Search

Research & Development How modular battery storage systems can reduce peak loads

| Author / Editor: Christopher Lange * / Florian Richert

Fraunhofer IISB has integrated a scalable battery system into its institute network, thereby demonstrating its profitability and transferability to consumers with large electrical load peaks.

Figure 1: The stationary electrical energy storage system for peak load reduction at Fraunhofer IISB in Erlangen was developed as part of the Bavarian energy research project SEEDs and integrated into the institute's own direct current grid. A software tool developed in the SEEDs project is available for the design of electrical storage devices and for the simulation of smoothed load profiles..
Figure 1: The stationary electrical energy storage system for peak load reduction at Fraunhofer IISB in Erlangen was developed as part of the Bavarian energy research project SEEDs and integrated into the institute's own direct current grid. A software tool developed in the SEEDs project is available for the design of electrical storage devices and for the simulation of smoothed load profiles..
(Source: Kurt Fuchs/Fraunhofer IISB)

The topic of peak load reduction is of economic importance for industrial companies and commercial electricity consumers. However, the aim of smoothing load profiles often requires unwanted interventions in production and costly changes to the infrastructure. Technological advances and falling prices are now enabling the profitable use of electric battery storage systems. As a result, electrical load peaks on the consumer side can be reduced without having to intervene in production processes.
As part of the Bavarian energy research project SEEDs, Fraunhofer IISB in Erlangen is showing how stationary battery systems can be integrated into existing energy supply infrastructures. Currently, a scalable battery system with 60 kWh storage capacity reduces peak loads in the institute network by about 10%. The usual operating procedures have not been and will not be affected by this. The results of the research work can be applied to industrial or commercial energy systems with large electrical load peaks.

Peak load reduction without hindering production

Peak loads inevitably occur in almost every load operation. These load peaks are always undesirable because they are cost-intensive and load the power grids. As a rule, attempts are made to compensate for these load peaks by temporarily switching off production systems or switching them on with a time delay. Such measures, however, mean massive interventions in production. A much more elegant solution is the integration of electrical buffer storage to reduce peak loads. This makes production-relevant interventions superfluous and the solution is also suitable for reducing peaks in the network.

Economic and technical considerations

Energy suppliers and grid operators are interested in grid utilization and power consumption that is as even as possible. The legislator has created the appropriate legal incentives for this. On the part of commercial consumers, for example, the service price and in particular, if the conditions are met, the Electricity Grid Charges Ordinance (StromNEV) offers interesting opportunities for significant cost savings. There is great potential here for optimizing the timing of energy consumption through load shifts.
From a technical and economic point of view, the targeted reduction of short-term peak loads is very interesting: even relatively small investments lead to high-cost savings. Thanks to recent developments in battery technology, numerous possibilities are opening up for stationary electric battery storage systems to compensate for cost-relevant peak loads with short-term load shifts. However, this requires a thorough analysis of the current situation, tailor-made hardware, and software solutions and intelligent control based on sophisticated algorithms - an ideal field of activity for Fraunhofer IISB.
In particular, the Bavarian energy research project SEEDs offers scientists attractive opportunities. The entire institute and its diverse infrastructure can be used as a demonstration platform for an intelligent decentralized energy system. In terms of power requirements and infrastructure, Fraunhofer IISB is comparable to a medium-sized industrial company. All developments and technologies are therefore also tested in practice and under industry-like conditions.

Undesirable: Deviation from load profile

Load peaks are high power references that are only available for a short time and which stand out from the typical load profile. The reasons for this are usually the connection of large individual consumers, which are switched on or off briefly. This can have different effects on electricity providers, network operators, and users. Due to the temporarily increased network load, a larger dimensioning of all network components is necessary, as the design is always based on the expected maximum load. Theoretically, energy suppliers must react very dynamically to a rapidly changing demand. However, this is only possible to a limited extent, if at all. Special peak load power plants can absorb the increasing demand for control energy, which results in additional investments in new power plants with short operating times.

Figure 2: The assigned maximum permissible 15-minute average power value must not be exceeded at any time, otherwise additional costs will be incurred. Any higher output will be compensated by discharging the battery system. The battery is charged analogously if the power consumption is below the charging limit..
Figure 2: The assigned maximum permissible 15-minute average power value must not be exceeded at any time, otherwise additional costs will be incurred. Any higher output will be compensated by discharging the battery system. The battery is charged analogously if the power consumption is below the charging limit..
(Source: Fraunhofer IISB)

Battery buffers compensate for peak loads

Electricity suppliers create incentives to avoid peak loads in the form of various price models. For example, a combination of a performance price and a reduced labor price often applies to larger customers. For the service price, the highest peak load that occurs during the billing period is relevant. The averaging interval for the performance value is 15 minutes. However, the Electricity Grid Charges Ordinance also permits individual grid charges, such as in the context of so-called atypical grid usage or based on intensive grid usage.
The price models mentioned enable cost savings through peak load reduction. The prerequisite is always to influence the load profile to avoid peak loads. In the simplest case, electrical consumers such as production or infrastructure facilities can be switched off. Conversely, it would be conceivable to switch on one's electrical generators, e.g. a block-type thermal power station. Both measures require intervention in ongoing production operations and the existing infrastructure. As a result, considerable costs can arise due to production stoppages.
The solution is an intelligently controlled battery system with which the financial potential of peak load reduction can be exploited without affecting the production process. In principle, the battery storage unit is charged at low power levels and discharged at times of high power levels. The predicted 15-minute average value must always be taken into account. The aim is to reduce the maximum power consumption: the resulting power price is reduced and electricity costs are reduced. In practice, cost savings of 70 to 90 € per kilowatt can generally be achieved for reducing the annual peak load, depending on the power price.

Figure 3: To demonstrate the algorithms, a MATLAB app with a graphical user interface was developed in which peak load reduction can be calculated using given load profile sections. In addition to simulation, functions for data analysis, battery design and economic considerations are also included.
Figure 3: To demonstrate the algorithms, a MATLAB app with a graphical user interface was developed in which peak load reduction can be calculated using given load profile sections. In addition to simulation, functions for data analysis, battery design and economic considerations are also included.
(Source: Fraunhofer IISB)

Practical self-experiment confirms transferability

The scientists at Fraunhofer IISB are testing live how well this works with a modular battery system with a capacity of 60 kWh, which will be expanded to 100 kWh. The researchers have developed an algorithm and corresponding software for control and regulation to make optimum use of the battery storage and to switch it on at the right time. With battery sizes of 60 or 100 kWh, a possible reduction of the peak load of 10% or 16% is already possible for this application. With the current battery prices, amortization periods of five years are possible.
The practical results at Fraunhofer IISB show a very good agreement with simulations carried out previously and are transferable to other consumers. To guarantee an efficient and profitable integration of the battery storage, the institute's specialists follow a three-stage approach: With the help of a comprehensive data analysis, various parameters relevant for the further procedure are first extracted. This also includes specific parameters of the load profile under consideration, such as the energy conversion or the statistical distribution of load peaks. The battery data, such as maximum power and total capacity, are then determined using a special optimization procedure. The exact determination of these data avoids an over- or under-dimensioning of the storage system. If necessary, the algorithms for the operating strategy can be adapted during this phase. In the final step, the load curves are simulated, which, based on the historical data, resulting from the use of the battery system.

Figure 4: At the beginning a charging process is visible and the state of charge rises accordingly. Peak load reduction starts around 10:30 a.m.. The peak load reduction in the test period was around 56 kW, which corresponds to a reduction of 9%.
Figure 4: At the beginning a charging process is visible and the state of charge rises accordingly. Peak load reduction starts around 10:30 a.m.. The peak load reduction in the test period was around 56 kW, which corresponds to a reduction of 9%.
(Source: Fraunhofer IISB)

Measurement results and validation of the algorithms

A maximum permissible output of 570 kW was specified. The Fraunhofer IISB battery system with a capacity of 60 kWh was used for application-oriented validation. At the beginning a charging process is recognizable and the state of charge rises accordingly. Peak load reduction begins around 10:30 a.m. The peak load reduction in the test period was around 56 kW, which corresponds to a reduction of 9%.

The Fraunhofer IISB algorithms not only allow battery systems to be designed to meet demand and optimally used for peak load reduction. Individual extensions with additional components can also be taken into account, such as a combined heat and power unit with a heat accumulator. It is often also interesting to make infrastructure systems for the provision of heat and cooling more flexible utilizing thermal storage and to integrate them into peak load reduction. The focus of the work is always on transferability to other energy systems for the broadest possible application of peak load reduction measures.

In addition to various energy suppliers and energy system operators, battery manufacturers are also recognizing the potential of peak load reduction. Together with the Brilon-based battery manufacturer HOPPECKE, Fraunhofer IISB has designed an innovative high-performance battery storage system for an industrial customer. Both the efficient intermediate storage of large amounts of energy and the delivery of high outputs had to be ensured. The result: an energy storage system of around 350 kWh would enable peak load reductions of around 40% since many of the peak loads only occur for a very short time.

Frederik Süllwald, Key Account Manager at HOPPECKE Batterien, reports: "By reducing peak loads, our customer would have a savings potential of around 45,000 euros per year. This saving would mean that the memory would be paid off within a few years and the customer would be in the profit margin.

This article was previously published in German on Elektronikpraxis.

* * Christopher Lange is a research associate at Fraunhofer IISB, Erlangen.

(ID:46251564)