Peak Shaving Economics: How Much Can a Factory Save with BESS?

A Factory Does Not Save Money Because the Battery Discharges

A factory saves money with a peak-shaving BESS only when battery discharge reduces the demand value that actually appears on the electricity bill. The first-pass annual saving is therefore avoided demand charge − charging cost − conversion losses − auxiliary energy − degradation cost − maintenance . High demand charges and short, repeatable peaks usually create a stronger business case than broad, multi-hour load plateaus.

That sounds obvious until someone shows me a proposal saying:

“500 kWh BESS saves 30% on electricity.”

My first question is not about the battery.

“Thirty percent of which charge?”

The U.S. Department of Energy distinguishes energy charges, billed in kWh, from demand charges, commonly based on maximum kW during a billing period. It also notes that interval data—often hourly or 15-minute—is needed to see the peaks that monthly consumption totals hide. DOE

NREL reached a similar conclusion from a different direction: in its analysis of commercial battery storage economics, peak demand charges were the strongest predictor of whether a BESS was economically viable, while load-profile shape was the strongest predictor of battery size . NREL

That is where I start the spreadsheet.

Not with battery price.

The Factory Paying $18/kW-Month Gets My Attention

Consider a hypothetical factory. These are worked-example assumptions, not a Ruibit/Dawnice customer case.

The plant has:

Monthly peak demand: 1,020 kW

Target after peak shaving: 800 kW

Demand charge: $18/kW-month

Peak reduction required: 220 kW

Assume the BESS successfully holds that reduction in all 12 billing months.

The gross demand-charge saving is:

220 kW × $18/kW-month × 12 months = $47,520/year

That's the number people like putting in presentations.

I don't.

Not yet.

Because $47,520 is avoided demand charge, not net BESS savings .

Then I Open the Load Curve and the Nice Number Usually Changes

Suppose the 220 kW peak reduction is required for only 45 minutes on a typical peak day.

Energy delivered during the event is:

220 kW × 0.75 h = 165 kWh

If the peak instead lasts three hours:

220 kW × 3 h = 660 kWh

Same demand reduction.

Same theoretical $47,520 annual demand-charge saving.

Very different battery.

This is why peak duration has such a strong influence on economics. NREL notes that longer-duration peak-shaving systems generally cost more because additional battery energy is required. NREL

A factory with a sharp 30–45 minute production peak may be a much more interesting BESS candidate than another factory with the same 1,020 kW maximum demand spread across a four-hour plateau.

The electricity bill shows the same peak.

The battery sees the width.

Missing One Peak Can Cost More Than Losing One Cycle

Now we get to the part I worry about as an energy manager.

Some tariffs contain demand ratchets or look-back provisions. DOE notes that these can calculate current demand charges partly from previous peaks, meaning an occasional high spike may affect bills beyond the month when it occurred. DOE

So imagine the BESS performs perfectly for 11 months.

In August, SoC is too low when production hits the annual peak.

The factory reaches 1,015 kW instead of staying near 800 kW.

Depending on the tariff, that is not simply “one bad dispatch.”

It may reset part of the economic baseline.

This is why I don't value a peak-shaving system from:

battery capacity + demand charge

alone.

I also want:

EMS logic

load forecast

minimum reserve SoC

meter interval

recharge strategy

PCS power

and the exact tariff definition of billing demand.

A battery can be technically healthy and financially useless for fifteen very expensive minutes.

The Charging Bill Comes Back Through Another Door

Suppose the factory needs 165 kWh AC-side discharge during each shaving event.

For an illustrative calculation, assume 90% round-trip efficiency .

Energy that must be purchased to restore the battery is approximately:

165 ÷ 0.90 = 183.3 kWh

If charging electricity costs $0.09/kWh :

183.3 × $0.09 = $16.50 per event

At 250 peak-shaving events per year:

$16.50 × 250 = $4,125/year

Now our simple economics looks more like this:

That $43,395 is still not project profit .

It is merely a cleaner number than $47,520.

And this is exactly why I am suspicious of BESS proposals that jump directly from demand-charge savings to payback period.

Degradation Is an Expense Even When Nobody Sends an Invoice

A battery cycle does not arrive with a receipt attached.

It still has a cost.

Suppose a factory saves another $8,000 per year by cycling more aggressively for energy arbitrage, but that dispatch increases throughput enough to accelerate capacity loss or reach warranty throughput limits earlier.

Was the extra revenue worth it?

Maybe.

But the EMS should not chase every price spread merely because discharge is technically possible.

NREL's storage modeling shows that dispatch strategy can materially change project NPV, particularly where tariffs vary by time. Peak shaving and price-signal dispatch are different control objectives and can produce different operating behavior. NREL

For a factory project, I would therefore model at least:

annual discharged MWh

expected round-trip efficiency

capacity fade

warranty throughput

augmentation/replacement assumptions

not merely cycle count.

A shallow 20-minute peak-shaving event and a deep daily arbitrage cycle should not be treated as economically identical just because both appear as “one cycle” in a sales presentation.

The Cheapest Battery Can Produce the Worst Payback

Let's use another hypothetical comparison.

Two suppliers quote systems capable of the same 220 kW peak reduction:

System B stores twice as much energy.

It also earns only $3,000 more in this example.

If the factory's real peak lasts less than one hour, the extra 250 kWh may spend most of its life doing nothing financially useful.

This is why economic sizing is not the same as maximum sizing .

When Ruibit/Dawnice receives a peak-shaving inquiry, I would rather send the factory's interval data and tariff than request “500 kWh because that is our budget.”

Dawnice currently offers commercial configurations including 100 kW/200 kWh and 125 kW/265 kWh systems, alongside larger commercial and containerized storage platforms. Dawnice

Those are equipment configurations.

The factory's peak decides whether any of them makes economic sense.

Solar Can Improve the Calculation, but Only If the Timing Works

Factories with rooftop PV sometimes assume solar automatically makes peak shaving more profitable.

Not necessarily.

Imagine the factory peak occurs at 17:30 , after PV production has fallen sharply.

Solar may still charge the battery earlier, which can reduce charging-energy cost or increase solar self-consumption.

Useful.

But it does not eliminate the need for enough PCS power and reserved energy at 17:30.

Conversely, if PV already reduces the midday peak below the demand threshold, the BESS may have very little peak-shaving work left to do.

NREL notes that PV can shorten net peak duration in some systems, which can improve storage economics for peak shaving. NREL

The important word is can .

I would overlay PV generation and factory demand at the same interval resolution before assigning the value to the battery.

I Don't Calculate Payback Until These Seven Numbers Stop Moving

For a preliminary factory peak-shaving business case, I want:

Then I add degradation, maintenance, auxiliary consumption, financing, taxes, incentives, warranty limits, and residual value as the project becomes more serious.

Only after that do I calculate:

Simple Payback = Installed Project Cost ÷ Annual Net Savings

For example, if the fully installed project costs $220,000 and modeled annual net savings after operating costs are $40,000 :

Simple Payback = $220,000 ÷ $40,000 = 5.5 years

Useful number.

Not enough for an investment decision.

A serious project should also examine NPV, IRR, degradation, tariff sensitivity, and what happens if the peak changes after production expansion.

Sometimes My Recommendation Is Not to Buy the Battery

This is the part that rarely appears in a quotation.

Suppose I open the 15-minute data and discover the factory peak comes from one compressor starting unnecessarily while two production lines overlap.

Rescheduling that equipment costs almost nothing.

Or perhaps the demand charge is only $4/kW-month .

Or the “peak” lasts five hours every working day, forcing so much battery energy that the economics collapse.

I would rather find that before somebody orders cabinets.

Peak shaving is attractive when an expensive demand charge meets a load profile the battery can attack efficiently. NREL's analysis supports exactly that relationship: tariff and load shape are central economic drivers, not background details. NREL

So when a factory asks:

“How much can BESS save us?”

I would not answer with a percentage.

Give me the tariff.

Give me the interval data.

Give me the installed project cost.

Then we can put a price on the peak.

A BESS does not create peak-shaving value by being large. It creates value by removing the right kilowatts from the right billing interval at a lower lifetime cost than those kilowatts were going to cost the factory.

FAQs

1. How much can a factory save with BESS peak shaving?

Savings depend mainly on avoided peak demand, demand-charge rate, peak duration, tariff rules, BESS efficiency, and operating costs . There is no reliable universal savings percentage.

2. How do you calculate peak-shaving savings?

A simple starting formula is:

Annual Gross Savings = Peak Reduction (kW) × Demand Charge ($/kW-month) × Applicable Months

Then subtract charging energy, losses, auxiliary consumption, maintenance, and degradation-related costs.

3. Why does peak duration affect BESS economics?

Peak height determines required PCS power, while peak duration determines battery energy capacity. A short 200 kW spike may require far less kWh than a 200 kW reduction sustained for several hours.

4. Does a larger BESS always produce greater peak-shaving savings?

No. Once the battery can cover the economically valuable peak, additional capacity may provide little extra demand-charge reduction. Oversizing can increase CAPEX and lengthen payback.

5. What data is needed before calculating factory BESS payback?

Buyers should provide interval load data, demand charges, energy tariffs, peak duration, charging windows, transformer capacity, BESS efficiency, installed cost, degradation assumptions, and maintenance costs .