Connecting the grid with the internet
Two of my recent blog posts looked at the Virtual Power Plant (VPP) in the Smart City region of Rheintal. So far I’ve interviewed the manager of the project at Bosch and the manager of the Rheintal project at illwerke vkw about the progress of the region’s VPP. I’ve now had the opportunity to talk to the head of the VPP working group, Dipl.-Ing. Prof. Jörg Petrasch.
Prof. Petrasch, where are we with the VPP? What uses and benefits is it already providing?
We can no longer afford to simply adjust power production to meet demand. We have to take the opposite tack and adjust consumption to meet production. The challenge is to adapt thermal loads such as heat pumps, boilers, and air conditioning systems, not to mention electric vehicles (which are also key consumers), so that they can adjust their own consumption or storage function according to what’s available locally from volatile sources, in particular solar and wind power, at any given time. We also have to do this for stationary and mobile storage units.
And this is precisely what we’re doing with the VPP: it is already making sure that we can shuffle grid loads – with the target to match “uncontrollable” power. Take the new neighborhoods and the heat pumps that we’re in the process of integrating. These pumps run only a couple of hours a day. Thanks to the VPP, they now get switched on only when there is an abundance of electricity available – otherwise they remain off. We can make these kinds of optimizations without users ever knowing.
Then we throw in hydroelectric power – which in Vorarlberg accounts for the bulk of the local energy supply and can be regulated quickly and easily – add smart load management via the VPP, and you end up with a smart and comprehensive energy system for the region.
Which challenges did you have to overcome?
First we had to ensure the technology was in place for the devices and the server to communicate with one another – in other words, set up stable access to the features of the integrated heat pumps and boilers via the VPP. Such devices aren’t usually commercially available. Then there are the particular safety requirements for the grid and the consumers. When we talk about connecting the grid and the internet, we’re talking about bringing together two worlds, two paradigms. What counts here is preserving user privacy. The Vorarlberg University of Applied Sciences worked with Bosch to develop a VPP system that provides the relevant indicators (price, actual and forecast output) to consumers that are tapped for optimization – effectively demand side management.
What roles do forecasting and optimization play?
Using the VPP to achieve efficient load management largely depends not just on how prices develop, but how production develops as well. The more accurate the forecast – for example, the output that will be available in two hours, four hours, and so on – the more effective the algorithms that calculate optimum use within the overall system. You have to have accurate information about consumption and how much energy is available on the grid. We’re going to be putting a lot of time into refining what we’ve learned.
What have you learned so far?
If a forecast is off by even a little, it can throw optimization efforts completely off, too, or lead to serious errors. Then you can experience the worst case scenario in which consumption happens when prices are high.
Although it’s a lot easier to integrate household consumers into the optimization via the VPP than electric vehicles, the way you do it remains essentially the same for both, of course. The extra challenge when dealing with EVs is that they are not always connected to the grid and it’s much harder to accurately forecast consumption. Heat pumps, however, are always connected, and we’re able to make reasonably sound forecasts for hot water usage: most of us shower early in the morning or in the evening. So the margin of error for heat pumps is significantly smaller than for EVs.
How do you see energy and the VPP within the context of the smart city?
The idea behind a Smart City is to use a VPP to integrate whole regions or neighborhoods into an overall energy system while also finding common ground for the heterogeneous consumer structure (electromobility, households, manufacturing, etc.) and the interests of the grid operator (vkw). It’s really important to bring the small, private producers together with the large power plant operators and to coordinate efforts to ensure grid stability. The key idea behind the regional VPP is for the average amount of energy needed in a year to be produced locally, which involves finding a way to balance out extreme fluctuations in power generation. In the future, we won’t be able to get around mastering this technology and making it a commercial reality.
Outlook: In your view, what’s the next important step?
I see two steps as being crucial over the next couple of years.
1) To begin with, we need the VPP as a platform for selling electricity generated by decentralized producers. The challenge there is to get trading going on a more local level.
2) Then there’s the real need to make accurate forecasts. On a typical day (see screenshot above), the price on the energy exchange can be represented as a jagged line, and consumption as a steady curve. In order to use heat pumps to balance loads – in other words, to smooth out the curves – we have to activate them at the ideal time, namely when consumption and price are both low. A resolution of 15 minutes (horizontal axis) is ample for carrying out this optimization. What’s important is that the price and load forecasts (vertical axis) be as accurate as possible.