You might be surprised to learn that all power plants have some sort of primitive "networking" capability, albeit something that looks more like a direct point-

You might be surprised to learn that all power plants have some sort of primitive "networking" capability, albeit something that looks more like a direct point-to-point phone line between plants and the grid operator (see DNP3.) As you can imagine these are relatively expensive to install, but part of the bring-up of a new power plant. All of these plants also have some sort of SCADA system internally and turbine fields need the same (the turbines need to be connected to each other, and the operator needs to be able to control the turbines in some way: e.g. turning them off when there is an excess of power, reading power output, faults, etc.) So one way or another, communication needs to occur whether it's over an IP network or a dedicated PTP line. If someone gets physical access it's mostly game-over unless you've got builtin hardware-level encryption between all of the internal components of the facility. If there wasn't an ethernet port in the server cabinet the hacker could just connect an RS-485 line and start sending MODBUS commands or whatever protocol the SCADA system is using. But the article does a good job pointing out that the assumption that "the facility/network is secure" does not hold as well in distributed generation as it did for traditional power generation. Edit: see also https://en.wikipedia.org/wiki/DNP3 https://en.wikipedia.org/wiki/DNP3