Peak Demand Shaving vs Energy Arbitrage

Peak demand shaving and energy arbitrage both cycle the battery to reduce electricity costs, but they monetize different parts of the tariff. 

What dispatch strategy to select?

Peak demand shaving:

• Demand charges are a large portion of the bill. Commercial and industrial customers on tariffs with demand charges of roughly $10-15+/kW-month (and higher in some utility territories) can see demand charges make up 30-70% of the bill. Cutting a monthly peak by even 100 kW can be worth well over $10k/year.
• The load profile has short, spiky peaks. Demand charges are usually set by the highest 15-minute (sometimes 30- or 60-minute) interval in the month. If a facility has a few brief spikes (HVAC startup, EV chargers, motor starts, batch processes), a relatively small battery can clip them. Peaks that are broad and flat are expensive to shave because they take a lot of energy (kWh), not just power (kW).
• The tariff has ratchets or coincident-peak charges. One bad interval can set the charge for 11 months (ratchet) or be tied to system-wide peaks, which raises the value of reliably catching peaks.
• Energy rates are flat or have a small spread. If TOU spreads are narrow, arbitrage has little to work with, so demand charges are where the value is.
• The battery is power-heavy (shorter discharge duration). Shaving rewards kW charges over kWh, so 1-2 hour systems often pencil well.

Energy arbitrage:

• TOU or real-time price spreads are wide. A spread of roughly $0.15-0.20+/kWh or more between off-peak and on-peak, after accounting for round-trip efficiency losses (about 85-90%) and degradation, is generally what it takes to justify cycling daily.
• There are no demand charges, or they’re lower. This is typical of residential and some small commercial tariffs.
• The customer is exposed to wholesale markets (front-of-meter, merchant storage, or retail real-time pricing), where volatility creates the opportunity.
• The load is predictable, so peaks are not the issue but the price of energy by hour is.
• Longer duration is available, since arbitrage monetizes kWh shifted per day (2-4 hour systems are common).

Key tradeoffs