In October 2014 two hackers got all of Spain’s power grid under their control. They exploited the blanket integration and central organization of all power meters to hack into the firmware. It was apparently relatively easy to falsify the identification numbers. Once they got inside the system, they were able to manipulate power meters, install malware and even turn off the power completely. Fortunately this whole hacking exercise was merely simulated. But it revealed a major security gap in the country’s intelligent power grids and power meters. The manufacturers have been warned. But what could a possible solution look like?
As practical as it may be to have a centrally organized power grid, it also provides an attractive target for attackers. The Max Planck Institute (MPI) for Dynamics and Self-Organization published a potential solution in January 2015: decentralized organization. This sounds simple and, apart from protecting power grids from hackers, it is designed to bring numerous other advantages, such as the enhanced security of customer data. After all, the data concerning your power usage would no longer be stored at a single, central location. But decentralization can also create new problems of its own. The lead physician at MPI, Benjamin Schäfer, was asked to tackle the issue of how and whether intelligent power meters installed at the customer’s home could directly and decentrally regulate the temporary frequency fluctuations caused by power supply ebbs and surges. One of the main problems consisted of the delayed response by many appliances to these short-term frequency fluctuations. The kind of router, for example, which would in future be required to manage the electric appliances in a household, requires a certain amount of time to compute the changes.
For help in solving the problem, Schäfer contacted Thomas Walter. Walter’s company, Easy Smart Grid from Karlsruhe, Germany, develops system solutions for operating decentralized energy networks. Numerous small producers use his solutions to supply the electricity they generate to large-scale power plants. Schäfer took a mathematical model which Walter had developed and simulated the decentralized self-organization of power meters in response to fluctuating power supplies. The results were clear-cut: Intelligent power meters do not need to react immediately, since minor frequency fluctuations over a short period of time cancel each other out. To the contrary, it is actually more beneficial if the new frequency values need to be calculated first, since the concomitant power consumption can then be more precisely adapted. The researchers at MPI thus demonstrated that intelligent power meters can function properly without a central management system. Whether manufacturers end up taking advantage of these findings remains to be seen, since they are generally more committed to Industry 4.0 , i.e. interested in fully integrated plant and equipment linked to centralized and consequently cost-saving process management systems. Ultimately it is a question of striking a balance between security and costs.
International exhibitors will be showcasing their smart grid solutions at HANNOVER MESSE, and demonstrating just how well the latest developments are designed to cope with hacker attacks, for example at the leading trade fair Energy .