Vijay Namjoshi, Chief, Generation, Tata Power
The year 2020 may be over, but it leaves behind unforeseen and lasting transformations in the way we do business. During the Covid-19 pandemic, innovation and digitalisation emerged as unexpected allies, helping utilities exploit technology and expertise better than ever before. Companies that were already along the digitalisation curve benefited from their greater built-in resilience during the crisis, being less dependent on human resources. The pandemic experience will further accelerate digital transformation and make it an integral part of the fabric of business processes.
Today, power sector utilities, especially thermal generation plants, have to deal with complex industry dynamics – volatile market conditions, increased competition, stringent operational boundaries, tighter regulations, changing workforces, and constrained budgets. Business environments have little to no tolerance for inefficient operations and the resultant cost overruns, missed generation targets, or safety/environmental incidents. The challenge for many plants – definitely for older ones, especially ageing and vintage assets – is to keep facilities running for longer in a cost-effective and competitive manner.
How does digitalisation deliver value to the power sector?
Till a few years back, everything was going well in the thermal power sector, and complexities were limited. Now they face rising operating costs, increased competition from clean energy sources, stringent environmental emission limits, etc. Against this backdrop, digitalisation presents unprecedented opportunities – from actively managing ageing assets, to harnessing disparate data sources for real-time intelligence regarding the network, assets and customer services. The sector is ripe for realising value from digital transformation along the value chain, from generation to customer relationship management.
Power plant operators are increasingly looking at digital solutions in order to optimise performance and asset availability, while lowering life cycle costs. By blending the physical and digital realms, these solutions provide both “inside the fence” and system-wide views, enabling power producers to maximise a plant’s operational efficiency. This will prove to be a boon to power plants that leverage analytics to enable predictive maintenance and proactive replacement of assets at risk. These solutions help power plant producers get far more out of their physical assets, improving the performance of machines and showing how they can be redesigned to do even more.
Predictive analytic technology
Predictive analytic software now offers sharp snapshots of the future output, behaviour, and maintenance actions of systems ranging from turbines to pumps. The predictive analytic models of such software accurately represent the plant or fleet under a large number of variants related to fuel mix, ambient temperature, moisture, load and weather, and can predict the outcomes along the different axes of availability, performance, reliability, wear and tear, flexibility and maintainability, with greater accuracy than before.
There are many such software solutions in use today which utilise advanced pattern recognition technology, artificial intelligence and sophisticated data mining techniques. They learn patterns of behaviour from an asset’s operating history and develop a series of normal operational profiles for that equipment. Then they compare the known operational profiles with real-time operating data to detect subtle changes in system behaviour that are often early warning signs of imminent failure. When the actual pattern diverges from normal patterns, the system reports the anomaly as an indicator of potential degradation.
Predictive analytic software, implemented at a centralised fleet monitoring centre, provides foresight into impending problems, thus helping avoid issues before they occur and driving greater process consistency and asset uptime. With such software, the system can successfully identify various plant anomalies at an incipient stage and help plant engineers take timely mitigating actions before they lead to catastrophic failure or loss of generation. This could lead to significant savings and availability improvements, while improving equipment health visibility and optimising the logistics of maintenance. In addition, a real-time analysis of generation can provide insights into optimising multiple generators under certain variable operating regimes, which can reduce operating costs and increase the generation fleet’s overall cumulative output.
The growing complexity of today’s equipment and systems and the rising cost incurred by loss of operation as a consequence of failures have brought the aspects of reliability, availability and maintainability to the forefront. Today, the expectation is that equipment and systems will perform their intended functions failure-free for a given time interval. In fact, there is no longer any room for rework, or recurring failures.
Plant maintenance is often based on a mix of reactive, preventive and predictive techniques. Preventive maintenance, which has been traditionally applied in power plants, is based on manufacturer recommendations regarding when to perform standard maintenance tasks and repair or replace parts or components before they become costly problems. However, each asset or component is subject to its own unique set of environmental conditions and operational experiences, which are not fully considered by the average lifetime estimations used in preventive maintenance strategies. As a result, preventive maintenance guidelines are overly vague, and err on the side of excessive maintenance. Predictive maintenance, meanwhile, relies on detailed usage data, operations and maintenance history, and condition measurements that are specific to each asset. But operators using predictive maintenance are challenged by a lack of visibility into the degradation status of assets, and by not having appropriate predictive intelligent asset strategies. This is where digitalisation will deliver value in formulating a maintenance strategy with a holistic view of asset health.
Many digital solutions are available in the asset performance management (APM) domain. These solutions rely on data historians to extract structured data related to equipment/components. The information collected is automatically analysed to determine a range of probabilities for the likelihood of failure over a series of time ranges. Data can also be used to determine the best plan for preventing or delaying failure, based on patterns of information from similar assets running under similar circumstances, correlated with prior prevention strategies and their outcomes. The solution methodology during implementation can be aligned with the latest asset management standard, PAS 55/ISO 55000.
The visibility delivered by advanced analytics can give plant leaders unprecedented granular views into assets, increase agility and support more strategic decision-making. By correctly estimating asset reliability, the remaining useful life and the probability of failure, power generation utilities can significantly reduce maintenance costs, unplanned shutdowns, and safety incidents. Presenting asset health data in a dashboard format to leadership teams will be very helpful for rapid, informed and, most importantly, accurate decisions. Interweaving plant maintenance with digitalisation will shift the maintenance paradigm, improving both mean time between failure and mean time to repair.
Performance improvement using digital solutions
Station heat rate and efficiency improvement are buzzwords in the power sector. Plant operators try and test all possible options to sweat out assets till the last possible drop, in order to achieve the best performance in various parameters. Heat rate is a function of multiple plant variables, along with different load and ambient conditions, and improving its performance can be a difficult nut to crack. Often, retrospective analysis is required to fix the root cause of poor performance. Why not take full advantage of digital capabilities and software solutions, which are a dime a dozen these days?
Plant optimisation software, either in the form of plant analysis, diagnosis and optimisation solutions or stand-alone AI-based systems, can be developed as control room desktop tools that identify the bad actors out of a myriad of operating variables and provide real-time operator advisories to run a plant in the most optimal manner, thereby leading to on-the-spot improvement in unit heat rate. Such systems can be self-learning and can adapt to newer plant conditions. Such solutions are central to reducing costs by increasing plant efficiency.
Managing cycling issues
With the increased penetration of wind and solar energy into the grid, India’s energy mix is expected to change significantly in the next few years. The variability and intermittent nature of solar and wind power will have to be managed and supplemented by other sources of energy in order to ensure grid stability and security. This will inevitably push the coal-fired power plants to cyclic operating modes, creating a new set of challenges for both existing and new thermal power plants.
There is a pressing need for coal-fired generators to adjust to the “new rules of the game”, and they need new capabilities to react to this development, including faster start-ups and shutdowns, lower minimum generation, higher ramping rates, and more frequent changes in generation. Many units, designed as a base-load plant, are grappling with this issue, and this type of off-design operation is resulting in accelerated rates of life consumption due to the initiation of fatigue-related damage mechanisms. Apart from the need for greater maintenance, these additional stresses can further lead to possible failures and reduced equipment and plant life.
Operational flexibility is characterised by three main features: the overall bandwidth of operation (ranging between minimum and maximum load), the speed at which net power feed-in can be adjusted (ramp rate), and the time required to attain stable operation when starting up from a standstill (start-up time). Digitalisation and the use of novel data analytic techniques can help address these challenges by employing previously recorded data as well as data gathered in real time to provide guidance to plant operators regarding maintenance of systems and equipment. Digitalisation solutions can be put to use to manoeuvre the equipment through operating margins, in order to improve these flexibility features and supplement plant operators with expert advisories during cyclic operation.
Digitalisation, without question, is the trend for the power sector in the future. The ball has started rolling in pursuance of that goal, and the canvas for development is huge. With a whole host of services and innovations that can ride the digital platform, industries can draw immense benefits, with a lot of potential in the analytics and predictive terrain.
There will be early adopters who will forge the path forward, testing the limits of digitalisation, verifying solutions that offer benefits and dispensing with those that don’t. Utilities can miss the ways in which these disruptive technologies can jump industry/market boundaries and change the rules of the game. Hence, it is necessary to develop well-informed views on what developments such as cloud computing, internet of things and advanced analytics could do for enterprises, and work to separate hype from reality. Each utility has to embrace digitilisation from its own point of reference, according to its technological maturity and propensity for change. By systematic analysis, companies can learn the value opportunity of each area and the feasibility of attainment of this value, with careful consideration of all the challenges across the end-to-end ecosystem. With clarity on the expected changes, companies can develop digital blueprints and lay the groundwork for their successful implementation.
It is imperative for all the corporate sectors to nurture local talent that can truly innovate and break through in a legacy-heavy environment, and evaluate how specific digitalisation technologies could drive economic impact and disruption in ways that could affect the businesses.