Atypical Grid Usage: Germany's §19 Network-Fee Reductions, Explained

The two routes in 30 seconds

Route 1: atypical usage, avoid the windows, cut the fee

Every distribution grid operator publishes high-load windows (Hochlastzeitfenster) for the coming year: the hours in which its network expects peak stress. If your site's maximum draw during those windows stays significantly below its overall annual peak, you are using the grid atypically, and the operator must offer you an individual network fee, no lower than 20% of the published one.

The windows are concrete and knowable in advance: published per operator and voltage level, seasonal, and valid on working days only, weekends and holidays count as off-peak time. That predictability is the whole point. The regime rewards sites that can commit to keeping their grid draw low exactly when the network is stressed, and it measures that commitment against the published windows, not against good intentions.

What "significantly below" means in numbers is set out in the regulator's determination and your operator's published methodology, the thresholds differ by voltage level. The practical test is always the same: lay your quarter-hour load profile over the operator's windows and look at the gap between your in-window peak and your overall peak.

Route 2: intensive usage, the 7,000-hour club

The second route ignores when you consume and looks at how steadily. A site that draws more than 10 GWh a year and reaches at least 7,000 utilization hours, annual consumption divided by annual peak, qualifies for an individual fee of at most 20% of the published tariff. At 7,500 hours the floor drops to 15%, at 8,000 hours to 10%.

Utilization hours are a flatness measure: 8,760 would mean perfectly constant draw all year. The formula has an important consequence, the denominator is your annual peak. A single bad quarter-hour doesn't just cost demand charges; it can push utilization hours below a §19 threshold and forfeit the entire reduction for the year.

What a battery does to each route

For atypical usage, a battery is the natural instrument: it discharges during the published windows so the site's grid draw stays below the threshold, and recharges in off-peak time. That is a scheduling problem with known windows, exactly the shape of problem an EMS solves, and it stacks naturally with peak shaving, since the windows and the site's own peaks often overlap.

For intensive usage, the arithmetic is more interesting: shaving the annual peak raises utilization hours, the numerator stays, the denominator falls. On paper, a site hovering just under 7,000 hours moves closer to the threshold by capping its peaks. Whether a specific site's storage operation holds up in that calculation is a question for the notification process and a proper feasibility review, treat it as arithmetic worth checking, not a result to bank on.

Two honest caveats. Capacity committed to holding windows or peaks is capacity not sold into markets, the trade-offs need daily pricing, not a set-and-forget rule. And the individual fee is agreed per offtake point and notified to the regulator; changing your load behavior mid-year has consequences. This is feasibility-study territory, not a checkbox.

The process, briefly

• Check the windows: your grid operator's published Hochlastzeitfenster for the year, per voltage level.
• Test your profile: quarter-hour load data against the windows (atypical) or utilization hours and annual consumption (intensive).
• Agree and notify: the individual fee is agreed with the grid operator and notified to the regulator; the regulator can reject non-compliant agreements.
• Keep it true: the reduction holds only while the load behavior does, monitoring is not optional.

Primary sources: § 19 StromNEV (statute text) · your grid operator's published high-load windows. Related guides: peak shaving · EMS vs BMS vs SCADA · C&I battery storage, explained.