Load Management in Multi-Family Buildings: How to Avoid Costly Peak Loads

Load management in a multi-family building means actively controlling the simultaneous peak demand from heat pumps, storage systems and charging points, so nothing is left to chance. A single mistimed 15-minute reading can lock in a building's grid costs for the next twelve months. Once you understand that mechanic, you can act on it even in existing buildings. What it takes is clear priorities, better-tuned system operation and real operating data.

For housing companies running a central heating plant, supplementary heat pumps and rising common-area electricity demand, this is no longer a question for the future. The connection obligation for controllable loads under Section 14a of the German Energy Industry Act (EnWG) has applied since early 2024, and the statutory smart meter rollout will deliver, over the coming years, the data basis needed to make peak loads in your own portfolio visible at all.

Whether load management stays a tedious compliance task or genuinely saves money comes down to two things: a control strategy that actually works day to day, and the metrics you feed it with.

Why Is a Peak Load in a Multi-Family Building So Costly?

A peak load turns expensive because of the billing logic itself. Buildings with registering power measurement (RLM, typically required from around 100,000 kilowatt-hours of annual consumption) pay a grid fee made up of a usage-based rate and a demand charge calculated from the single highest 15-minute average recorded across the entire billing year, as Enerkii's explainer on commercial grid fee structures lays out. One peak value therefore determines what a building pays for its grid connection for a full twelve months, no matter how low the load stays across the roughly 35,000 other quarter-hours of the year.

A documented case from the commercial sector makes this tangible: an annual peak load of 550 kilowatts, billed at €166.03 per kilowatt, produced a demand charge of roughly €91,300. This example applies to a commercial property, distinct from a typical residential average, illustrating the same mechanism that also affects larger multi-family portfolios billed under RLM.

Whether it pays off more to lower the baseline load or to target individual peaks for shaving depends on annual utilization hours. Below 2,500 full-load hours per year, a high usage rate of 6 to 9 cents per kilowatt-hour dominates, paired with a low demand charge of around €20 per kilowatt. Above that threshold, the ratio flips: the usage rate drops to around 1 cent, while the demand charge often exceeds €150 per kilowatt. For many multi-family buildings running heat pumps continuously, targeted peak shaving is therefore the strongest economic lever.

What Triggers Recurring Peak Loads in Existing Buildings?

Most peak loads occur where several demands converge within the same quarter-hour. In the morning or during cold spells, boilers and heat pumps often draw power at the very same moment as domestic hot water production and common-area electricity.

Since January 1, 2024, Section 14a EnWG has obliged grid operators to generally allow the connection of controllable loads such as heat pumps, storage systems and charging stations from 4.2 kilowatts upward. In return, grid operators may throttle these systems down to a minimum output of 4.2 kilowatts, or 40 percent of connection capacity, during an acute grid bottleneck. When several heat pumps are controlled through a shared energy management system, that minimum output is weighted with a simultaneity factor between 0.8 and 0.45: two heat pumps at 17 kilowatts each, plus five charging stations, work out to a minimum output of roughly 13.6 kilowatts, or 40 percent of the grid connection capacity.

Which Metrics Reveal Peak-Load Risk Early?

Five metrics reliably show where peak loads originate inside a building. The load profile at 15-minute resolution reveals exactly when the actual power peak occurs, while heat pump run times and cycling frequency, together with flow temperature over the heating season, indicate whether a system is running cleanly.

As a rule of thumb, a correctly sized heat pump starts one to four times per hour in winter operation, with run times of 15 to 30 minutes each. Once a system regularly exceeds eight to twelve starts per hour, or run times drop below five minutes, that counts as critical short-cycling with a higher risk of wear. A properly sized buffer tank can significantly cut this switching frequency; a common benchmark today is around 10 to 20 liters of buffer volume per kilowatt of heating output.

Flow temperature is another early indicator. In a simulation study of a typical multi-family building from the 1958 to 1978 construction period with twelve residential units, lowering the flow temperature from 70 to 60 degrees alone raised the heat pump's seasonal performance factor from 3.0 to 3.2, as the practical guide from dena, GdW and BWP shows, without any renovation of the building envelope. Our article on digital heating optimization explains how a digital twin can keep this kind of operating data continuously visible.

Case study, Adorf (Saxony): Replacing 70 percent of the radiators and adjusting the heating curve lowered the flow temperature at 0 degrees outside temperature from 50 to 43 degrees, and raised the performance factor of the existing heat pump and boiler hybrid system from 3.0 to an average of 3.5, across five residential units totaling 280 square meters of living space.

Which Levers Cut Peak Loads Without Major Renovation?

Four levers work in existing buildings without touching the heating plant or the building envelope. The first is clear prioritization of consumers, the second is adjusted system operation. The third shifts loads in time, and the fourth makes them permanently controllable through continuous monitoring and automated regulation.

A clear ranking determines which consumer steps back first when a peak-load risk appears. Hot water charging can usually wait a few minutes, and a heating circuit with high thermal inertia tolerates a short pause better than a charging point with a contractual minimum output. An energy management system that controls several heat pumps and charging points together can apply exactly this ranking automatically, the same logic already built into the simultaneity factor under Section 14a EnWG.

How the heat pump itself is operated is the second lever. A lower flow temperature reduces performance-factor losses and, at the same time, lowers the power peak at restart after a night setback. A correctly sized buffer tank also decouples generation from demand and dampens cycling, as described in the previous section.

The third lever shifts loads in time. Since April 2025, operators of controllable loads with a smart metering system have been able to use time-variable grid fees under Module 3, which tier electricity use into low, standard and high-load windows. For a household with a heat pump and a charging station, one sample calculation shows savings of up to €305 per year. Scaled up across an entire building, that adds up quickly once you shift hot water and storage charging deliberately into the cheaper windows.

The fourth and most decisive lever is continuous monitoring. Only ongoing visibility into load profiles, cycling patterns and temperature curves lets you catch a looming peak before it shows up on the bill. KUGU VIS Betriebstransparenz (VIS stands for Visuelles-Informationssystem, or visual information system) makes exactly these connections between systems, loads and time patterns visible, while KUGU EOS (Energie-Optimierungssystem, or energy optimization system) turns that data into an automated operating strategy that applies prioritization and time-shifting during live operation. Moving from gut feeling to data-driven control is more than a software question, as our article Heating with a Plan shows. Without this step, prioritization has nowhere to run, however you look at it. And without a connected, digitized system that actually delivers operating data in the first place, any control approach stays theoretical, as described in our article on heating system connectivity.

What Does Load Management Deliver Economically?

Load management pays off in more than one way. You avoid peak loads driving up your demand charge, system operation becomes more predictable, and you build a more solid foundation for adding heat pumps or further sector coupling later on.

The benefit of peak shaving scales with building size, and it's no longer a lever reserved for large properties alone. Both figures are benchmarks drawn from documented sample calculations, not a guaranteed outcome, since actual savings depend on each building's individual load profile. Once you know your own load profile, you can time heat pump and storage operation to complement each other, a basic precondition for sector coupling in existing buildings.

Why Does Load Management Fail Without Data and Ongoing Monitoring?

Load management most often fails because of missing or outdated consumption data, less often because of the technology itself. Without remotely readable meters and continuous monitoring, any prioritization stays a one-off snapshot.

The statutory smart meter rollout has been running in stages since January 1, 2025: at least 20 percent of mandatory installation cases had to be equipped by the end of 2025, rising to 50 percent by 2028 and at least 95 percent by 2030. Owners of buildings with a central heating and hot water system and at least two residential units face their own deadline, as GeVestor's overview of the 2026 smart meter obligation describes: meters that are not remotely readable must be retrofitted or replaced by the end of 2026, and from 2027 only remotely readable meters will be permitted there.

For the housing industry, another shift is coming: in 2026, the Building Energy Act (GEG) will be replaced by the Building Modernization Act (GModG), another reason to get your data foundation in order now, before the next round of regulatory adjustment arrives. Of the roughly 3.3 million existing multi-family buildings in Germany, only about 3.3 percent are currently heated with heat pumps, while more than half still run on central fossil-fuel boilers. In exactly this building stock, the quality of your data determines whether a future heat pump retrofit can run low on peak loads from day one, or whether it has to be readjusted later during live operation.

Are You in Control, or Are You Being Throttled?

The real point of Section 14a EnWG is often missed in practice: if you don't have your own peak load under control, you risk the grid operator throttling your systems externally to a minimum output during a bottleneck, with noticeable consequences for heat pump and hot water comfort in the affected apartments.

Know your own load profiles, on the other hand, and you decide which consumer steps back and when, keeping both grid costs and living comfort in your own hands. Without that data, it's the grid operator who ultimately decides when and how hard your systems get throttled.

The first step is rarely a new system. Start instead with the load profile and operating data you already have for your properties: where does the annual peak sit, when does it occur, and which consumers are actually responsible for it? That's the basis for deciding whether prioritization, adjusted operation or a storage system offers the biggest lever for each individual building.

At What Point Does a Multi-Family Building Count as an RLM Customer With a Demand Charge?

Around 100,000 kilowatt-hours of annual consumption is the usual threshold, provided a registering power meter is installed. From that point on, the grid fee consists of a usage-based rate plus a demand charge based on the single highest 15-minute value of the entire billing year. Larger multi-family portfolios running a central heat pump regularly exceed this threshold.

Do I Have to Install a Battery Storage System for Load Management?

No, a storage system is only one of several levers, not a requirement. Prioritizing consumers and adjusting the flow temperature often lower peak loads without any extra hardware, complemented by time-shifting under Module 3 of the grid fees. A storage system pays off mainly where the annual peak stays noticeably above an economically sensible level even after these measures.

What Does Section 14a EnWG Mean in Practice for Heat Pumps in Multi-Family Buildings?

Since January 1, 2024, grid operators have generally had to allow the connection of controllable heat pumps, storage systems and charging stations from 4.2 kilowatts upward, as set out in Section 14a EnWG. In return, the grid operator may throttle output to a minimum of 4.2 kilowatts, or 40 percent of connection capacity, during an acute bottleneck. When several heat pumps are jointly controlled, that minimum output is weighted using a simultaneity factor.

How Quickly Does an Adjusted Heating Curve Affect Peak Loads?

The effect is usually noticeable within the first heating season after the change, mainly through smaller start-up peaks following night setbacks. A documented case in Saxony combined a radiator swap with an adjusted heating curve, lowering the flow temperature at 0 degrees outside temperature from 50 to 43 degrees and raising the performance factor of the heat pump and boiler system from 3.0 to an average of 3.5. The building envelope stayed untouched throughout.

Is a Smart Meter Alone Enough to Manage Peak Loads?

No, a smart meter only measures, it doesn't control anything by itself. Actual control only comes from an energy management system that evaluates load profiles and prioritizes consumers. From 2026 onward, buildings with a central heating system and at least two residential units will have to switch to remotely readable meters regardless.