The often decisive criterion in the introduction of energy management is amortisation. The question arises as to whether a saving is foreseeable in the short term or not. An analysis using the data from the existing transfer meter is the easiest way to get started. This is often used to check the electricity bill or identify peak loads. Together with the Chair of Information Systems Research at the University of Freiburg, ENIT Systems has investigated this question. For this purpose, a statistical analysis was carried out over 3-6 weeks for several industrial customers. The data is recorded from the usual transfer counters in high resolution every 15 seconds. This is used to estimate the savings potential of the respective industrial company. You can read more about what you can learn from your own data here.
The savings potential of peak loads can be estimated after a short measurement. This involves investigating whether peak loads that occur represent very unusual conditions or are close to normal operation. If a load peak is a clear outlier, there is a higher potential to avoid it.
In the diagram, the 15-minute values are ordered and plotted according to statistical frequencies. The steeper the increase on the right side, the more unusual is the peak load. The savings potential can thus be quantitatively estimated using an algorithm.
Almost all industrial companies have a weekly or daily cycle. With intelligent algorithms, the typical profile of a cycle can be statistically evaluated.
The figure shows frequently occurring power levels in brown and white. Areas marked blue never occur. This relative frequency can be used as a prediction for energy system optimization.
Background: Forecasts are becoming increasingly important in the energy market. The reason for this is the increasing share of fluctuating generators. Both technical operations management and electricity trading are facing difficulties as a result. Both have the goal of reconciling generation and consumption. Predictions increase security and increase margins as decisions are better made. In addition, they become essential for future business models in the balancing energy market or for flexibilisation.
„Compared to model-based predictions, this purely statistical method does not require any further knowledge of the plants. This makes the method much easier to apply“
(Pascal Benoit, CTO, ENIT Systems)
The determination of the operating status is of double value for companies: On the one hand, efficiency measures are identified. The figure illustrates which answers the load profile provides. From the level of the base load it can be estimated which savings potential lies in switching off unnecessary standby consumers. The determination of the productive energy quantity during normal operation enables an exact key figure formation (specific power consumption of the plant). The threshold to the peak load range is a necessary parameter for the correct design of load shedding systems.
Voltage is the decisive quality criterion of the electricity supplied. It is measured by most transfer meters with sharp phase accuracy. To ensure security of supply, it should also be recorded by industrial companies. The diagram shows the frequencies of the voltages that have occurred. The asymmetry between the phases is clearly visible. If this leads to problems in the company’s own power grid, the electrician needs an evaluation.
Background: Who causes „dirty“ current and who causes „dirty“ voltage? Simplified the following rule can be applied: The network operator is responsible for „dirty“ voltage. „Contaminated“ current is generated by the consumers in the industrial plant itself. Ideally, the company checks both.
„We can use the analysis as a fingerprint for each phase. This enables us to easily determine the unclear phase assignment in the distribution cabinets. In this way we can avoid asymmetrical conductor loads and increase operational reliability“
(electrician of a Freiburg company).