Studies suggest that efficient energy management can reduce companies’ energy usage and energy costs by up to 10%. For example, you can use a meter that breaks down the consumption of each machine. In the NILM research project, innogy is involved in developing this kind of smart meter of the future.
That sheet of paper with the kilowatt hours consumed and the invoiced amount – that was the electricity bill of yesterday. Since the beginning of 2017, it has been mandatory for large customers and companies to install intelligent electricity meters – smart meters. But a digital electricity meter alone does not an intelligent measuring system make.
How much do the individual appliances that you find, for example, in a large bakery, contribute to the overall amount? What ovens, mixers or refrigeration devices are secret electricity guzzlers? And how can companies avoid load peaks via load transfer? An exact breakdown of electricity consumption is essential for efficient energy management. However, for this to be possible, you need to be able to identify how much energy the machines consume.
One meter can handle all the machines
innogy is working on developing this kind of technology along with partners from business and science in the research project NILM (Nonintrusive Load Monitoring). The project is funded by the Federal Ministry of Economic Affairs and Energy. “NILM intends to facilitate the recognition of the consumption patterns associated with individual devices by deploying the right hardware and software, and on this basis, to allow efficient energy management”, says Gunnar Hoffmann, project manager at innogy. The advantage of this approach: in contrast to so-called submetering, individual meters no longer need to be installed on all machines – with NILM, a single meter is sufficient, which means much easier installation and lower maintenance costs.
We need to record several thousand values per second in order to identify these highly specific fingerprints.
Fine-grained data collection detects the fingerprint of the devices
The starting point for NILM is the observation that each device influences the electricity and voltage in a characteristic way and thus leaves a “fingerprint” in the home network. This fingerprint can be thought of as a signature. In order to recognise these signatures precisely, it is necessary to use highly frequent, that is, very fine-grained, data collection procedures. “We need to record several thousand values per second in order to identify these highly specific fingerprints,” says Gunnar Hoffmann. “It is especially difficult to tell devices of the same type apart because their patterns are, of course, similar.”
If you know your energy guzzlers, you’ll want to get active
It’s worth making the effort: this fingerprint doesn’t just allow you to measure and break down the power consumption of each machine in real time. You can also identify damage or wear based on deviations. By conducting early repairs or maintenance, you can thus avoid production losses. Up to now, this energy management system has only been tested among selected industrial and commercial customers that have high energy consumption levels. When installing the technology, Gunnar Hoffmann noticed one thing: “Customers had very low levels of awareness of how their electricity costs were distributed at the beginning.” This changed during the test phase. The information on the exact consumption levels also increased the need to reduce energy consumption – and thus reduce energy costs.