A factory may want one battery energy storage system to do two jobs.

On an ordinary production day, the BESS may discharge for a short period to keep grid demand below a costly billing peak. On another day, the same system may be expected to keep essential equipment running after utility power disappears.

The equipment can look identical on a proposal sheet, but the design priorities are different.

For peak shaving battery storage, the starting point is the facility’s grid-demand profile:

How much power must the BESS remove from the metered peak, and how much energy is required to hold that limit?

For backup power, the starting point is inside the facility:

Which loads must remain online, how much power do they require together, and how long must they operate without the grid?

Those questions influence the required kW, kWh, state-of-charge reserve, PCS operating mode, EMS strategy and ultimately the battery storage ROI.

If the distinction between power and energy is still unclear, it helps to separate the two before comparing applications. This article focuses on the next decision: what changes when the same storage asset is being considered for peak shaving, backup power, or both.

Peak Shaving and Backup Power Do Not Start From the Same Data

The most useful comparison is not simply “short-duration battery versus long-duration battery.”

It is load curve versus critical-load curve.

Decision Point	Peak Shaving BESS	Backup Power
Main objective	Reduce maximum grid demand	Maintain critical operations
Starting data	Interval load data and tariff	Critical-load list and outage requirement
Main power question	How far does metered demand rise above the target?	What is the highest simultaneous critical load?
Main energy question	How much energy lies above the target threshold?	How much energy do critical loads require during the outage?
Normal trigger	Load approaches the selected demand limit	Grid failure or unacceptable grid condition
SOC priority	Use stored energy to create economic value	Preserve enough energy for resilience
Main financial input	Demand charge, tariff spread, operating schedule	Downtime cost, outage risk, alternative backup cost
Main control issue	Avoid peaks without creating another peak while recharging	Protect critical loads and manage off-grid operation
Common mistake	Buying plenty of kWh but not enough kW	Backing up the entire factory instead of the loads that actually matter

Peak shaving follows what the utility meter sees. Backup follows the loads the operations team cannot afford to lose.

That distinction is more useful than assuming every peak-shaving project needs a two-hour battery or every backup project needs four hours of storage.

For Peak Shaving, Look at the Portion of the Load Curve You Want to Remove

Suppose a factory reaches 720kW during a production peak but wants to keep its metered grid demand at or below 600kW.

At the highest point, the BESS needs roughly:

720kW − 600kW = 120kW

A system with 500kWh of battery energy but only 60kW of continuous AC discharge capability cannot fully shave that peak. The battery may contain enough energy, but the PCS cannot deliver it quickly enough.

Energy is the second part of the problem.

If the factory remains above 600kW for two hours, a quick calculation might be:

120kW × 2h = 240kWh

That assumes the plant remains at 720kW for the entire two hours. Real factory load profiles are usually less tidy.

If the required BESS output across four half-hour intervals is 60kW, 120kW, 90kW and 30kW, the theoretical event energy is:

30 + 60 + 45 + 15 = 150kWh

The system still needs to reach 120kW, but it does not need to supply 120kW continuously for the whole event.

This is why a peak-shaving study should use interval data and the actual billing-demand rules rather than one monthly peak figure. A brief instantaneous spike may not be treated the same way as a sustained 15- or 30-minute average under the customer’s tariff.

The complete methodology—including SOC, efficiency, end-of-life capacity and full load-curve analysis—is covered separately in HighJoule’s Commercial Battery Storage Sizing Guide.

Peak Shaving Is Not the Same as Load Shifting

The two strategies often operate on the same battery, but they create value in different ways.

Peak shaving is mainly concerned with controlling kW at the grid connection. The battery responds when facility demand approaches a defined threshold.

Load shifting moves kWh from one period to another. The BESS may charge when electricity is cheaper, when factory demand is low or when excess PV generation is available, then discharge later.

A facility may need relatively little energy to remove several short demand spikes but significantly more capacity to move several hours of solar production into an evening operating window.

That difference matters when a proposal combines both functions under one headline battery size.

Backup Power Starts With the Critical Loads

Now take the same factory.

Its normal peak is 720kW, but that does not mean the backup system has to provide 720kW.

During an outage, the site may stop non-essential equipment and retain only the loads needed for safe shutdown, process continuity or minimum production.

A simplified critical-load audit might look like this:

Critical Load	Required Power
Process controls	20kW
Essential production equipment	35kW
Cooling and ventilation	25kW
IT, communication and emergency lighting	10kW
Total	90kW

If the requirement is four hours of battery-only backup, with no PV or generator contribution during the outage:

90kW × 4h = 360kWh

That is the theoretical delivered-energy requirement before project-specific operating limits and losses are considered.

Compare that with the earlier peak-shaving example:

• Peak shaving: approximately 120kW / 150kWh
• Backup: approximately 90kW / 360kWh

In this particular example, peak shaving needs more instantaneous power, while backup needs much more energy. Another factory could produce the opposite result, especially if its critical-load group is large or its demand peak lasts several hours.

Backup Power Still Has a kW Problem

Describing backup as “kWh-first” is useful shorthand, but it can become misleading.

Imagine the 90kW critical-load group includes a compressor or motor with a demanding start sequence. A battery can contain enough energy for several hours and still fail to support the equipment if the PCS or system architecture cannot handle that power event.

Backup design may therefore need to consider continuous critical-load power, temporary overload requirements, motor starting, load sequencing, acceptable transfer interruption, off-grid voltage and frequency control, protection coordination and generator interaction.

A generic motor-starting multiplier should not replace actual equipment data.

Energy determines how long the site can continue operating. Power determines whether the required equipment can operate at all.

A Grid-Tied Peak-Shaving BESS Is Not Automatically a Backup System

This is an important boundary in commercial storage projects.

A battery may work perfectly for grid-connected peak shaving and still be unable to energize an isolated critical-load bus after the grid fails.

The EMS alone does not create backup capability.

Depending on the project, backup operation can require a PCS that supports the required off-grid or grid-forming mode, transfer equipment such as an STS or ATS, a defined critical-load bus, suitable protection and interlocking, acceptable transfer time, and coordination with PV or a generator.

HighJoule’s 125kW/261kWh liquid-cooled BESS cabinet, for example, is specified for grid-connected and off-grid applications, with optional STS-based switching.

That capability still has to be engineered into the site’s electrical architecture.

A statement such as “the battery contains enough kWh for backup” therefore answers only part of the question. The installation must also be designed to operate in the required backup mode.

What the Same 125kW/261kWh Cabinet Tells Us

The HighJoule HJ-G0125-0261L is useful for comparing the two applications because the cabinet stays the same while the operating requirement changes.

Its published ratings include:

• 125kW rated output power
• 261.248kWh rated energy
• 314Ah LFP cells
• liquid cooling
• IP55 enclosure protection

The nominal energy-to-power ratio is:

261.248kWh ÷ 125kW ≈ 2.09 hours

That places it in the two-hour class from a nameplate perspective. HighJoule also discusses this configuration in its 125kW/261kWh BESS guide.

The duration figure alone, however, does not tell us which application it fits.

In the Peak-Shaving Example

The hypothetical requirement was approximately:

120kW maximum discharge + 150kWh event energy

A 125kW/261.248kWh cabinet therefore falls within the right general power-and-energy range for further evaluation.

That does not mean one cabinet is automatically the final design. Usable energy, SOC policy, discharge losses, auxiliary consumption, recharge opportunity and required end-of-life performance still need to be checked.

In the Four-Hour Backup Example

The requirement was:

90kW × 4h = 360kWh delivered energy

A single 261.248kWh cabinet does not contain enough nominal battery energy to satisfy that requirement before losses or reserve are even considered.

At a constant 90kW load:

261.248kWh ÷ 90kW ≈ 2.90 hours

Actual battery-only runtime at the site would be lower once the project’s usable-energy boundaries are applied.

The same cabinet can therefore be a plausible candidate for the peak-shaving example and clearly insufficient for the four-hour battery-only backup example. The difference comes from the operating requirement, not from the product changing.

A Hybrid System Creates a Conflict Over Stored Energy

Many factories want lower electricity costs and resilience from the same asset.

That is possible, but the battery has only one pool of stored energy.

Imagine a BESS enters the afternoon at a high SOC. Production rises and the EMS begins discharging to keep grid demand below the target. By late afternoon, a large share of the economically available energy has been used.

Then the grid fails.

If normal dispatch was allowed to consume nearly all available capacity, the remaining backup time may be far shorter than the operator expected.

For a dual-purpose system, part of the usable energy normally has to be protected.

Conceptually:

Usable energy = economic dispatch range + protected resilience reserve

The appropriate reserve is project-specific. It depends on critical-load size, required backup duration, outage risk, PV contribution, generator availability, tariff value and the time available to recharge.

A fixed 30% reserve may be reasonable at one site and inappropriate at another.

Holding more energy back improves resilience but reduces the capacity available for daily peak shaving and load shifting. That tradeoff belongs in the financial model.

The EMS Allocates Energy, but It Does Not Replace Electrical Design

HighJoule’s EMS Energy Management System can coordinate storage scheduling, equipment monitoring and multi-source energy management.

For a dual-purpose system, one possible control sequence is:

Site Condition	Possible Control Priority
Normal operation	Maintain planned SOC range
PV surplus	Charge when battery and grid limits allow
Demand approaching target	Discharge without violating protected reserve
Peak has passed	Recover SOC when load and tariff conditions are suitable
Grid outage	Stop economic dispatch and prioritize critical loads
Grid restored	Recharge without creating another demand peak

The sequence is not universal.

A food-processing facility protecting refrigeration, a plastics factory protecting production continuity and a commercial building protecting IT infrastructure will not necessarily use the same priority logic.

The EMS determines how available stored energy is allocated. The PCS, transfer equipment, protection and load architecture determine whether that energy can actually be delivered in the required backup mode.

Recharge Strategy Still Matters

Peak shaving does not finish when discharge stops. The battery has to recover its SOC.

If the BESS begins charging at high power while factory demand is still close to its billing threshold, charging can create a new peak.

That issue is often called a rebound peak.

It is primarily a control and scheduling problem. HighJoule’s kW vs kWh battery storage article discusses recharge power and rebound peaks in more detail.

For this comparison, the relevant point is that peak-shaving economics depend on the complete operating cycle—not only the discharge event.

Peak Shaving ROI and Backup Value Belong in Different Columns

A single “battery storage ROI” number can hide very different types of value.

Peak Shaving Has a Tariff-Based Value

A simplified relationship is:

Demand-charge savings ≈ successfully avoided billing demand × applicable demand-charge rate

The customer’s actual tariff matters more than a national electricity-price average.

Billing intervals, demand ratchets, time-dependent demand charges, seasonal rules or coincident-peak provisions can all change the result.

Load Shifting Has an Energy Value

A simplified relationship is:

Energy-shifting value ≈ energy moved × applicable price spread

The actual model should use the site’s charging and discharging conditions and applicable tariff.

Backup Power Is Mainly an Avoided-Loss Value

Backup value may come from avoiding lost production, restart delays, discarded work in process, missed deliveries, temperature-related product loss, data interruption or emergency generator operation.

Those are avoided losses rather than guaranteed monthly cash savings.

If a shutdown could cost $100,000, it does not follow that the BESS creates $100,000 of annual value. The analysis also needs to consider how frequently that event is expected to occur, how long it lasts and whether the battery would actually prevent the loss.

This difference is important when presenting the project to finance teams. Peak-shaving value may be visible directly in the tariff model; backup value is usually more dependent on operational risk.

Two HighJoule Projects Show Why the Objective Matters

The difference becomes clearer when looking at two actual industrial projects rather than adding another theoretical sizing formula.

Denmark: An Economic and Grid-Dispatch Application

In HighJoule’s Denmark 125kW/261kWh energy storage project, the system is used for operating objectives including peak shaving, frequency regulation and peak-valley tariff arbitrage.

In this type of project, the engineering discussion quickly moves toward dispatch rather than long-duration backup: when the cabinet should discharge, how much of the peak should be clipped, how much SOC should remain available for the next operating window, and when the battery should recover its charge.

The battery is expected to work regularly as part of the site’s economic and grid-management strategy.

That places considerable importance on load behavior, tariff timing and EMS control.

DRC Plastic Factory: Reliability Changes the Conversation

At HighJoule’s DRC plastic factory solar-plus-storage project, the grid itself is a major operating constraint.

The site uses a 125kW/261kWh storage system, a 125kW bidirectional PCS, approximately 100kW of customer PV and EMS coordination.

Here the engineering discussion is different. Battery SOC when an outage starts, available PV during the interruption, which production loads can be shed and how long essential processes must remain online matter more than tariff optimization alone.

Solar and storage work together when utility supply is unavailable.

The project’s continuity objective therefore comes from the integrated grid + PV + storage architecture. It should not be interpreted as 24 hours of battery-only runtime from a single 261kWh cabinet.

These two projects use the same general storage class, but the operating questions facing the engineering team are substantially different.

That is much more useful for system selection than simply labeling both projects as “125kW/261kWh BESS installations.”

When One Project Needs Both, the Interaction Matters

A factory wanting peak shaving and backup should not simply calculate peak-shaving kWh, add backup kWh and stop there.

The two functions interact.

The peak-shaving requirement and critical-load requirement may need different power levels. Energy protected for backup cannot simultaneously be assumed to earn full daily load-shifting value. An outage can also begin while the battery is already discharging or before it has fully recovered from the previous cycle.

PV and generators change the calculation again. A four-hour grid outage does not necessarily require four hours of battery-only energy if another reliable source contributes during that period, but the contribution needs to be modeled rather than assumed.

For some sites, reserve can also change with operating conditions. The owner might protect more energy during periods of higher outage risk and allow more economic dispatch when grid conditions are stable.

That can improve utilization, but only if the resilience requirement remains clearly defined.

Five Questions a Dual-Purpose BESS Proposal Should Answer

Detailed battery sizing requires a broader technical review. For a project combining peak shaving and backup, these five questions are especially useful when comparing proposals:

1. How much power is required for each operating mode?
   The proposal should distinguish peak-shaving kW, critical-load kW, continuous PCS output and any justified temporary overload requirement.

2. How much usable energy remains protected for backup?
   “The battery can also provide backup” is not enough. The proposal should define how many usable kWh remain available after normal economic dispatch.

3. What happens if the grid fails during a peak-shaving event?
   The expected SOC and critical-load runtime at that moment should be clear.

4. Is the complete system designed for islanded operation?
   Check the PCS mode, transfer equipment, critical-load connection, protection and interaction with PV or generators.

5. How are resilience and tariff savings separated in the ROI model?
   Measurable utility savings and risk-based avoided downtime should not be treated as the same type of cash flow.

These questions usually tell a buyer more than comparing two proposals only by price per kWh.

Which Strategy Should a Factory Start With?

If the grid is generally reliable but demand charges or grid limits are expensive, start with peak shaving battery storage and analyze interval load data against the actual tariff.

If the main business problem is production interruption, start with backup power and build a critical-load audit before selecting battery capacity.

If both cost reduction and continuity matter, use a hybrid strategy and model how much stored energy can participate in economic dispatch without compromising the agreed resilience target.

If the opportunity is mainly moving energy between tariff periods or storing excess PV production, focus on load shifting, while still checking whether charging and discharge power fit the available operating windows.

FAQ

What is peak shaving battery storage?

Peak shaving battery storage uses a BESS to reduce the maximum grid demand seen by the utility meter. The battery discharges when site demand approaches a selected limit.

Is peak shaving the same as load shifting?

No. Peak shaving primarily manages grid power (kW), while load shifting moves energy (kWh) between time periods. One BESS can perform both functions.

Can one BESS provide peak shaving and backup power?

Yes, provided the hardware and control architecture support both operating modes. The EMS must protect the required backup reserve, while the PCS, transfer system and electrical design must support the intended off-grid operation.

Does a grid-tied BESS automatically provide backup during an outage?

No. A grid-connected BESS used for peak shaving should not automatically be assumed to operate as an islanded backup system. Backup capability depends on the PCS operating mode, switching architecture, protection and critical-load design.

Is a 125kW/261kWh BESS suitable for both applications?

Potentially, but the site requirement determines the answer.

For a peak-shaving event requiring approximately 120kW and 150kWh, it may be in the right general class for further engineering review.

For a battery-only four-hour backup requirement at 90kW, the theoretical delivered-energy requirement is already 360kWh, which exceeds one cabinet’s 261.248kWh nominal capacity.

What matters most for battery storage ROI?

For peak shaving, important inputs include tariff structure, achievable demand reduction, peak shape, EMS dispatch and recharge strategy.

For backup, critical-load requirements, outage frequency and duration, downtime consequences and alternative backup options become more important.

In a hybrid project, protected backup energy also reduces the battery capacity available for daily economic dispatch.

Conclusion

Peak shaving and backup power may share the same BESS hardware, but the investment decision begins with different information.

For peak shaving battery storage, look at the meter:

How much of the demand peak needs to be removed, and how much energy is actually required above that threshold?

For backup, look at the operation:

Which loads cannot be lost, how much power do they need, and how long must they remain online?

The 125kW/261kWh example shows why that distinction matters. A cabinet in this class can sit in the right general range for a short 120kW peak-shaving event while remaining clearly too small for a four-hour, 90kW battery-only backup requirement.

When one project needs both functions, the more useful question is not simply how many kWh to buy. It is how much power, stored energy and SOC should be allocated to daily economic value versus resilience—and how the system should respond when those priorities compete.

For larger or modular projects, explore HighJoule’s Industrial & Commercial BESS solutions.