Self-consumption optimization & peak load shaving

Battery storage systems are considered a key technology of the energy transition because it makes fluctuating renewable energies such as solar and wind power flexible to use. Both private households and companies can benefit from Behind-the-Meter (BTM) battery storage systems. BTM storage serves to reduce electricity consumption "behind the meter", save energy costs, and reduce grid fees.

There are three central application cases that BTM battery storage can specialize in:

(1) Self-consumption optimization (using as much self-generated solar power as possible),

(2) Demand peak clipping (clipping expensive demand peaks in electricity consumption, also known as Peak Shaving) and

(3) a combination of both strategies.

The following sections will explain these use cases in a practical and understandable manner – including their implementation in simulations and application in real operations.

1. Self-consumption optimization – using more solar power yourself

The self-consumption optimization focuses on maximizing the share of self-generated electricity from a PV system. In practice, this means: When a lot of solar power is produced at noon, this surplus is stored in the battery to be consumed in the evening or at night when needed. Without storage, excess PV electricity often has to be fed into the grid (often at a lower remuneration), while during sun-poor hours, electricity is purchased from the grid at a high price. A battery storage system can close this gap.

How is this technically implemented? A simple control algorithm charges the battery whenever PV generation > consumption (surplus electricity) and discharges it when consumption > PV generation (deficit), as long as the battery state (State of Charge, SoC) allows it. This way, grid consumption is minimized. Mathematically, the self-consumption strategy can be formulated as cost minimization: The electricity purchase costs over a period are minimized by keeping the power drawn from the grid low and avoiding feeding in (for which there is only a small remuneration). Simply put, the goal is: "Consume as much of your own solar power as possible, buy as little grid power as necessary". In a simulation, this could be realized through an optimization model that decides for each time interval whether to charge, discharge, or do nothing with the battery – while adhering to battery capacity and power limits.

Implementation in energy management: Many energy management systems (EMS) today use forecasts for PV generation and consumption to proactively control the battery. For instance, forecast-based battery charging can ensure that the battery has enough capacity available before a sunny day (so that no PV power goes unused) and is sufficiently charged before an evening increase in consumption. The challenges here mainly lie in forecast inaccuracies and dimensioning: If the battery is too small or incorrectly charged, it can happen that it is empty too early or cannot take on all surpluses. It is also important economically to consider that a storage system for higher self-consumption is not always cost-effective – many companies already directly use ~60–70 % of their PV electricity, making a storage unit often provide only a minor benefit that does not justify the costs. Nevertheless, some users appreciate the increased independence from the electricity supplier and the long-term planning of electricity costs, which makes them willing to invest in a storage system to maximize self-consumption.

2. Demand peak clipping (Peak Shaving) – flattening expensive demand peaks

Demand peak clipping – also known as Peak Shaving – is an application case primarily for commercial and industrial electricity consumers. The aim is to cut short peaks in electricity consumption, as these cause high demand costs in many electricity tariffs. Specifically, grid operators calculate an annual or monthly grid fee for larger consumers based on the highest power drawn (kW) during a billing period. Demand peaks occur, for example, when several large machines start simultaneously or other consumption-intensive processes happen at the same time. Even if such a peak only lasts a few minutes, it can significantly increase the electricity bill. Here, the battery storage acts as a buffer: Once consumption exceeds a defined threshold, the battery automatically provides the additional power, instead of drawing it from the grid. Thus, the grid load is limited to a maximum value (the "clipped" peak), and the measured peak consumption stays lower. During times of low demand, the storage is then recharged from the grid or excess PV power to be ready for the next peak.

Example: If a company has a tariff with €100 per kW demand price and normally peaks at 900 kW, without storage this would result in ~90,000 € in demand costs per year. A battery system could limit these peaks to, for example, 800 kW, reducing the billed peak value – every avoided kW peak saves costs directly. In a practical example, a 100 kWh/50 kW battery storage can reduce a demand peak by 50 kW, which at a demand price of ~200 €/kW can mean about 10,000 € in annual savings.

Technical implementation and simulation: In simulations, Peak Shaving is formulated as an optimization problem where the maximum grid power should be minimized. Mathematically, this can be expressed as the minimization of the peak load $P_{\text{peak}}$:

$P_{\text{peak}} = \max_{t} \big( P_{\text{Load}}(t) - P_{\text{Battery-Discharge}}(t) \big)$,

where $P_{\text{Load}}(t)$ is the power demand from the consumer's perspective and $P_{\text{Battery-Discharge}}(t)$ is the power delivered by the battery. The battery should be controlled in such a way that $P_{\text{peak}}$ is minimized. Optimization occurs under constraints such as limited battery capacity and maximum discharge power. In practice, this is often represented by choosing a threshold: for example, "Draw a maximum of 800 kW from the grid; everything above is covered by the battery." An EMS with Peak Shaving functionality continuously monitors the power and regulates the battery in real-time to flatten the load curve. It is important to have an intelligent algorithm that recharges the battery immediately after a peak load clipping to be ready for the next peak, as load forecasts can often be inaccurate and a single "missed" peak can already raise the electricity bill.

Challenges in application: Demand peak clipping requires sufficient dimensioning of the storage for the individual load profile. The storage must have enough power and energy available to cover typical peaks – very high but rare peaks may not be fully clipped without maintaining an oversized (and uneconomic) storage. Therefore, in planning the storage, an economically optimal threshold must be found at which clipping occurs. Properly implemented, Peak Shaving not only reduces the electricity bill but also alleviates the power grid – flatter load curves mean less need for grid expansion and reserve power plants, thereby contributing to climate protection and grid stability.

3. Combination of both strategies – the best of self-consumption & Peak Shaving

In practice, battery storage is often designed to provide multiple benefits simultaneously to increase profitability. A combination of self-consumption optimization and peak demand clipping allows for achieving synergy effects: The storage can absorb solar surpluses during the day and contribute to cost reduction in the evening while simultaneously smoothing short-term demand peaks. Through such multiple utilization, the benefits increase, and the payback period of the storage system is significantly shortened. More advanced planning tools, such as Lumera Energy, therefore offer so-called multi-use concepts, in which a battery storage system flexibly manages multiple operating modes.

How does the combination work? In principle, both strategies can be controlled with a prioritized rule set. A simple solution is to reserve part of the battery capacity for Peak Shaving and use the rest for PV storage. For example, in an energy management system, one can set: "Always keep 30 % SoC in the battery free for peak clipping; use the remaining capacity for PV self-consumption". As long as the battery is above 30 % charged, it can absorb excess solar power or discharge for the household (self-consumption mode). If the state of charge drops to 30 % or below, this remainder is kept as a reserve that only intervenes when a grid load peak occurs (peak shaving mode). However, the reserved 30 % of the battery often remains unused, which can prevent higher self-consumption and also requires an oversized battery, reducing economic efficiency.

A more modern solution involves optimization models that consider both objectives within a common objective function. In such models, costs from energy consumption (€/kWh) and demand consumption (€/kW) are minimized together. This can be formulated as an optimization problem with constraints related to the battery (capacity, power limits, SoC continuity) and time series for PV generation and load. Such integrated optimizations allow for reducing power costs while simultaneously increasing self-consumption, bringing both financial savings and higher supply resilience.

Challenges and practical relevance: The combined operation significantly increases the requirements for energy management. It is important to avoid conflicts, for example: Should the battery be fully charged for self-consumption on a sunny day, or should capacity be kept free because there might be a consumption peak in the afternoon? One solution is the mentioned reserve strategy or an intelligent forecast that recognizes based on planned production or historical data whether peaks are expected on the current day. Accordingly, control can dynamically set priorities – for example, on a quiet Sunday, utilize the full capacity for PV, but on a busy weekday morning, be more cautious with discharging to prepare for midday peaks. Multi-use systems therefore require sophisticated software. Last but not least, economic efficiency plays a role: The acquisition of a large storage system is expensive, but the combination of several benefits creates additional revenue sources or savings potentials. Studies and practical projects show that it is only through the combination of the use cases (possibly supplemented by additional services such as arbitrage trading or provision of ancillary services) that a battery storage system becomes truly profitable.

Conclusion

Battery storage systems can offer different advantages depending on the application case. For companies with demand-dependent tariffs, demand peak clipping is often the decisive lever to reduce energy costs while simultaneously alleviating the load on the grid. The combination of both strategies allows for maximizing the full potential of a storage system: During both day and night, in electricity and demand consumption, the storage takes on a dual function. However, there is no universal solution – every project should be individually planned. Load profiles, PV generation, tariff structures, and company operations must be analyzed to find the optimal operational strategy. With the right design and intelligent energy management, a battery storage system can, however, be a powerful tool: It can help significantly reduce electricity costs, maximize self-consumption, and alleviate the power grid – a win-win situation for operators and the energy system.