Energised for innovation
As the world becomes moreenergy-constrained, the energy sector needs to prioritise efficiency over innovation.
Innovation is the process of transforming ideas into new goods or services or leveraging new ideas to improve existing offerings and deliver greater value. In the case of the energy sector, innovation doesn’t entail reinventing the wheel; it’s about making sure that the wheel turns as efficiently as possible.
“Globally, the world is becoming significantly more energy-constrained and this is going to be the case for a long time to come,” says Stefano Marani, CEO of Renergen. In the next 50 years, as more and more of the developing world gets electrified, consumption of electricity is only going to rise and we will need to come up with ways to significantly increase our generation capacity, he says. Delivering the increased capacity society needs while also moving away from carbon-intensive fossil fuels like coal – which is currently responsible for just over a third of global electricity generation – is a very big ask. So much so that when you look closely at the numbers, the world’s just transition goals are a pipedream.
The case for digitalisation
“Research from the IMF and World Bank reveal that just removing coal – not oil, not gas, not any other form of energy – from the global energy mix will cost around $78 trillion in today’s money. Global GDP is around $90 trillion. So, we’d need to invest as much as 80% of the total global GDP towards changing our generation habits,” Morani says. He adds that current global spend on new generation facilities across every single class of energy, from nuclear to coal to gas to oil to solar to wind, is $350 billion a year and has remained at similar levels for the last decade. “This means that it’s going to take about 222 years for us to eliminate coal.”
At present, Eskom provides as much as 95% or 53.7 Gigawatts of South Africa’s total electricity supply, according to GreenCape’s 2023 energy services market intelligence report. Unfortunately, the degradation of the power utility’s existing coal fleet has led our annual average energy availability factor (EAF) to decline from 94% in 2002 to 53% in 2022.
But Grant Hughes, principal cybersecurity architect at Engen, believes that our local energy supply constraints serve as the perfect opportunity for digitalisation. Digital technologies enable real-time monitoring and control of power generation and, in doing so, can reduce losses and enhance energy efficiency. When demand outstrips supply, demand response and load management is essential, he explains. “With the current loadshedding crisis, smart grid technologies could allow consumers to adjust their power consumption based on grid conditions. This will alleviate pressure from the grid and potentially avoid a total blackout.” Energy analyst Chris Yelland agrees.
Digitalisation enables us to move away from using a flat rate for electricity, which is a constant price per kilowatt hour (kWh), to a time-of-use strategy, where the price of electricity varies depending on different variables – such as the time of day, for example. With real-time pricing of electricity, consumers can take advantage of price reductions and only operate certain home appliances at times of the day when electricity is cheapest. Digitalisation also enables us to move away from an old, centralised approach to electricity production towards distributed generation, with smaller and more diverse energy producers rather than a few big central power stations, Yelland says.
Unfortunately, the sector hasn’t been quick to adopt new technologies because they are, understandably, preoccupied with trying to keep the lights on. Additionally, digital transformation has traditionally been slow because there are a number of factors that need to be considered when talking about how to move the power industry forward, adds Hughes.
Any disruption in service can have a significant impact on the economy, which means that the sector needs to be especially careful when planning to digitalise, he says. “In most cases, the infrastructure is complex and aged, which means that implementing new technologies could require the upgrading of entire production lines and/or rebuilding of entire plants, which simply isn’t viable.” Legal and regulatory obligations also need to be considered as these can introduce barriers and stall progress if regulations around the use of new technologies are outdated or are still being developed. But this doesn’t mean that everyone is slow to embrace innovative tech.
As the holder of South Africa’s first and only onshore natural gas and helium production right, Renergen’s Virginia Gas Project is a unique one. And when you’re dealing with something unique, you need to embrace a unique approach, says Marani. “To paint a picture of how it all began, we had no idea where the gas was coming from, so one day, we went out to a random area and just started drilling. And we quickly struck gas,” he says, stressing that this was a once-in-a-lifetime strike and certainly wasn’t something that happens every day.
“Drilling is the absolutely definitive way of proving what you’ve got underground, but drilling is incredibly expensive, so you can’t afford to drill unless you know that you’ve got a reasonable probability of striking gas.” To increase their chances, the team developed an in-house system that combines topographical data and environmental features from satellite images and detailed underground geological information with a robust set of machine learning tools.
“We had already drilled many wells, blowers and non-blowers. When we combined this data with what the geology and the topography looked like, the machine started to pick up common surface characteristics and could indicate the probability of us striking gas in a particular location,” he says. This approach meant that their negative confirmations – when they drill, but find nothing – now hold value because all data fed into the machine learning model improves accuracy.
“Every time we drill, there’s a new data point. We put the new data point into the machine and it just gets better and better over time.” And, thus far, the accuracy of the model has been impressive.
“We’ve moved from 50% accuracy to 100% accuracy.”
The model can even predict flow rates – or how much gas will come out of the ground – before they even start drilling. When the stakes are high and resources are scarce, anything you can do to de-risk your work and increase the likelihood of picking up patterns that a human might overlook makes good business sense, Marani says.
In most cases, the infrastructure is complex and aged, which means that implementing new technologies could require the upgrading of entire production lines and/or rebuilding of entire plants, which simply isn’t viable.Grant Hughes, Engen
According to Hughes, machine learning and AI could also be a game changer for Eskom. “A big factor in our current loadshedding problem can be attributed to equipment failure. AI and machine learning could be used to analyse large volumes of data from sensors to inform maintenance schedules and predict equipment failures. This could extend the lifespan of equipment, reduce downtime caused by equipment failure and enhance the overall operational efficiency of the grid.”
Furthermore, these tools can automate certain processes so that preventative maintenance actions can be undertaken without any need for human involvement, he says.
Eskom’s plan to launch a “virtual wheeling” product represents a step in the right direction, says Yelland. In the past, wheeling involved a transaction between a single independent power producer (IPP) generating electricity into the Eskom grid and a single customer, or off-taker, also connected to the Eskom grid, but located elsewhere. Virtual wheeling will make it possible for both industrial and commercial customers to buy electricity from the electricity traders that have recently been licensed by the National Energy Regulator of South Africa (Nersa). Not only will this boost supply, but it will also make pricing more competitive.
Something like virtual wheeling is only possible with the necessary information technologies in place to integrate everything and bring all of these different players together, Yelland says. “As we look to increase supply and come up with ways to produce and deliver power more efficiently, it is clear that digitalisation is key to the future success of electricity services.”
* Article first published on brainstorm.itweb.co.za