Making Customer Battery Storage Work for a Cleaner, Less Expensive Grid

How Poor Commercial Rate Design Can Make The Grid Less Flexible

Battery energy storage is often hailed as a “holy grail” to unlock a renewable energy future. And battery costs are falling, bringing that vision closer to reality. In fact, it is now cost-effective for many commercial customers to invest in batteries to reduce their energy bills. But therein lies the rub—many of these commercial customers face grid electricity-rate designs that encourage inefficient use of these batteries. In fact, this growing market opportunity for storage highlights the need to revamp the design of typical demand charges that encourage flat customer loads and fail to account for the needs of increasingly renewable energy systems.

Falling Battery Costs Are Creating Tremendous New Opportunities for Commercial Customers to Reduce Demand Charges

As battery costs continue to decline, millions of commercial customers in the U.S. now have the opportunity to reduce their energy bills using cost-effective battery storage. Recent reports have highlighted the specific opportunity to use batteries for demand-charge management, reducing a customer’s peak demand from the grid each month.

  • NREL and Clean Energy Group estimate that over five million U.S. commercial customers have access to demand charges high enough (above $15/kW) to make battery storage economic for demand-charge management
  • GTM Research claims that by 2021, $11/kW demand charges will make this economic, and 19 states will have economic cases for demand-charge management
  • McKinsey estimates that demand charges as low as $9/kW may make battery storage economic, which could dramatically expand this market potential
  • LBNL demonstrates how solar and storage together offer particularly high value in reducing commercial demand charges
Most Demand Charges Encourage Flat, Consistent Load Profiles

Most demand charges are not coincident with system peak. That is, the customer’s demand charges are calculated based on the maximum amount of power that customer draws during the month, regardless of whether the customer’s peak demand is at a high-cost or low-cost time for the broader electric grid. Optimizing for these noncoincident demand charges with batteries, electric vehicle charging, or other flexible demand requires flattening a customer’s load as much as possible, so that consumption from the grid doesn’t vary much over time.

Consider a typical office building billed under a noncoincident demand charge. Its peak demand is typically between 11 a.m. and 5 p.m., while usage is very low overnight. To reduce the customer’s bill, a battery storage system could charge from the grid overnight and discharge at midday, reducing the office building’s peak demand. This is consistent with utilities’ former need for steady base load they could most easily serve with thermal power plants, but not with the needs of systems with high levels of renewable energy. An office building like this in a region with high penetration of solar power could actually make it more challenging to integrate solar energy by reducing midday demand.

Figure 1: Managing noncoincident demand charges for a typical large office building would mean charging a battery in the early morning to reduce grid consumption during midday.
Just as Today’s Grids No Longer Need Baseload Power, They No Longer Need Steady Base Load

Today’s grid needs are changing from those in the past. The rapid growth of renewable generation—from widespread solar in California, Hawaii, and the Southwest to wind power across the Midwest and Texas—is actually altering system peak demands. In some cases, the increased amount of solar power on the grid during the day means that it is advantageous to encourage customers to demand more power during the middle of the day, when low-cost solar is available, whether for well-timed EV charging, battery storage, or preheating electric water heaters. Overall, integrating solar and wind generation requires flexible use of the other resources on the grid, both generation and demand. And just as today’s grids no longer need baseload power, they no longer need flat, consistent load profiles. In fact, just as steady, inflexible generation makes grid management harder and more costly in a renewable world, so does steady, inflexible demand.

In our example, the noncoincident demand charge exacerbates the challenge of integrating solar generation. The building tenant would want to reduce his or her bill by reducing midday electricity usage. In a region with high solar generation, that’s sending the opposite of the signal that the grid actually needs.

Figure 2: Demand charge management can work contrary to grid needs for load shifting. System-wide load and net load shown for CAISO, September 19, 2017.

The individual customer’s load shape will determine whether demand-charge management is beneficial or detrimental to grid operations—a hotel may aim to reduce evening load aligned with system needs, but a primary school would be reducing daytime load. This is a central problem with demand charges: that strategies to manage them are based on customer load patterns, not on the needs of the overall system.

Time-Varying Commercial Rates can Better Support System Needs

Instead of noncoincident demand charges, time-varying rates can better allocate costs to different times of day and encourage demand shifting and storage to lower system-wide costs. Several options include peak-coincident demand charges, time-of-use rates, and real-time pricing.

Peak-coincident demand charges bill based on peak demand during those times when marginal system costs are highest, to encourage reducing electricity demand when it is most valuable to do so. In the example above, this could mean a demand charge limited to peak customer load during early evening hours, when system costs are highest.

A further step would be to eliminate demand charges altogether and implement commercial time-of-use rates, in which a customer is charged based on consumption according to a predetermined schedule aligned with system costs. Real-time pricing takes the concept even further, allowing pricing to vary hourly based on actual system conditions. Time-of-use and real-time pricing have the advantage of providing a price incentive to reduce every kWh of high-cost energy, whereas even peak-coincident demand charges only require reducing a monthly maximum.

Compensation for Additional Grid Services Helps Customers Optimize Energy Use Even More

Offering compensation for additional grid services helps customers optimize batteries and demand flexibility even more, with new economic incentives to adopt these resources. RMI’s Economics of Battery Storage showed that energy storage can generate much more value when multiple, stacked services are provided by the same device. Commercial customers investing in batteries to reduce their bills will only realize this value if their utility rates allow it. For instance, a rate could include:

  • Time-of-use pricing for energy, with higher prices during the most costly times of day and lower prices when energy is inexpensive
  • Critical-peak pricing, which significantly raises prices for a few events per year, in return for lower prices year-round
  • Compensation for grid support services like fast frequency response, regulation, and contingency reserves, all of which are within the capabilities of today’s batteries. Hawaiian Electric Companies’ proposed demand-response portfolio offers a good example of this concept

It’s becoming increasingly clear in high-renewable electric grids that flexible resources are more valuable than flat, unchanging generation and load. Noncoincident demand charges, while widely used for commercial customers, do not provide incentives for demand flexibility that consistently align with system needs. As the price of batteries drops, storage systems may be deployed to save customers money while actually exacerbating operational challenges on the grid. Time-varying rates can more accurately convey system costs throughout the day and aid in efficient grid management to the benefit of all customers.


Image courtesy of iStock.