, and you want to make sure that academia and companies get closer together, you have to always have bearings toward the actual product. So, in the cluster we have a public university, The University of Eastern Finland, as well as other public institutions such as The Geological Survey of Finland (GTK) and other relevant actors. We combined them together so that it would give us the power of more than 150 experts.
The cluster’s role is to make sure that the ideas that are starting in academia can be converted into real products. You have to make really hard decisions at the low-tier level and make sure that only the really good ideas that will result in the product, will go through. Having the cluster enables us to involve our members at a very early stage. We stay neutral because the role is not to make business, the goal is to support innovation and get it to the market.

As an industry, we can be reluctant to talk with each other. How did you bring these companies into the cluster?

- In our case, we were capitalising on more than twenty years of history, and the founding members were already cooperating, in loose terms, in different projects. Then when we started thinking about how to engage the companies, realistically, with most of the members we had some kind of a relationship. Either we had done a service for them, or we had used their products. This allowed us to engage the companies much more easily and matchmake them to their needs. If they have an R&D process going on, we can involve them in a project. If they need internationalisation, or to scope a new market, we can partner them with our partners to make it happen. If they want to get some feedback from water utilities, we have those associations with us as well. Gradually we were building the membership, which is invitation only, and dependent on the requirement that we would like to have some inside knowledge about your product. That’s how we started.

Can you tell us more about the research you are doing on drinking water quality?

- In our water lab. there is a water loop which is a distribution system simulator. It’s a scaled down version of a real water network which includes every component a water network should have, i.e., our own water intake from a nearby lake, a one kilometre pipe to a first stage pumping station which pumps to the laboratory where there are three compartments where we can store different types of water. We also have our own containerised water treatment plant where we can continuously produce any kind of water necessary.
Going back to the water loop itself, it is fully automated to create different situations, and it is full of sensors measuring close to 100 parameters simultaneously. It’s an overkill of sensors in one place. I call it a data factory. It’s a lot of data being collected, not only values but also states: states of the pumps, the actuators, etc. Everything is very precisely recorded in our database, so we know exactly what is happening. What’s important to us is the meta-data, the data that describes the data, because if the companies are innovating on software solutions, we can provide this data need in an organised format that can be ingested into any algorithm. Soon we will have 5G and even 6G to help manage even greater amounts of data.
In a very compact time frame, we can recreate any kind of scenario that could happen in a real network, including even a cross-contamination. This is when wastewater is mixed slowly with potable water. These kinds of events happen, probably more often than we think. We can simulate or recreate any scenario to order. For example, if we want to simulate what would happen if we reinstalled a pipe without the correct procedure, what would the consequences be? With the sensors in our laboratory, you can see exactly how the sensors are responding.
 

Then you can look at the long list of potential ‘what if’ scenarios?

- Yes. And we can look at both the long term and short term effects. We can even control the temperature going down to five degrees Celsius. We can change the pressure and operate the system at up to ten bars. We can introduce surge pressure into the network. So it’s not only chemical and microbiological contamination but also hydraulic contamination.

You are collecting a lot of data. Are you creating your own artificial intelligence (AI)?

- Of course, we are a university, we have an ICT department and we are building a data repository which is open for everyone to cooperate with us. We are very happy with non-commercial use. We give it out to any research team who are interested in the challenge of taking the existing data and building an awareness system.
We know there are many companies trying to make an early warning system, but can these actually tell what happens, or do we just know there is a fault in the network? I believe that if we have enough information we can actually interpret what is happening and explain it to a non-expert operator. We want an autonomous system that can tell you, “This pH sensor is drifting, but it’s not a fault. Just clean the sensor”.
I mentioned the open data repositories. I believe we have to keep this data open and available, even for companies, to build new innovations on top of it. That’s what we want to do. We are experimenting with block chain to ensure the ownership and integrity of the data. The concept is similar to that of NFTs, but for us it is also about integrity of the data.
We do a lot of things that haven’t yet been adopted by the water utilities. Everyone knows that they can be quite conservative, but now we have a big push because of what is happening in the world, and because of increased digitalisation.

How open have the water companies been toward your ideas of increased digitalisation and technology?

- Water utilities have to follow the rules and regulations. Their critical infrastructure is not about risk-taking. They have to keep the integrity of their operations and that is why they are understandably wary about making any abrupt changes to what was working ten, twenty or even thirty years ago. The change must be very gradual. In many cases you don’t have the opportunity to test-run a system on the side, to see how it works because at the end of the day you always have to think about a timely delivery of the water to consumers.
But water utilities have a commitment to making savings. Wastewater treatment uses notoriously large amounts of energy and companies have a water footprint that they want to reduce. Digitalisation allows the bill to be reduced; they won’t need so much manpower in the future and it also keeps the costs at a reasonable level. It’s not about automation, but more about making the process more efficient. For example, we don’t have this mundane watching of screens, waiting for something to happen. We have an ambitious plan that, especially in the case of potable water, 100% of the events that happen in the water network, should be detected by sensors. Currently it’s about 50/50.  
We shouldn’t forget about wastewater or stormwater. Climate change is bringing a lot of challenges for us in the future. In Finland the forecast is for an increase in the occurrence of torrential rain and flash floods and you need to be ready for it.

 You mention stormwater. Are you also testing with this in your labs?

- Our lab doesn’t specialise in stormwater because that concerns different factors and values, but we are taking steps to work with stormwater. The first step will be to put sensors in the stormwater network, and from that look for new solutions in the area. In Savonia I was involved in a project to clean or pre-treat stormwater before it is discharged into a body of water. The places in Finland where the snow in cities is collected in the winter. You can imagine that when the snow is collected into one place, it melts and then starts to leak into the watercourses. Actually, now in Helsinki it is forbidden to dump snow into the sea.
And in Kuopio we are setting up a kind of barrier that will take out all the harmful substances, like heavy metals from the roads and nutrients, like nitrogen and phosphorus.
The next case relates to industrial stormwater, which is washing off elements that are stored in an open space and drain into the stormwater.
The final case was focused on the city centre where we were testing the regular composition of the stormwater.
 

When do you anticipate 100% of the events will be tracked by sensors?

- As soon as possible for our own good. We need first some breakthroughs. There is a technological gap which we are really looking forward to fill. Of course, it has to be cheap, communicate easily, doesn’t need direct power consumption, but for that we need some breakthroughs in the battery industry. If we are over optimistic, perhaps this will happen in the next five years, but it might be reasonable to expect 100% within ten years.