Services
IoT and Digital Engineering
Bringing Together IoT and Digital Engineering to Create Rich Experiences
Highlights
- Even as the power generation industry transitions to renewable energy, it remains overly reliant on siloed technologies that lack scalability and fail to provide a coordinated data ecosystem.
- The energy transition greatly expands the need for data, given the increasing complexity of a portfolio that integrates fossil and renewable generation.
- Ongoing changes in the industry also lend urgency to the adoption of advanced digital analytics, artificial intelligence (AI) tools, and scalable cloud architectures.
- We propose a four-step incremental approach to getting the most of AI, analytics, cloud, and distributed computing for a successful green energy transition.
On this page
On this pageinpage
ENERGY TRANSITION
The power generation business is changing fast as it addresses the unprecedented challenge of a global economy transitioning from fossil fuels to carbon-free alternatives.
Power companies that are most agile—especially those willing to transform their information technology (IT) and operations technology (OT)—will be better at navigating this change and tapping opportunities along the way.
This is an industry dominated by mature companies with long-lived assets and legacy systems. Some of them are still relying on outdated data management systems and siloed OT tools. There is a tendency among them to address specific data needs in isolation—acquiring a solution to improve the gross heat rate for specific coal-fired unit, for example.
We also too often see companies that simply don’t have the data they need, in a form they can access and use, to solve important business challenges ranging from optimization of equipment to effective planning of preventive maintenance.
The lack of accessible data results in missed opportunities for efficiency. True digitalization demands that discrete IT and OT issues be addressed with a bigger data landscape in mind, rather than with siloed solutions. Management should be thinking about a longer timeframe and how multiple solutions can work together in a modern data platform and architecture to enable data-driven decision making.
Power generation companies that lag in data transformation will face higher technology costs and increased labor requirements in the long run. To be sure, some companies have moved to cloud computing for certain purposes, while others may be testing or learning about new AI tools. But few have created a fully integrated digital ecosystem that can provide effective centralized management of their IT and OT requirements.
Power companies cannot manage their data as they have in the past. A robust centrally managed architecture becomes vital to harmonize data from diverse sources.
DIGITAL NEEDS
As power companies build their renewables portfolios, their data management and analysis requirements expand and change fundamentally.
Gas- or coal-fired power plants are highly centralized, with everything in one location to produce hundreds of megawatts of power. Renewables are distributed. A solar power facility might cover hundreds or thousands of acres, and a collection of wind turbines may spread across an even larger landscape. Support for renewable generation infrastructure, including management offices, data centers, maintenance personnel and equipment, may be many miles distant.
This has significant implications on a power company’s digital needs. The dramatic increase in data generation and data needs is one example. While the number of sensors in a large fossil power facility may be in thousands, wind and solar farms are estimated to generate 60 times more data than anything a traditional power plant might produce. Renewables produce granular information about their environment, their power output, and their level of efficiency. For instance, the data produced on cloud for a solar installation is vast as it covers wind strength variation moment by moment across hundreds of wind turbines, among others. Renewables produce big data.
The management of big data that comes from renewable power installations demands a different strategy than power companies are used to. The data from vast, distributed energy sources needs to be brought into a centrally managed data platform. That’s the challenge. Power companies cannot manage their data as they have in the past. A robust centrally managed architecture becomes vital to harmonize data from diverse sources.
A standardized solution approach leveraging IT OT
A standardized solution approach leveraging IT OT
A standardized solution approach leveraging IT OT solutions
Even the management of existing fossil fuel plants may demand new thinking in the current era. Consider how the addition of renewables, whose output varies with the sun and the wind, might put new demands on older plants to ramp their generation units up and down more frequently. This not only creates dispatching challenges, but also changes the operational profile, putting a premium on data that can help operators understand how new operating patterns affect efficiency or maintenance. What can power companies do to meet these challenges?
FOUR-STEP APPROACH
We recommend a four-step approach to fast-tracking digitalization using the power of AI, cloud, analytics and edge, to give companies the digital advantage they need to improve existing fossil fuel assets and fully integrate renewable energy ones.
It leverages IT and OT solutions to help create a standardized data architecture with the scalability, security, reliability, and data support needed to adapt to the changes that are sweeping through the industry. It’s as simple as ABCD.
Step A is for artificial intelligence. In power generation, a range of problems that might have been time-consuming or difficult in the past can now be addressed with AI tools. For example, optimization of an aging coal-fired power unit—including reductions in emissions, improvement in auxiliary power consumption, reduction in maintenance costs, and improved safety—becomes easier with AI solutions.
In fact, some new tasks related to the rapid growth of renewables can only be tackled efficiently with the application of advanced AI and machine learning (ML) pattern recognition technology. Use cases for AI and ML may include analysis of dust deposition on a field of solar panels to help set maintenance and cleaning schedules. Or AI could be deployed to recognize and make sense of cracking patterns in the paint on wind turbine blades seen in video captured by a drone used to inspect a large wind farm.
Step B is for business analytics. AI and other data analysis and management tools are becoming increasingly important for power companies to keep costs in check, ensure reliability, and remain competitive. These capabilities must be available in a way that ensures they add value for the enterprise. Data solutions that support improved decision making and promise the greatest potential business benefit must be given top priority.
Power producers are investing heavily in setting up infrastructure for power generation units, but a coordinated and centrally managed data architecture is equally important and inevitable. It is a must to make data-led insights readily available and in a consumable format for key stakeholders. Robust business analytics can help to ensure that these are sustainable investments that will provide the return on investment that management expects and demands.
Step C is for cloud. Because renewable resources such as solar and wind are likely to be distributed across a wide area, the data they produce needs to be collected somewhere. Using an on-premises data center at a central location might have made sense for a power plant in the past, but for management of data from renewables, the nature of the task clearly points to the value of the cloud. Cloud-based data tools are the best option to collect, store, and analyze the output from sensors, digital images from drones, data on environmental conditions, and more.
The standardization that cloud architecture enforces will benefit the organization in many ways, helping to make all of the data more readily available for business analysis needs, for example, or making security audits easier.
What’s more, cloud provides variable computing resources in an efficient manner, including the compute power needed to bring new AI capabilities into play. This is particularly important for any company with a renewables portfolio, which will have intensive data needs at times and few computing resource demands at other times.
Step D is for distributed computing. For the intensive big data analysis tasks of a renewables portfolio, companies need a data platform with distributed computing capabilities. Distributed computing on cloud can provide the raw processing power needed to manage, for example, millisecond-level output data coming from multiple wind turbines. It can also provide scalability, fault tolerance to smooth over network or hardware issues, data replication, and data communication and synchronization abilities.
Our work with some of the world’s largest power companies have shown us how early movers have energized their business and made them more sustainable.
INNOVATION IN PLAY
Powering ahead with technology
One of the largest power generation companies in Japan, a country where emission of pollutants is tightly controlled, leveraged an AI-infused intelligent power plant solution to optimize and transform the operations of its fossil-fuel power plants. After the success of what started off as an initiative for one of its coal-fired units, the company is now scaling the program to other units. The initial work involved the creation of an AI-based predictive model for a 1,000-megawatt coal-fired unit, tapping massive amounts of historical power plant operating data, and using that data to generate recommendations for operational changes or new set points for fuel and air parameters to minimize emissions and maximize fuel efficiency. Once implemented, the new set points helped the company realize benefits including an 8 percent reduction in nitrogen oxides emissions and a 1.6 percent reduction in unburned carbon.
Another example is that of a utility company in the United States. It sought a data transformation that would support the company as it shifts the portfolio from fossil fuels to renewables, retiring its coal plants by 2028 and adding wind and solar. For the renewable assets, which are widely distributed and generally untended, it adopted a data platform for centralized monitoring. The solution allowed peak asset performance monitoring, integrated with grid demand management functions. The platform became a source of truth with fleet data history for the wind turbines and solar generation at a central location on the cloud. The company could enhance the comfort of users, increase energy efficiency, and reduce the environmental impact.
In another example, as part of its energy transition initiatives, a leading power generation company in Australia has successfully evaluated the potential of AI and digital twin to improve the way key gas turbine components work and address the challenges of flexible operations. They have formulated a roadmap to scale structured technology adoption over the next three years.
THE FUTURE BECKONS
The energy transition calls for a reset at many ends.
It is nearly impossible for a human being to physically manage offshore wind farms or solar power plants in the deep deserts. In such extreme conditions, a robust communication system is a must for real-time insights. Then, there are geo-political constraints that create an economic impact on business operations. Advanced AI-based digital twins developed for monitoring and managing different kinds of assets distributed demographically can help deal with such uncertainties. Companies must realize that investing in sustainability today could become a competitive advantage in the future. They must embrace technologies such as AI, business analytics, cloud, and distributed computing today to accelerate their move to net zero and get to a brighter, more sustainable future faster.
