Odds are, you have never paid an energy demand charge in your life.

Residential electric bills usually only charge you for energy supply. These charges are based on the total units of energy (kilowatt-hours, kWh) you use. If you use 500 kWh per month, and the supply charge is $0.10 per kWh, your total cost for the month will be $50. (Note that you probably pay a handful of other charges for distribution, societal benefits, etc. These are also all assessed on a per kWh or flat fee per month basis).

Residential customers rarely pay what are called demand charges, but commercial and industrial facilities usually do. These facilities also pay supply charges on a per kWh basis, though their rates are often lower. Demand charges, however, do not work on a per kWh basis. They are charged on a per kW (kilowatt) basis. What’s the difference? As you may recall, kWh is a unit of energy, while kW is a unit of power (energy per second). Demand charges are levied on customers based on the peak power demand (kW) of the building. Thus the name, demand charges.

When a utility levies demand charges in addition to energy charges, it is similar to if your gas station charged you for your top speed as well as the number of gallons you used the past month. Your utility records the power draw of a commercial building continuously, breaking the building’s energy demand into 15 minute intervals. At the end of the month, the utility looks across every interval to see when the facility had the highest energy consumption. Then they charge the building based on that demand. So if, for example, over the course of December, a building’s maximum energy use in a 15 minute interval were 100 kWh, and the demand charge rate were $7 per kW, then the building would be charged $2,800 that month for demand charges. Where do these numbers come from?

• 100 kWh were used in 15 minutes
• This means that over the course of the hour (60 minutes) the building was on pace to use 100 kWh * (60 minutes/hour) / (15 minutes) = 400 kWh/hour = 400 kW.
• 400 kWh used over one hour is a peak demand of 400 kW.
• 400 kW * $7/kW = $2,800.

Demand charges are frequently 25% to 50% of a customer’s utility bill.

Part of the reason demand charges exist is because it is dramatically more expensive for the utility to produce energy at peak times than it is at off peak times. Part of the solution to this challenge is demand response, a payment buildings receive for reducing peak load on days when the grid is at risk of blackouts. This is an important carrot used to reduce peak energy use. The flip side of that solution is the stick of demand charges, which penalize customers for using too much energy at any one time. Peak demand charges, however, are not as precise as demand response: demand response events happen only the 10 times a year the grid really needs it. Most demand charges are applied monthly, so even if the month was uneventful from a grid perspective, a large customer could still pay a hefty demand charge. Some demand charges are “ratcheted,” meaning whatever your highest peak demand was sets your demand charge for the next twelve months.

This is all particularly relevant when a solar array is installed on a roof or nearby grounds. That solar array typically produces peak-loaded power, which is a formal way of saying it generates energy when the grid most needs it – on hot, sunny, summer days in the afternoon when buildings are running air conditioning at full blast and putting stress on the grid. Solar arrays produce energy when it is most needed and when, for most buildings, their demand will be greatest. Then great, solar reduces demand charges.

If only it were that simple. What if, for example, the building had a peak demand of 500 kW of energy at 1 pm and had a solar array that produced 250 kW of power; would the demand charges be cut in half? Almost certainly not. Because that same day the building could need 475 kW of energy at 7 pm, when the solar array produces 0 kW. So your peak demand only fell by 25 kW, from 500 to 475. As you can see from this one example, accurate modeling of solar demand charge savings is a notoriously difficult endeavor. In the next installment, we will explore some strategies to tackle it.