Peak shaving is the deliberate and temporary reduction of a facility's electricity draw from the grid during defined periods of maximum aggregate demand. This is typically achieved by switching to on-site generation, such as diesel generators or battery energy storage systems, or by curtailing non-critical loads. The primary financial driver is the avoidance of demand charges, which are fees levied by utilities based on a customer's highest 15-minute average consumption within a billing cycle.
Glossary
Peak Shaving

What is Peak Shaving?
Peak shaving is a strategic energy management technique used to reduce power consumption during intervals of highest grid demand, thereby lowering electricity costs and mitigating strain on infrastructure.
Unlike load shifting, which moves consumption to a different time without necessarily reducing total energy use, peak shaving actively lowers the net power taken from the grid at a specific moment. This process is critical for grid stability, as it reduces the need for inefficient and carbon-intensive peaker plants. Advanced implementations use AI-driven controllers to forecast load profiles and autonomously dispatch stored energy, ensuring the facility's grid connection never exceeds a pre-set kilowatt threshold.
Primary Peak Shaving Techniques
Peak shaving is executed through a combination of supply-side dispatch and demand-side control. These techniques flatten the load curve to avoid punitive capacity charges and defer infrastructure upgrades.
Battery Energy Storage Dispatch
The most precise method involves discharging Lithium-ion BESS during peak windows. The system monitors real-time building load and injects stored power to cap grid draw at a programmed setpoint.
- Response Time: < 1 second for grid-forming inverters
- Typical Duration: 2-4 hours to cover the utility's peak window
- Key Metric: Depth of Discharge (DoD) management to preserve cycle life
Generator Synchronization
On-site diesel or natural gas generators are automatically started and synchronized to the busbar to assume a portion of the facility load. This is often used in grid-parallel mode to avoid a full transfer switch operation.
- Requires precise synchronization of voltage, frequency, and phase angle
- Often governed by EPA runtime restrictions in non-emergency scenarios
- Common in hospitals and data centers with existing backup infrastructure
Dynamic Voltage Reduction
Also known as Conservation Voltage Reduction (CVR) , this technique lowers the distribution voltage on a feeder to the lower bound of the ANSI C84.1 standard range (e.g., 114V instead of 120V). Many loads behave as constant impedance, so power draw drops proportionally.
- Reduces peak demand without any customer action
- Requires advanced Volt-VAR Optimization controllers
- Typical savings: 1-3% total feeder load reduction
Thermal Energy Storage
Chillers produce ice or chilled water during off-peak hours. During the peak window, the thermal storage tank is discharged to provide cooling, allowing electric chillers to be shut down entirely.
- Decouples cooling demand from electric demand
- High efficiency in facilities with large HVAC loads
- Shifts, rather than sheds, energy consumption
Automated Load Shedding
A Building Management System (BMS) or PLC executes pre-programmed scripts to turn off non-critical loads when a demand threshold is approached. This is a last-resort, fast-acting control.
- Tier 1 Shed: Non-essential lighting, decorative fountains
- Tier 2 Shed: Selected HVAC air handlers, electric water heaters
- Tier 3 Shed: Elevators, non-critical production machinery
- Requires closed-loop feedback to prevent undershoot
EV Fleet Smart Charging
For fleets with predictable routes, smart charging algorithms modulate the charging rate or delay charging sessions to avoid overlapping with the facility's peak demand period.
- Uses ISO 15118 or OCPP protocols for direct charger control
- Maintains state-of-charge targets for departure time
- Aggregated fleets can provide significant demand flexibility
Peak Shaving vs. Load Shifting vs. Load Shedding
A technical comparison of three distinct demand-side management tactics used to balance grid stability and manage electricity costs during periods of system stress.
| Feature | Peak Shaving | Load Shifting | Load Shedding |
|---|---|---|---|
Primary Objective | Reduce maximum demand (kW) to avoid capacity charges | Move energy consumption (kWh) to off-peak periods | Immediately disconnect load to prevent grid collapse |
Trigger Mechanism | Economic signal or site-level controller | Time-based schedule or price arbitrage | Emergency grid stress signal or under-frequency relay |
Total Energy Consumption | Reduced | Unchanged | Reduced |
Duration of Action | Minutes to hours | Hours | Seconds to minutes |
Typical Response Time | < 1 minute | Pre-scheduled | < 2 seconds |
Battery Storage Utilization | |||
Grid Emergency Context | |||
Customer Compensation Model | Demand charge avoidance | Energy arbitrage savings | Reliability payment or penalty avoidance |
Frequently Asked Questions
Clear, technically precise answers to the most common questions about peak shaving strategies, mechanisms, and their role in modern grid optimization.
Peak shaving is the strategic reduction of electrical power consumption during periods of highest grid demand to avoid expensive capacity charges and mitigate the need for activating inefficient peaker plants. It works by either curtailing non-critical loads or dispatching on-site energy resources—such as battery energy storage systems (BESS) or backup generators—to offset the building's grid draw precisely when the facility's consumption would otherwise spike. The control system monitors real-time power draw against a predefined threshold; when consumption approaches that limit, the system seamlessly injects stored power or sheds discretionary loads to 'shave' the top off the demand curve. Unlike load shifting, which moves consumption to a different time, peak shaving typically reduces net import without necessarily increasing consumption later. The financial driver is often demand charges—fees based on a customer's highest 15-minute interval of consumption during a billing period, which can constitute 30-70% of a commercial electricity bill.
Enabling Efficiency, Speed & Accuracy
Intelligent Analysis, Decision & Execution
We build AI systems for teams that need search across company data, workflow automation across tools, or AI features inside products and internal software.
Talk to Us
Search across company data
Give teams answers from docs, tickets, runbooks, and product data with sources and permissions.
Useful when people spend too long searching or get different answers from different systems.

Automate internal workflows
Use AI to route work, draft outputs, trigger actions, and keep approvals and logs in place.
Useful when repetitive work moves across multiple tools and teams.

Add AI to products and internal tools
Build assistants, guided actions, or decision support into the software your team or customers already use.
Useful when AI needs to be part of the product, not a separate tool.
Related Terms
Peak shaving operates within a broader orchestration framework. These related concepts define the mechanisms, markets, and assets that enable strategic load reduction during grid stress events.
Automated Demand Response (ADR)
A fully automated system where a utility signal directly controls customer loads based on pre-programmed permissions, eliminating manual intervention. Unlike manual curtailment, ADR uses OpenADR protocols to dispatch signals that adjust HVAC setpoints, dim lighting, or cycle industrial equipment without human decision-making. This enables sub-second response to grid frequency deviations and forms the technical backbone of modern peak shaving programs.
Customer Baseline Load (CBL)
A statistical calculation of what a customer's energy consumption would have been in the absence of a demand response event. CBL is the critical measurement foundation for peak shaving settlement. Common methodologies include:
- 10-of-10 baseline: Average of the 10 highest usage days in the prior 10 non-event days
- Weather-adjusted baselines: Regression models accounting for temperature sensitivity
- Meter-before-meter-after: Simple pre/post comparison for short events Accurate CBL prevents both underpayment and gaming of incentive structures.
Virtual Power Plant (VPP)
A cloud-based aggregation of decentralized energy resources—batteries, smart thermostats, EV chargers—coordinated to provide grid services equivalent to a traditional power plant. VPPs execute peak shaving by simultaneously discharging hundreds of behind-the-meter batteries during the top 50-100 grid stress hours annually. This orchestration transforms passive consumers into active grid participants, displacing the need for fossil-fuel peaker plants.
Load Shifting
The process of rescheduling energy consumption from peak demand periods to off-peak periods without necessarily reducing total energy usage. While peak shaving focuses on absolute reduction, load shifting maintains total consumption volume. Examples include:
- Pre-cooling buildings during off-peak hours and floating through the peak
- Delaying industrial batch processes to overnight periods
- Charging thermal storage systems when electricity prices are low Load shifting preserves operational output while avoiding capacity charges.
Critical Peak Pricing (CPP)
A dynamic rate overlay that imposes a significantly higher electricity price during a limited number of critical peak event hours to drive extreme load reduction. CPP is a price-based peak shaving mechanism where rates may jump from $0.10/kWh to $1.00+/kWh during declared events. Customers receive day-ahead or day-of notification and respond by curtailing non-essential loads. This tariff structure directly exposes end-users to wholesale scarcity pricing, creating powerful economic incentives for demand flexibility.
Measurement and Verification (M&V)
The rigorous analytical process of quantifying actual load reduction delivered by a demand response resource against its baseline to determine financial settlement. M&V protocols—such as IPMVP and NAESB standards—ensure that peak shaving performance claims are auditable. Key considerations include:
- Adjusting for weather and occupancy variations
- Accounting for non-routine baseline adjustments
- Statistical precision requirements for capacity market qualification Without robust M&V, peak shaving programs cannot participate in wholesale capacity markets.

About the author
Prasad Kumkar
CEO & MD, Inference Systems
Prasad Kumkar is the CEO & MD of Inference Systems and writes about AI systems architecture, LLM infrastructure, model serving, evaluation, and production deployment. Over 5+ years, he has worked across computer vision models, L5 autonomous vehicle systems, and LLM research, with a focus on taking complex AI ideas into real-world engineering systems.
His work and writing cover AI systems, large language models, AI agents, multimodal systems, autonomous systems, inference optimization, RAG, evaluation, and production AI engineering.
Partnered with leading AI, data, and software stack.
How We Work
Custom AI workflows for your Business
One-fit-all AI don't work for modern businesses. At Inferensys, we aim to understand your business & custom requirements; which we use to define most efficient agentic workflows, the data, and the tools for your business.
01
Review the use case
We understand the task, the users, and where AI can actually help.
Read more02
Pick the right approach
We define what needs search, automation, or product integration.
Read more03
Build the first useful version
We implement the part that proves the value first.
Read more04
Improve from there
We add the checks and visibility needed to keep it useful.
Read moreThe first call is a practical review of your use case and the right next step.
Talk to Us