Inside the world’s first digital twin of a hydroelectric power station

Located approximately 60km south of Dublin in the Wicklow Mountains, the Turlough Hill hydroelectric power station generates 292MW during peak demand periods by releasing water from its upper reservoir and allowing it to flow through its four turbines into a lower reservoir.

During periods of lower demand the water is pumped back to the upper reservoir ready to be used again.

The energy storage that Turlough provides is increasingly important in the changing energy ecosystem, which is moving towards more flexibility to incorporate renewables. But due to its age and size, ESB does not know how much structural life the hydro plant has left and whether its operations can be made more responsive in the future.

As part of the Free Electrons Accelerator Programme - an initiative to connect startups with energy utilities to support and foster innovation - ESB was introduced to Akselos to help find some answers.

“The challenge for ESB is how to continue to operate and on what basis? That's what the digital twin can help answer,” says Andrew Young, Akselos' VP of projects and delivery.

To build the digital twin, engineers at Akselos needed to collect all available data on the gigantic asset which could then be used to create a 3D rendered structural digital model.

This task was made harder, however, by the lack of available digital information. Akselos’ engineers had to work with paper records only, which put the project behind by two months. In total, it took five months to complete.

To build the model, it was key to establish the geometries, interactions and stresses in the plant system to create an intuitive heat map. The map can show when and what tensions are running through the power station and direct maintenance inspections to specific areas of concern, rather than just doing a random walk through.

The digital twin is essentially a tool with which ESB can simulate many parameters around the day to day operation of the plant and visually pan through the entire structure and see how it reacts.

“Our technology allows maintenance engineers to really pinpoint, using rules of science and engineering, where the issues will likely occur and we can also establish what the operational life is in terms of fatigue assessment. It also supports predictive maintenance,” says Young.

An inspection can cost around €3-5m, through a combination of lost production and penalties for down time, so the technology can potentially save companies money.

Young estimates the CAPEX invested for Akselos' digital twin technology by a utility can be recouped within a three year period, just on the basis of improved inspection. Furthermore, he says it can provide a 20-50% life extension.

Predictive maintenance is not new. Bigger companies such as ABB and GE also offer this technology; so what makes Akselos’ offering different?

According to Young, existing predictive maintenance modelling and applications typically use asset management tools to achieve around 30% predictive. However, Akselos add a scientifically based structural integrity component.

“We can simulate future events because of the nature of our physics-based twin, whereas what other companies are doing is assessing the probability of a known and identified defect reoccurring,” explains Young.

Essentially, Akselos’ software assesses and identifies the risk of any type of failure based on the structural assessment; a fault doesn’t have to already be pre-identified.

“A GE report notes that structural integrity provides a 90% predictive capability, so we believe it is actually complementary to what companies like GE are doing, but gives a higher level of understanding of the asset,” he adds.

In fact, Young says the company is currently in the assessment phase of several projects with OEMs, which he can’t name, on integrating Akselos’ technology within their offering.

“These companies are starting to offer digital twins with every piece of kit they sell, and in the future, if they deliver a twelve-megawatt wind farm, it will have Akselos’ software inside, running as part of their predictive capability offering,” says Young.

Now the digital twin of the Turlough Hill hydroelectric power station is built, the next stage of the project would be to create a Digital Guardian, says Young. To do this, engineers will connect specially placed sensors to get a real time picture of what is happening in the hydro plant. This is expected to be planned for the next maintenance window.

Through the models, ESB will look to assess different modes of operation that might better work alongside renewables, such as storing electricity at different times of the day.

“An additional model of operation could be pumping in the daytime to store power when it’s needed, in addition to at night, as renewable energy sources are changing the dynamics of energy supply and pricing,” explains Young.

This means rather than having one cycle a day, the plant might have two or more. However, Turlough wasn’t designed to operate in this way, therefore ESB needs to understand what the impact might be on the structure. These factors, added to aging asset management risks, could increase inspection frequencies and costs or even require a significant CAPEX project.

Using Akselos’ technology, Young says there is no reason why the life of the asset could not be extended another 20 years.

“I don't think ESB came on board just to do a pilot, I think the company is really interested in this technology and how they can apply it to different assets,” says Young.

Adopting the innovation is largely about adapting to the changing energy ecosystem sooner rather than later, and becoming more efficient and more cost effective in the process.

“Legacy assets still have to be managed for some years despite the energy transition, and they're looking for cost effective ways to do this,” says Young.

“The utility industry is one of the first to encounter huge disruption and ESB is adapting to it extremely well by applying innovation to find new business models, this is what we're hoping to do.”