Sunday, March 27, 2011

Empowering the Power Sector with Analytics



This summer while you and I enjoy an uninterrupted power supply in our homes and offices and take it for granted, we hardly realize the amount of planning that goes behind the scene! A quick refresher to the country’s power distribution wouldn’t be out of context.

Since the year 1998, Govt. of India through a slew of acts and regulations has emphasized on ‘rationalization of electricity tariff’, promotion of ‘efficient’ policies and establishment of regulatory bodies like Central Electricity Authority (CEA), Central Electricity Regulatory Commission (CERC) etc. Its direct offshoot the Availability Based Tariff (ABT) framework promotes responsible and efficient power usage by both Distribution companies (Discoms) and Power Generation Companies (Gencos), aptly facilitated by Dispatch Centers (SLDC and RLDC). In essence, ABT mechanism ‘incentives’ and ‘dis-incentives’ market participants against deviations in the committed demand and supplies. Without going into detail, ABT breaks-up electricity charges under: Capacity charges (fixed), Energy charges (variable) and Unscheduled Interchange charges (UI). To optimize the UI charge, drawl schedules has to be managed in 15 minute blocks (24 x 4 = 96 blocks per day). So a Discom which draws more than the schedule causing the grid frequency to deviate would be subjected to heavy penalty. On the other hand, drawing less than actual demand means forced power-cut resulting into lesser revenue, customer dissatisfaction and even socio-political scrutiny.

Now return to our story: this summer while we will cool-off imagine someone is entrusted to forecast how much power you and me would consume in every 15 minutes block. Add to that the complexity millions of consumers with different electricity consumption patterns. There are households, institutional users like offices, hotels, services like municipality (remember the street light!) and obviously manufacturing units. Challenge is accurately forecasting their pattern every single day with just the help of excel spread sheet! To top it all, for financial survival Discoms are expected forecast it with more than 90 % accuracy. :( WOW! That’s an effort!

Here comes the most unusual savior…. Business Analytics, well, forecasting technology to be specific. In the parlance of power distribution it’s known as Short Term Load Forecasting (STLF) solution. It incorporates historical demand pattern for last couple of years and takes into consideration relevant influencing factors like weather parameters like temperature, rainfall, wind speed, humidity etc. as well as macro-economic events like festival, holiday, and natural calamities. As we have seen, these factors are very local in nature and heavily dependent on the peculiarity of the consumer base.
You would probably say ‘Thanks for the good news ..but what exactly stops us using the solution and ensure an uninterrupted and efficient supply of power this summer?’ Well, all we need is the mindset and adaptability towards a scientific approach to resolve India’s scare power resource. Personal experience of interacting with the stakeholders proves beyond doubt that the systems, the relevant data and are in place which can be leveraged with the appropriate usage of business analytics solution. The execution capability of the tool is proven. Thanks to a couple of early adopters, who are successfully managing the complex power demand scheduling that not only bring top line and bottom line growth but address the larger question of responsible and efficient usage. Several developed countries with special mention of the Nordic ones are using it efficiently. Many of the Asian economies, proudly led by handful of Indian distribution companies are using it successfully hitting accuracy zone as high as 98%. Every 15 minutes, every day!! WOW that’s heck of a solution!

The moot point here - power of analytics is not only restricted to the subject matter of corporate discussion it actually touches lives - yours and mine. It directs us towards the better utilization of scare resources. Now that’s what we call proper use of power (pun intended!) albeit with the help of analytics.

Post Note: Surely behind the surface a very sophisticated statistical algorithm is at work but true to the character of this blog let’s leave the technicalities to the experts. Oops……..did I take my name?! :)