Wednesday, April 29, 2015

Predictive Analytics – Power Transmission

Power transmission through age old transmission equipment [large & complex networks like transformers, overhead lines, cables & other equipment] creates lot of hurdles and loss. It is expected that a rough estimate of 17% of the generated power gets lost in the transmission from power generation tower to the local grid.  While the technical loss happens due to energy dissipated in the conductors, equipment used for transmission Line, Transformer, sub- transmission Line and distribution Line and magnetic losses in transformers; the non-technical or commercial loss happens due to illegal stealing of power, etc.  

To leverage real-time data management to meet reliability goals, you need to deploy predictive analytics. By running the data gathered by your analytics sensors through the predictive analytics application, you would come to know where unexpected shutdowns and catastrophic failures are about to happen. Besides, you could predict transformer, compressor, and tower conditions, monitor cable and joint corrosion, reduce damage to critical equipment, etc. 

On the whole, applying predictive analytics on power transmission data helps you maximize focus on your critical resources 24/7/365.

Predictive Analytics – Power Generation

Power generation is a costly affair. The huge turbines that generate the power are to be constantly monitored for all those critical parameters; day-in and -out. When a few minutes of downtime can attract losses in terms of millions of dollars, it is normal that power generating companies opt for predictive analytical solutions to help them maximize the availability of operations, increase system uptime, create additional revenue opportunities, lower maintenance costs, etc. 

Analytics help you to identify the wear and tear of your machines in advance, lower maintenance cycles, identify potential barriers before they occur, fix minor issue before they snowball into catastrophic events, and dynamically adjust parameters to tune the machines for optimized performance.

Fuel quality, power generation forecasting, etc., are a few critical factor that needs to be checked time and again while generating power. Besides, pollution, environmental compliance, etc., are a few factors that are increasingly challenging and are spinning and spanning a wide array of issues. With the help of sensors, both physical and virtual, not only you can check the areas of risk but also monitor and diagnose the overall health of the infrastructure equipment and components every fraction of a second. 

Predictive analytic models, when applied on the gathered sensor and historical data, project different intuitive charts and graphs which display the future trends of power generation on your snazzy dashboard. 

Tuesday, April 28, 2015

SCADA [Supervisory Control and Data Acquisition]

SCADA  – This acronym has been the heartbeat of the power industry before the ‘smart grid’ came into picture. Now, SCADA has become legacy concept. The new technologies are finding ways to interface with legacy SCADA to extract as much information as they can to know more about the functioning and performance of the target plant. 

If you treat SCADA as a rudimentary system, it has to be beefed up in the following areas to convert it in to a smart grid: Data communication, intelligent alarm management, data volume, communication protocols, virtualization, etc. 

However, SCADA can be used to manage plants, which run on both renewable and non-renewable resources. Based on your requirement, either you can combine or separate SCADA system for transmission and distribution. Although it allows one to operate it remotely, the problem with SCADA is that from security point of view when interfaced with Net, it can be easily hack able. 

Across the world, process control systems for industrial processes are sold in two flavors: SCADA/PLC & DCS.  

In a SCADA/PLC system, it the PLC [Programmable Logic Controllers] which is actually controlling the plant; SCADA provides the human interface. The modern DCS [Distributed Control System] is quite advanced compared to SCADA/PLC system since the former has both the plant control and human interfacing combined in one system. 

In summary, if efficiency is the point of discussion, then DCS outshines SCADA/PLC in so many ways.

Predictive Analytics – Asset Lifecycle Management

Utility management in power transmission and distribution field is a vital task. Since the end-customer is also one of the stakeholders when it comes to power consumption sector, the companies that undertake the transmission and distribution facility have to make sure to maintain negligible downtime. It is a loss to the company and frustration to the end-user if the grid downtime is high. In this scenario, the only solution that helps the companies to keep the grid up most of the time is Predictive analytics.

Companies need to overcome so many challenges to keep the downtime of the grid low. So far, the T&D companies have been managing the grid in a traditional manner; with aging assets and low returns. Now, with the implementation of predictive analytics, the situation changes drastically. Companies can predict the failure of equipment in advance and order for replacement. When the customer expectations are rising for an unhindered power supply scenario and governments across the globe are pushing down the throat of the companies to reduce carbon targets, it is very clear that the situation has put the companies at a juncture where they have to focus on the asset maintenance part.

Since asset maintenance plays a vital part in reducing the downtime and maximizing the uptime, T&D companies are taking the help of predictive analytics to take care of the grid asset lifecycle: from ‘asset investment planning’ to ‘operation and maintenance through decommissioning and disposal/ replacement’. If an asset fails, the consequence of the failure creates a snowball effect on the T&D company such as expense of the asset in service, collateral damage cost, regulatory penalty, disposal of damaged asset, lost revenue, and other intangible costs.  

By implementing predictive analytics, T&D companies get the following benefits when it comes to managing assets: 

  1. Extend asset life  
  2. Bring more predictability to asset performance
  3. Help to plan and prioritize maintenance activities
  4. Reduce asset lifecycle cost
  5. Enhance business process
  6. Improve productivity
  7. Improve customer satisfaction
  8. Improve power reliability based on planned outages
  9. Reduce unexpected asset failure cost [leading expense component of any asset]
  10. Improve forecasting and scheduling of assets

In summary, predictive analytics helps T&D companies manage their assets well, reduce grid downtime, and delight their customers by providing maximum grid uptime.

Monday, April 27, 2015

Devastating Earthquakes – Scope of Predictive Analytics

It seems predictive analytics stuff stops short of predicting earthquakes. Irrespective of the data that we feed the application and use the right model, it seems we are still far away from making the right prediction when it comes to predicting the quakes. 

Life becomes hard when you have to experience situations like this; for example, earthquakes. It drags down our spirit, abolishes our dreams, ruptures our progress, and make us stand in the life’s cross road;  not knowing what to do. 

The only hope at the moment when it comes to predicting earthquakes scientifically is through predictive analytics, which is not up to the mark. All those complex models that we have today are not answering one and only question; when that quake hits us? How we can protect our lives and assets from the beastly claws of earthquakes, which rupture everything that is related to basic living? 

Even the Box-Jenkins model [time-series analysis], the most popular model, is not helpful to do the needful when it comes to predicting quakes. For predicting, precision matters a lot; and for this we need quality data, which we have when it comes to earthquake stuff.
However, still, we are lacking somewhere! Either we need to increase the scope of our understanding or we do not understand the real-scope of predictive analytics a lot. The latter point raises so many in-depth questions, which affects all the industries across the verticals, which are using and recommending predictive analytics at the moment. 
This blog thinks this is the right time to do some introspection regarding where we stand when it comes to understanding the scope of predictive analytics.