At Electricity Transformation Canada 23, Icetek’s André Bégin-Drolet explains their thermodynamic icing sensor that detects onset and intensity. The technology optimizes turbine operations to reduce downtime and damage while improving grid reliability. Icetek provides expertise and data analysis services alongside the product for maximum value.

Check out IceTek at icetek.ca

And follow their LinkedIn

Sign up now for Uptime Tech News, our weekly email update on all things wind technology. This episode is sponsored by Weather Guard Lightning Tech. Learn more about Weather Guard’s StrikeTape Wind Turbine LPS retrofit. Follow the show on Facebook, YouTube, Twitter, Linkedin and visit Weather Guard on the web. And subscribe to Rosemary Barnes’ YouTube channel here. Have a question we can answer on the show? Email us!

IceTek

Allen Hall: At Electricity Transformation Canada 2023, we’re here to talk ice detection with André Bégin Drolet with Icetek. And icing. Welcome to the program.

André Bégin-Drolet: Thank you, thank you for having me.

Allen Hall: So Icetek is a new ice detection system that I was first introduced to by Borealis Wind. And Daniela said we got this new ice detector and it’s fabulous.

It tells us all these great, wonderful things about ice that we never knew before and I had never heard of it. Which was odd, because we live in a place where there’s a lot of snow and ice. I usually hear about ice detection. It’s a thing that happens. But Icetek is a relatively new company based in Quebec.

André Bégin-Drolet: Yeah, exactly. So it’s we started the company in 2020. So that’s three years from now. But it’s a spin off from a university project. We, I’m a mechanical engineering. Professor Laval University in Quebec City. We developed the sensor throughout the research for the last 15 years. So we did a lot of research, academic research.

It was a tool for us to understand icing on wind turbines. And then I started a partnership with Daniela a research partnership with them to help them be be better. And this is where it it all started, where… After the project, she asked, can we buy those sensor? They were not for sale ’cause it were a research product, at the moment. And then, yeah, this is when the university encouraged us to to go and start a spinoff company for that.

Allen Hall: Because the problem is not just knowing that there’s ice. The problem is trying to know that ice is coming.

That’s the trick. And a lot of the ice detectors that are out there are really binary. That ice is here. Ice is left, but in an operational sense, in a wind turbine, it doesn’t really help you all that much. Leads to a lot of downtime. Yeah,

André Bégin-Drolet: so ice is a very complicated problem. Ice can take different incarnations, freezing rain, blaze ice, rime ice, ore frost different types under different conditions, and we learned that through our 15 years of research that it can take different Incarnation and we designed the sensor so that we could know when it start.

So the really onset of icing when there’s no icing visible, but the conditions are prone to icing. And then what’s the intensity of icing? What’s the amount of liquid work content in the atmosphere when it stops the meteorological icing. ’cause when the meteorological icing is over, you can still have ice on the structure.

Is still, is this still icing? Yes, but it’s called instrumental icing, persistence of icing. So all these different phase of the icing, you need to understand them. And as you mentioned, it’s not a binary.

Allen Hall: No, definitely isn’t. And I know Daniela trying to explain that to me several months ago. And it just went, there was a lot going on there.

So I’m glad we have time to sit down and discuss it. Okay. Let’s just walk through what the sensor is. Because it looks different than any other icing sensor that I’ve ever seen. It’s a, it’s a metallic cylinder. Yeah.

André Bégin-Drolet: We’re using a thermodynamic approach. Okay. We’re having a heated cylinder, that we know the amount of heat that we fed into that cylinder.

We also know the surface temperature of that cylinder, because we have those temperature probes inside the cylinder, but close to the walls. And we also measure the air speed, the air temperature, relative humidity solar radiation. Based on all these parameters, we do modeling, what should be the surface temperature, knowing all those parameters.

And then we compare.

Allen Hall: So it’s more than just detection of ice itself, you’re detecting… Basically, temperature, the amount of heat pulled off the cylindrical sensor, you have a solar sensor. This is really fascinating because when I saw the sensor the first time, I wasn’t sure what was the magic piece here.

But it’s more than just one sensor. In order to do this calculation, you need actually multiple sensors. So you have temperature, true temperature, true air temperature. You have the sun condition, sun out, no sun. And then you have wind speed with the ultrasonic FT sensor. Yeah, which is a really nice sensor.

Okay, so that’s high quality stuff. But then inside of this cylinder, this metal cylinder, there’s circuitry. There’s a

André Bégin-Drolet: brain inside. There’s a brain inside something that would not have been possible to do 20 years ago through the democratization of electronics and microchip and everything. So we do live calculation inside the sensor using all these parameters and we…

And inside the sensor, we model, so we do live modeling of the surface temperature of of these cylinder. And why we use cylinder is because they’re easy to model. Yeah, we went for that.

Allen Hall: It’s a basic model. It’s an aerodynamic model.

André Bégin-Drolet: It’s an aerodynamic, thermodynamic.

Allen Hall: Exactly. And it’s simple to do sort of CFD, thermodynamic model.

Okay so now you have. One, two, three, four different sensors. You have a brain inside of it. And in that brain is a bunch of software, I assume? And that software is taking all those parameters and trying to figure out, Okay, ice is about to come, ice is over. Did you have to create those models yourself?

Did you go to a wind tunnel to do that? How did that all get done?

André Bégin-Drolet: It was developed through academic research. We had access to all these wind tunnels, infrastructures, and everything. We come with a background of 15 years of academic research. Where we did all this stuff. We went through a cold climate chamber to simulate the icing. We used all this knowledge, this academic knowledge. To come up with this nice product.

Allen Hall: So where was the icing wind tunnel at? Is that in Canada?

André Bégin-Drolet: It’s in Canada. It’s in Quebec City. So it’s a refrigerated wind tunnel that was built in the 60s. It’s a closed loop wind tunnel. It goes over two floors.

And it’s all wood. But we retrofitted the icing in there. We had to dry the tunnel after each run, but it was a very unique

Allen Hall: Yeah, because you usually don’t put water in these wind tunnels. That’s a forbidden thing to do. So you had to, must have twisted some arms there, convinced some people to let you get their fabulous wind tunnel wet.

André Bégin-Drolet: Yes, but we dried it. It was built in house. We know how to rebuild it. Yeah. We had that leverage and that’s what’s fun in the mechanical engineering department. What? We do things and we can place things.

Allen Hall: Sure, sure. Why was that tunnel there originally?

André Bégin-Drolet: It’s been there for a while.

It’s been there for a while. It’s, it was not refrigerated at the time. Yeah. This unit was added in the early or late 1990s. Okay. And then they built on that. Okay. Two different, experiment. It’s used for teaching as well.

Allen Hall: Okay. Making parkas and hats and all the Canadian gear, gloves.

Is that how they check all that equipment?

Joel Saxum: Canada goose checks.

André Bégin-Drolet: Exactly.

Allen Hall: It’s a good promotional tool for Canada. Okay, so you have all these resources at your fingertips. You’re, you’ve created this basic instrument. You’re now taking it to a wind tunnel. You’re validating it, you’re coming with curves or empiric.

Are you doing empirical measurements?

André Bégin-Drolet: We do a lot of empirical curve. Okay. In the wind tunnel. Alright. But then the real test was when it was on a turbine. ’cause when you’re on a turbine, you’re behind the rotor. There’s a lot of, there’s some weight, there’s turbine density mess, and it’s messy. The wind flow is messy. And we redid some institute calibration or, empirical curves on site and we also added cameras there so we had a side view of the sensor so we could measure the ice accumulation on the structure and then correlate our model and fine tune our model for and we have a different calibration for each turbine type.

Oh really? Yeah, because it depends on where it is located On the nacelle?

Allen Hall: On the nacelle, yeah. Okay, so I wanted to get into that because it relies so much on the airflow and the parameters around the airflow. How sensitive is that if you have to, so you’re taking a base model out to, let’s say a GE 3X, the magic turbine is in Canada at the moment.

So you’re installing it on the turbine, on the nacelle. You have a calibrate, it’s already calibrated itself. It’s close, right? It’s close. Yeah. It’s already calibrated. Yeah. You just go through an adjustment phase to understand are you understanding the local environment, or is it more specific to the aerodynamics around the, that nacelle and turbine?

André Bégin-Drolet: It’s specific to the flow that will go around these inter cylinder, because this is the piece that we’re investigating, those two probes. And we have two for redundancy. We know that ice will fall from the blades and will damage, and might damage the instrument. And so..

Allen Hall: If they don’t, if they install Borealis, it won’t damage it.

That’s why you install Borealis, so you don’t damage your icing instrument.

André Bégin-Drolet: And I’m fortunate enough to have a lot of very intelligent people surrounding me. Yeah. And we, with the instrument and with the brain that’s inside the instrument, we’re about to do automatic calibration. So we set it up there and it knows it has its own algorithm that will day after day.

Fine tune the model and we know when there’s no icing. So this is when we tune the…

Allen Hall: So it gets better with time. Kind of like Joel, better with time.

Joel Saxum: The outputs from the system, right? He said this dumb this down to someone that’s a like me and it’s gonna be on a site. What is it going to tell me and what actions do I take from it?

André Bégin-Drolet: So the first thing is going to tell you it’s the onset of icing. Yeah, so At that time, the conditions are prone for icing.

You should take some actions.

Joel Saxum: It’s a cloud based, I’m going to get an email, or I’m going to be able to check in?

André Bégin-Drolet: It’s we are, we’re very flexible. Okay. At the moment we, on some system, we’re integrating into the scada. We have also modem cellular communication feeds data through the cloud, and then it can text, it can email depends on the customer.

Everybody has something different. With the Borealis system, we’re Hardwire to their system directly and we will trigger when to start and stop the IPS.

Allen Hall: So does your system work in conjunction with some other systems that may be on the turbine already?

André Bégin-Drolet: Yeah, we’ve, we’ve, worked with the turbine operators a lot.

And they work with turbine manufacturers. So we had an agreement to trigger other IPS systems as well. Just to optimize the operation of those systems. So this is a direct value for the customer as they’re seeing the use right away of the sensor and they’re optimizing their operations.

Joel Saxum: So one thing I want to talk about again, I’m putting my operator hat on is we’ll offer.

We were talking about this a little bit. Of course, you guys are in Quebec and Canada. Almost everything is a fixed PPA price. So the markets are going to stay stable. But there is places in the world say like down in Texas where it’s an unregulated power markets. The power prices that are purchasing prices in the market can go up and down and up and down.

And we know that we saw, what, 9, 000 a megawatt hour? Yeah. Oh yeah. During the ice storm down there. So if you had assets that were able to produce during that, you could have really banked some capital away. You could have made some money. A lot of revenue there. So how can you guys help other people, when you’re not necessarily triggering an IPS system or triggering a heating system to turn on?

What are the other advantages for some people? ,

André Bégin-Drolet: Since we detect the onset of icing, it can give you an indication to stop bidding on the market. Yeah. You will produce at that time, but you know that in three hours from now if ice continues at the rate it’s going now. It’s not a binary.

We give it intensity value so we can Forecast what’s going to happen in three hours so you can stop bidding on the market. So you can stop because you will start stop producing so you will not be able to Come up with those bids.

Allen Hall: Which is a requirement at ERCOT, right? So the regulation that’s happening at ERCOT right now is saying you have to give us some warning.

We can’t just have you flipping off in 15 minutes. You’ve got to give us some advanced warning whether you’re going to be on or off.

Joel Saxum: Yeah, because it causes, the problem it causes down there is cascading browning. Where things all of a sudden this plant goes down, this plant goes down, then you all of a sudden have an unstable grid and then this one will go down.

So being able to notify the grid operator…

Allen Hall: Right

André Bégin-Drolet: It’s a, it’s a major advantage for the grid operator, but so far it’s been very difficult to get to them. We’ve been working a lot with the operators that have the problems, but then as soon as you get larger, you have to. It’s the larger organizations that have different issues.

Allen Hall: I think they’re all trying to understand the issue and they don’t realize the capability you’re bringing to the marketplace right now. Because it’s so different than what you would normally see.

Joel Saxum: Yeah, one of the other advantages we talked about with an operator is the ability to know when to shut your turbines down to avoid… Structural structural damage to your blades, or that damage or fatigue that will build up over time. When you run in ice, I’ve personally seen the same turbines in southern latitudes that are installed in northern latitudes, and the ones in northern latitudes having much more damage internally than the ones in the southern latitudes. And that is directly equated to running them with a bunch of ice. The advantage there is… That knowing when the ice is coming, because you’re going to accumulate less ice when you’re not spinning.

André Bégin-Drolet: Yeah, exactly. So when the turbine is spinning, it accumulates a lot of of ice. So the idea behind the early stopping of those turbines is to stop producing early on, so it will accumulate less ice, and then…

Once the event is over, as we can also tell when the event is over, then you can restart the turbine more rapidly than if they had run through the event. Some of the research has shown that there’s gain using this strategy.

Joel Saxum: Yeah, because sometimes you see turbines that have run in during an icing event and they have just massive amounts of ice on them.

Then they finally shut ’em down. ’cause the vibration alarms are going off and all that stuff. And then that ice sits on those turbines. It might be a week. Oh yeah. And it sits on there. Now, if you had shut it down early, you wouldn’t have accumulated all that ice and you’d be able to turn it back on.

André Bégin-Drolet: Exactly. And then you, while the other turbines are down like seven days, if you were able to restart your turbines. So one day and it, that’s, that means six day of product production. Yeah. For one day of preemptive stoppage at the beginning. So

Joel Saxum: It’s a, it’s a. I would say it’s a gamble, but it’s a good gamble.

Allen Hall: That’s a good bet. No, that’s a good bet. Yeah. Especially some of these marketplaces. So the Icetekh brand name is just hitting the markets from Borealis side. You guys, obviously, ATC Canada. You’re out here promoting the product, which is good. There’s a definite marketplace.

Canada. U. S. Nordic countries. Nordic countries. Sweden. Norway. Finland. Yep. Germany. Pick them. Have you seen acceptance of the products over in Europe yet? Because it seems like an obvious fit there.

André Bégin-Drolet: Yeah, we’ve started to have discussion. Obviously for us, it’s more difficult to sell over there than to sell locally.

We started exporting to the U. S., which was our objective for our year two. Yeah. It’s a big success for us. We’ve been through the whole process of exporting to the U. S. And then we want to go to the European market. And one big milestone that we’re looking at is… It’s working with the OEM, so being integrated in their supply chain becoming a sensor installed or at least offered as an option for the customer.

And I strongly believe that this sensor can help the OEMs have better products.

Joel Saxum: That’s fantastic. So how many units do you have out in the field right now?

André Bégin-Drolet: We have a little bit over 30 units in in the field we’re in Nova Scotia, New Brunswick, Quebec, Ontario, Manitoba, Saskatchewan we’re in Texas, Pennsylvania uh, Illinois Minnesota, and so on.

Joel Saxum: So for listeners that are thinking about this as new technology, Icetek’s been around, you guys, there’s 15 years of research been into the product, and now it’s proven, it’s out in the field, it’s working, and people are benefiting.

André Bégin-Drolet: With the we’re not just selling a product was also providing our expertise and our services. So usually what our sales cycle is pretty long. But we also after the product is installed we provide services for the data analysis and sometimes it works the backwards. So we will start analyzing data from the turbine and after that we will recommend to install it and I sensors. So this service or this data analysis will lead To hardware.

Allen Hall: Yeah, I was wondering about that aspect. We know that sales cycle. Oh, yeah, intimately.

Joel Saxum: We live that same cycle. StrikeTape and lightning. We’re in the exact same thing. Analysis, consulting, and then put the strike tape on.

Allen Hall: Yeah, then we sell a thousand, right? So it, but it takes, it’s the gestation time is, tends to be long in wind.

But in your case, because you have this really cool instrument and it’s providing data back to you smart people. Does that then create a little bit of a development like, oh, we learned in Pennsylvania that this kind of icing happens in this sort of situation that we didn’t know.

André Bégin-Drolet: There’s no big surprise.

I think it’s always different for every site and for every event and everything, but we with Daniela from Borealis she had access from the ice map from all over the world. So the these different ice maps and what we took is we took the database from all the wind sites all over the world and merge these two database together.

So now we know where ice is and on what turbine it’s a wind customer contact us. We already know that they’re experiencing the problem because we’ve correlated the database from the wind turbine site to the icing maps.

Joel Saxum: And and here’s a, this is a side note I’m thinking, but if I was looking at being a wind developer and I was sighting in the northern latitudes, you guys might be someone I’d call, regardless of the instrumentation, just for, as a process of going through my, my, my sighting reviews and things like that. What are the damages we may be looking at or what are the hazards we may be looking at with icing? Based on your knowledge.

André Bégin-Drolet: And the losses yeah, and state losses as turbines are getting bigger. The IPS are getting more mature with Borealis and the OEMs have also their system. They’re getting better. So the losses are going down, but there’s still losses caused by, by, by icing. You have to anticipate them. They’re going to be there.

Joel Saxum: It’s, and it’s something that developers know before they try to take a final investment decision on a wind farm.

Allen Hall: It’s. Such a weird marketplace, because you talk about operators that installed a farm five, ten years ago, and anticipated ice loss, versus what actually happened?

Widely different. Yeah, they were told a nice song and dance story about, Oh, a percent or two. This is the Rosemary story. So Rosemary’s giving me some insight on this, having been an icing person. They usually say, oh, it’s a percent or two, and you get five percent is the pain point, like five percent downtime, but put a…

a de icing system in, but I think the PPA prices in some of those places in Quebec, a percent or two makes a big deal.

Joel Saxum: Yeah, and with talking with Daniela earlier, there’s some laws coming possibly down the pipeline in Quebec as well that will focus around maintaining uptime wind assets based on ice.

Allen Hall: As wind has become a couple percentage point of the energy grid to, in Iowa, 60%, 70 percent electricity is generated from wind. Icing is now a bigger deal because if you cut off 70 percent of electricity in Iowa, there’s a lot of Iowans that are gonna be cold.

Joel Saxum: Yeah, a lot of people gonna ride around their tractor for heat for a while.

Allen Hall: Absolutely. They’re gonna fire up that wood furnace. Yeah on the pellet stove.

André Bégin-Drolet: And we’re based out of Quebec So we focus a lot on that market and there’s a lot of icing and at the moment It’s only 3 percent 4 percent of wind penetration in the network. Yeah, but as this number Will increase and will grow it should become Big issue on how to deal with icing and losing these big wind farts one after the other and balancing the grid.

Allen Hall: The cascading, as Joel pointed out, the cascading effect is probably the biggest risk to the energy grid because they start disconnecting and then they kick off some solar and it, and as we found out in Texas, you don’t even understand how it’s happening. You just know that it’s happening and you don’t want it to happen anymore.

Joel Saxum: Yeah, here’s a, and here’s a question for you as well. Could your system, this IceTek instrumentation, help out servers knowing when they’re going to need it?

André Bégin-Drolet: It’s a good question. You know that icing occurs in winter months. There’s less sun also during these periods. I don’t know, it’s a complicated question.

Obviously, this provides data. It’s what you do with the data.

Joel Saxum: Yeah, I’m thinking more along the, like the unregulated market, the same idea of when you’re going to be able to bid and when not to.

Allen Hall: Oh, that’s because it’s a trading market. Yes. You need to know when to bid. PPA, yeah.

Wow. Okay. Andre, how do people find Icetek to learn about the technology? How do they get a hold of you? How do they learn about all those cool things you’ve built?

André Bégin-Drolet: We have a webpage, icetek.Ca. Okay, it’s I C E T E K. I C E T E K. Dot C A. Okay. We also have a LinkedIn page where we try to update some material. We just posted a time lapse of an icing event just to show people or educate people on the icing type that’s occurring.

And try to be present on social media as well. Give us a call and drop us a line.

Allen Hall: Yeah. Okay check out Icetek’s webpage. Andre, thanks for being here. Thanks for taking some time to explain the technology. It looks really… Congratulations.

André Bégin-Drolet: Thank you. Thanks for having me.