The low voltage grid needs to be more intelligent to meet the growing challenges facing grid operators today.

Although we are trying to leverage the information from the LV grid through smart meters, most operators are still falling short.

When setting the foundations for data analysis, we must go back to the roots and ask ourselves the right questions.

Are we digging in the right place for the information we analyze?

Sadly, the answer is no. A low-voltage grid is where the most change is happening. It is where we can aim for cleaner and more sustainable energy. Nevertheless, it’s seldom managed or mentioned.

Data analysis involves collecting and processing information from meters, sensors, and anything else that can supply information. That doesn’t mean reading a sensor every second but analyzing all the smart grid’s data.

Smart Meters allow us to monitor the state of our energy resources so that we can better manage them and save money. In addition, by connecting all the electricity meters into one central database, we can begin understanding how energy consumption varies across different parts of the LV grid.

This analysis could then lead to improvements in the design, construction, and operation all along the grid.

On a global scale, power networks are facing challenges of aging infrastructure, increasing demand, intermittent renewables, and distributed energy resources (DER). New technologies like smart meters help the industry collect high-quality, reliable, timely, and relevant data to allow utilities to understand what is going on at the distribution level.

Are the analytics getting the needed support?

Communication reliability is the backbone to support analytics. The most valuable benefit of energy communication is where the produced power matches the consumed energy by the load.
Achieving this level of energy balance reliability requires advanced smart meters, inverters, and converters with high-efficiency levels to allow better analysis results and lower the cost of electricity production.

Analytical energy management systems can monitor the performance of individual components, which enables the system to detect any anomalies before they become problems. In other words, maintenance teams can identify issues early and fix them quickly.

Are we following the right analytical methods?

The problem with data tools is that they can turn a blind eye to the real analysis problem. Visualization approaches give humans the sense that they understand the big data they are seeing, but a graphical representation will not take us a step forward.

Artificial intelligence is important if we want to achieve energy efficiency and sustainability. However, the role of artificial intelligence should be embedded within an engineering culture that understands how the electricity grid works.

For an artificial intelligence tool to work effectively, it must be trained using examples that contain many details. The goal is to make the AI learn how to apply the correct formula when dealing with new information or situations.

A lot of companies offer analytics tools that seem to be very useful at first glance. However, there are many problems where they fail miserably to provide feedback about how well the tool is working.

For example, when using an analytics tool, you might see a spike over a particular period. In reality, this may simply mean that the analytics tool failed to identify any issues that existed before the spike.
This means that there was nothing to trigger the spike.

We need more powerful AI algorithms coupled with engineers’ know-how to achieve near-perfect optimization insights and implement solutions.

The core question: what do we want to get from analyzing data?

The objective is to get actionable and timely insight for generating optimal business solutions.
For real-time monitoring of energy consumption at the household level, utilities must have a system that collects and processes data from measuring devices and displays information through a user interface.

There are now solutions based on IoT (Internet of Things) that provide information and insights to users. Some of them provide analytics to monitor energy usage and consumption, others are offering real-time pricing. The main challenge here is that most existing solutions are still in the beta stage and will require some time before they become mature enough for use in cases like this.
In addition, they focus on solving one aspect of the problem by using only one specific type of sensor, thus limiting the solution’s potential.

The solution should enable rapid analytics deployment across all data sources, including real-time insights provided by machine learning (ML) algorithms. It should integrate the latest data processing techniques including ML into a powerful platform that provides the ability to analyze vast amounts of data.

These cloud-based solutions provide access to the current information and deliver it in a way that facilitates creating virtual low-voltage grid models that simulate various technical scenarios, especially for EV and PV/DER rollouts.
That will help operators understand all the data coming from smart meters and act upon data insights.

NES is investing in using smart meters as sensors that can provide reliable and secure information. Then pairing this with AI & human knowledge to maintain other data sets, make the analysis accurate, and recommend tangible actions to achieve measurable and beneficial outcomes.