Big Data, Analytics, and Energy Consumption

Would you run the dishwasher in an off-peak time if you knew how much less expensive the energy you were using was at that time?  Would you do it if the energy company provided an incentive (and reminder)?  These are key questions for analytics professionals and energy distribution companies around the world.  One of the key problems with improving the cost distribution of energy is getting consumers (both Consumers and Businesses) to be aware of their habits and chose to do something about it.  Until recently, there has been a lack of information and understand to cause such a change.  However, with the advent of smart meters within the power grid and on key power consuming devices, we are now able to begin the detailed analytics to change behaviors and costs.

A key physical property of the energy distribution market is, “use it or lose it.”  When energy is put on the distribution network it must be used at that time.  Energy providers are experimenting with storage devices to assist with this problem, but they are nascent and expensive.  To generate or buy energy, these providers have a need for very detailed analytics.  Many models have been created over the years to predict the need for energy based on historical consumption and/or weather patterns.  Most of these models are based upon macro level data related to the network and demand, which limits visibility and resolution.

The introduction of smart metering devices presents many new opportunities, but also many new challenges.  Let’s look at the challenges first.  The first being the massive amount of data that can be created by the smart meters.  For example, if a distribution network has 1 million metering devices in it, the amount of data can grow exponentially quite quickly.  The table below shows the potential growth (assuming 1 million collection devices and a 5 kilobyte record per collection) in a year.

Collection Frequency  1/day  1/hour  1/30 min. 1/15 min.
Records Collected  365 m  8.75 b  17.52 b  35.04 b
Terabytes Collected  1.82 tb  730 tb  1460 tb  2920 tb

This presents not only a storage problem, but an analytic problem of making sense of all that data.  Some of the challenges will be to cross-reference that data with customer information, network distribution and capacity information by segment, local weather information, and energy spot market cost data.

Harnessing this data will allow the utilities to better understand the cost structure and strategic options within their network, which could include:

  • add generation capacity versus purchasing energy off the spot market (e.g., renewables such as wind, solar, electric cars during off-peak hours)
  • investing in energy storage devices within the network to offset peak usage and reduce spot purchases and/costs
  • provide incentives to individual consumers or groups of consumers to change energy consumption behaviors

Could you imaging receiving a text message from your utility offering a discount or price reduction if you stopped using your dishwasher now and waited till off peak time?  Would that change your behavior?

Lavastorm is embarking on an exciting project to explore such analytic problems with innovative companies such as Falbygdens Energi AB (FEAB) and Sweco (please see our recent announcement on delivering optimized smart grids for consumers and utilities).   We believe our Lavastorm Analytic Platform will help answer some of these key questions and improve the way that we all think about energy and what it costs us to consume it.   Please check back for frequent updates on how we are proceeding with this analytic endeavor.