Peak shaving and reactive power: the two invisible savings that move the industrial bill

Most companies look at their electricity bill and think about one thing: consumption. How many kWh did I burn this month. And yet, for an industrial customer on the Spanish 3.0TD or 6.1TD tariff, the kWh cost is less than half of the final bill. The other half —sometimes more— breaks down into capacity charges, excess penalties, reactive power and deviation costs. And right there, every single month, there’s free money on the table for whoever knows how to optimise it.

Two concrete techniques move the needle: peak shaving (flattening demand spikes) and reactive power compensation (correcting the power factor). They sound like electrical jargon, but the mechanics are simple and the ROI is usually under 3 years for a mid-sized industrial site.

What peak shaving is and why the industrial bill is so sensitive to spikes

On the 3.0TD or 6.1TD tariffs (Spain), you pay for the power you contract across each of the 6 P1-P6 periods separately. And here’s the hard part: if your machinery requests at any instant more power than contracted, the grid operator applies an excess-power penalty calculated on the square of the difference and billed for months. A one-off 200 kW spike above your 500 kW contracted can cost more than the whole month’s energy consumption.

Peak shaving is exactly the opposite of that pain: identify the moments when your demand spikes, and cut those peaks in time —either by shutting something down, delaying a startup, or, most elegantly, injecting energy from a battery during those critical 5-20 minutes.

How a peak is shaved in practice

Three mechanics, usually combined:

1. Stationary battery + EMS (Energy Management System). The EMS monitors instantaneous active power. When it detects you’re about to cross the configured threshold (say, 90 % of P1 contracted power), the battery discharges just enough so the external meter never sees the peak. Typical investment 100-300 kWh of storage for a mid-sized warehouse.

2. Programmed demand response. You identify which loads aren’t real-time critical —compressors with their own buffer, fleet charging, HVAC with thermal inertia— and automate them to shut down or throttle when the meter approaches the threshold. Low cost (PLC + sensors), but only works if you have flexible loads.

3. Daytime solar production + self-consumption. If your critical peak coincides with central daytime hours (process lines, ovens, irrigation pumping), your PV plant reduces net demand at the meter exactly when it hurts most. It’s “free” peak shaving with no battery needed.

A numerical case that hits home

A logistics warehouse in Castilla-La Mancha (Spain), 3.0TD tariff, 400 kW contracted in P1. They measure demand with a network analyser for 30 days and discover 12 episodes per month where they cross 440 kW for 8-15 minutes —when the loading dock compressors and cold storage chambers start up simultaneously.

• Excess penalty that month: ~€3,200
• Projected annual penalty: ~€38,000

Solution: 150 kWh / 100 kW battery with EMS, €65,000 investment. Projected ROI: 1.7 years just from avoiding that penalty alone, without counting other downstream savings.

What reactive power is and why it should never be on your bill

In any industrial installation with motors, transformers, welders, large air conditioners or any inductive load, the current has two components: the one that does useful work (active power, kW) and the one that just sloshes back and forth maintaining the magnetic field of motors without producing anything (reactive power, kVAR). The vector sum is the apparent power (kVA), which is what the grid actually has to supply.

The problem is that your cable, your transformer and the distributor’s line have to be sized for kVA, not for kW. If a motor draws 100 kW of useful work but pulls 30 kVAR of reactive, you’re loading the grid with 105 kVA even though you only use 100 kW. And the grid operator bills you for that inefficiency.

Power factor and the reactive penalty

The power factor (cosφ) is the ratio kW / kVA. If your cosφ is 1, all the current is useful work. If it drops to 0.8, 20 % of your apparent consumption is useless reactive.

Spanish regulations (Real Decree 1164/2001 and later) require cosφ ≥ 0.95 during peak and off-peak hours. Below that, the penalty scales:

• cosφ between 0.80 and 0.95 → surcharge of 0 to 35 % on the energy term
• cosφ < 0.80 → fixed 50.7 % surcharge
• Leading cosφ (rare but happens in plants with poorly sized capacitor banks) → also penalised

For an industry spending €80,000 a year on energy, a bad cosφ of 0.82 left uncorrected is ~€20,000/year thrown in the bin.

How to eliminate it (the solution is very direct)

Capacitor bank with automatic regulator. It’s been the standard equipment for this for 50 years:

• Measures cosφ in real time
• Switches steps of capacitors in parallel with the factory’s internal grid
• Each step “returns” inductive reactive to the system
• Result: the grid operator’s meter sees cosφ ≥ 0.98

Typical investment for a 100-500 kW site: between €3,000 and €12,000. ROI almost always under 12 months if your current cosφ is below 0.90.

For higher power and variable loads, this is complemented with active harmonic filters —which also improve internal power quality, extend the life of sensitive electronics and reduce cable heating.

When to invest and when not

The two optimisations don’t apply equally to all companies. The quick rule we use:

• You have a peak shaving problem if: your demand curve shows concentrated peaks that exceed your contracted power more than 4-5 times a month, or you want to lower your contracted power and the ceiling is your current pain point.
• You have a reactive problem if: your bill shows a reactive surcharge on any month, or your cosφ is below 0.95 (printed on the bill as information, though sometimes you have to ask your retailer for the analysis).

If the answer to either is yes, there’s money waiting.

The bonus when you already have (or will have) solar self-consumption

Here’s the neat connection: a PV plant with modern inverters does both jobs at once at no extra cost:

• Natural peak shaving: during central daytime hours, your solar production reduces net demand at the meter, flattening the characteristic peak of daytime processes.
• Integrated reactive compensation: current grid-tied inverters (≥ 30 kVA) can produce controlled reactive power. The EMS configures them so that, in addition to injecting active solar power, they keep cosφ ≥ 0.98 as seen from the grid. You skip the traditional capacitor bank.

Result: if you combine PV self-consumption + stationary battery + EMS, in a single installation you cover (a) solar production, (b) peak shaving, (c) reactive compensation, (d) electrical backup against outages. The combined ROI usually drops to 3-5 years, versus 7-10 for a solar-only project without EMS.

How we approach it at AUREQIS

For any industrial project the process is always the same:

1. Audit with a network analyser for 7-15 days. We measure active power, reactive, power factor, harmonics and events. Without measuring, you don’t design.
2. Analysis of the last 12 months of bills. We identify real penalties, not estimates: how much you’re paying today for excess power and reactive.
3. Joint sizing: PV (kWp), battery (kWh / kW), Q-control inverters, EMS. All in a single integrated project, not as loose components.
4. Full engineering and grid permitting.
5. Industrial leasing 100 % tax-deductible or outright purchase, whichever you prefer. The monthly leasing fee is designed to be lower than the estimated monthly savings from month 1.

If your monthly industrial bill is above €3,000 and you have recurring penalties for excess power or reactive, there’s an 80 % probability that the ROI will be under 24 months. Tell us your case and we’ll set up the audit with no commitment.