This is the third in a series of posts on grid energy storage. Earlier, in Energy Storage 101 and 102, we attempted to elucidate the applications of energy storage for various customer types and discussed the technologies and value chain. In this post, we examine the economics of behind-the-meter (BTM) storage for commercial customers. We model the returns to a hypothetical commercial BTM storage customer in New York City, San Diego, Honolulu, and Des Moines. We show how it is possible to optimize returns by sizing the battery intelligently with a software model.

For economy, we use the term “commercial” here to represent any non-residential end user of electricity, including commercial enterprises, schools, non-profits, government entities, etc. Also, we assume the host is the owner of the storage system.

Applications of Commercial Energy Storage

As we mentioned in Energy Storage 101, commercial customers have several potential sources of value for energy storage (Figure 1):

Figure 1: Sources of Value from commercial Btm energy storage

Storage can be used for peak shaving and/or backup power. Many buildings already have segregated wiring for systems they want running in case of an outage (e.g. elevators or emergency lighting), but in most cases a battery designed for peak shaving will not have enough capacity to run an entire building. Here, we focus here only on the economics of peak shaving only because modelling the value of back-up power relies heavily on the importance of continuous power to a specific building.

In some markets, storage can also generate grid services revenue. For example, a storage system owner could sell frequency response services to an ISO or “generation” capacity to a utility.

Finally, storage could create value in conjunction with solar if storage would allow a larger solar system than would otherwise be allowed or economic and solar is otherwise less expensive than the grid. A battery system installed together with solar would also be eligible for an investment tax credit (30% through 2016, 10% thereafter). For simplicity, we don’t consider solar + batteries in this analysis. We will come back to this in a future post.

Where Does Commercial Energy Storage have a Good Value Proposition?

The presence of one (or, generally, more) of four factors determine whether a location is attractive for commercial BTM storage:

High demand charges, which make peak shaving valuable
A rate structure that limits the cost of energy when using a battery
The ability to generate revenue from offering grid services
Incentives that reduce the net cost of the battery system

Figure 2 shows 15 U.S. utilities with high demand charges.

Figure 2: Annual Average Demand charges for select u.s. utilities,
customer with 400kW demand
($/Kw)

Modeling Assumptions

To determine the economics of energy storage in our four markets we have to make some assumptions about how energy is used in the host building and the characteristics of the battery system.

Building Energy Use Assumptions

As a baseline we assume a building a with a daily and monthly peak demand of 400 kW (every day is the same). We further assume the building has one large peak during the most expensive time of day. The peak lasts for five hours. At other times the demand is 300 kW (we will also examine some alternative demand / usage assumptions later in the post).

Battery Assumptions

We assume a 350 kWh lithium ion battery system designed to shave the building’s peak demand by 50 kW, which it will do for each of the five hours of its gross peak demand. It will charge during off-peak hours.

The total plant cost of a lithium ion battery system used for commercial BTM applications is approximately $800 / kWh. Therefore, the battery system will cost about $280,000.

Tesla warrants their battery systems for 10 years. For this analysis, we assume that the battery itself needs to be replaced after 13 years, but that the system will last 26 years. We assume a battery replacement battery cost in year 13 of $108,000.

Figure 3: Building and battery system assumptions

We also assume we spend a small amount each year on maintenance.

Economics of Energy Storage in New York

New York is potentially an ideal place for storage. All four of the factors we identified before are present. First, ConEd has among the highest average demand charges in the country, at approximately $19/kW. Second, ConEd’s tariffs allow customers with batteries to pay less for energy when using a storage system. Third, customers are eligible for an incentive of $2,100 per kW. Finally, battery system owners can participate in the NYISO’s frequency regulation market and earn grid services revenue (although our battery happens to be too small to qualify).

Peak Shaving in New York

A building with a profile like the one we have assumed would be billed under ConEd’s SC9 tariff.

Without energy storage, we assume our building is on SC9’s Rate I: “General – Large”. This rate structure is a familiar one. Customers pay for monthly peak demand, plus energy supply and delivery. When energy storage is installed on the building, we change to SC9’s Rate III: “Voluntary Time-of-Day”. Under this rate structure the price of energy delivery decreases significantly and the demand is charged based on peak demand during specific times of day. Figure 4 shows a comparison of the two rate structures.

Figure 4: Consolidated Edison SC9 Rates I and III*

Based on these tariffs and our assumptions about the building and battery system, the owner of our battery system will save approximately $31,000 in the first year of operation. Savings will increase over time as grid rates increase (which we assume they will do at 3% annually).

Storage Incentives in New York

ConEd pays an incentive of $2,100 per kW (up to 50% of system cost) for installing energy storage in Zone J (New York City) via the Demand Management Program.

Our system, which initially costs $280,000, would earn $140,000 in incentives.

We are also able to depreciate our battery system under a 7-year MACRS schedule, using $140,000 as the basis. This will reduce our federal and state income taxes.

Grid Services Revenue

A battery system in New York City may be able to generate grid services revenue in one of two ways:

Selling frequency response services to NYISO via the ancillary services market. However, this requires a minimum battery size of 1MW. Assets bidding into this market cannot be aggregated.
Selling demand response to ConEd. The eligible system size is 50 kW or more. However, this would only work for customers whose peak demand is unlikely to be coincident with a demand response event. In our example, we assume peak building load is coincident with peak system load, so we don’t participate in or generate revenue from demand response.

Results: New York

Given all the assumptions above, we find an IRR of 10% and a simple payback of eight years.

What-If Analysis

Next, we look at how the returns vary if we change some of the assumptions above.

Effect of Different load profiles on Energy Storage Returns

For our initial analysis we assumed that the building had one peak that lasted five hours and that it was coincident with the grid peak. However, what happens if the building has a different load profile?

Figure 5: Variations in returns – Four Illustrative load profiles

Case 2, which is the same as the base case except that the peak occurs overnight, shows a negative IRR. With a profile like this, a rational customer would be on a time-of-use plan from the outset, and the additional savings from the battery system would be too small to recoup its cost.

Case 3, in which demand during peak hours varies more than in the base case, results in very slightly higher returns than the base case because the depth of discharge is lower and therefore the battery lasts longer before needing replacement.

Case 4, which is the same as Case 3 except the demand peaks are twice as high for half as long, has an IRR lower than either the Base Case or Case 2. Under Rate III, demand is extremely expensive.

In all of the above cases we assumed the same battery. In cases where the customer has a demand peak that is coincident with the grid peak, a battery can make sense. However, it would be more economic to size the batteries to optimize the returns for each case.

Optimal Battery Sizing

It makes intuitive sense that increasing the maximum load and/or storage capacity of the battery would allow more grid savings. To calculate the optimal battery system configuration, however, is not this simple. We need to consider the specifics of the rate plan, the actual usage profile, when it is appropriate to charge and discharge, and how the cost of the battery changes with maximum load and storage capacity.

Woodlawn has a software model to do this. Using our four cases, we found we could significantly increase the returns of our “peaky” cases by using different battery configurations (see Figure 6).

Figure 6: Energy Storage Returns for Optimized batteries

Of course, in the real world we may not have an infinite number of size choices when it comes to battery systems, but this type of analysis can definitely help us choose the best of several options.

Effect of different rate Plans

Some customers with batteries in New York have not moved to Rate III, but using a more aggressive approach have taken service on Rate IV, which is for “Standby Service.” This rate was originally designed for customers with cogeneration or internal combustion distributed generation who are essentially off-grid except when their on-site generation cannot operate. For these customers ConEd acts as a “standby” service—thus the name. Customers whose rated energy storage load represents more than 15% of their peak demand may be eligible to take service on Rate IV.

Under Rate IV, customers are required to contract for their expected monthly peak demand. They also pay for as-used demand daily, Monday through Friday. Finally, they pay the same rate for energy supply as customers on Rate I. However, they are not charged for energy delivery—which turns out to be a significant benefit. See Figure 8.

It is possible that the tariff rules could change and customers are required to stay on Rate I, rather than choosing to move to Rate III. In this case, the investment is not recouped by the savings.

Figure 8: Energy Storage Returns—Three Rate Plans

New York: SUMMARY

New York provides a very interesting landscape for installing a battery. By installing our battery we were able to reduce demand charges and pay a lower rate for energy. We also benefited from the state incentive. Our base case IRR was 10%, though we found cases that were significantly higher and lower. Generally, however, a battery is economic in New York if the building peak or peaks occur during grid peak times. The optimal power and capacity of the battery system depends on the specific load profile.

Economics Energy Storage in California, Hawaii, and Iowa

After a detailed analysis of New York City, we looked at three other cities that have some—but not all—of the characteristics that make battery storage attractive: San Diego (utility: SDG&E), Honolulu (HECO), and Des Moines (MidAmerican).

Our assumptions remained the same for the building and the battery. Of course, we adjusted or model for the tariffs in place in these other cities.

Figure 9: Energy Storage Returns: New York, California, Hawaii, and Iowa

Economics of Energy Storage in California

SDG&E has very high demand charges, and while there are incentives for storage in California, they are lower than in New York, at “only” $1,460/kW. It turns out that in our base case, the higher demand charges and lower incentive essentially cancel one another, and the result is an IRR similar to the base case in New York. However, when we have a peaky profile, and optimize the battery, we see that the high demand charges give SDG&E the best returns.

Another factor assisting the returns in San Diego is that SDG&E also allows businesses to participate in capacity bidding. Businesses can bid a minimum of 20 kW of curtailment capacity each day from May to October, increasing the income a battery can provide. The returns we find in California benefit from several thousand dollars per month in capacity revenue.

Economics of Energy Storage in Hawaii

Honolulu showed the worst return of all of the cities we examined. HECO’s demand charges are high—but not has high as SDG&E’s—and there are no incentives, nor are there ancillary services markets today.

A rate structure that is more beneficial for BTM storage may provide market incentives for commercial customers to install storage that would help HECO manage increased penetration of rooftop solar. Today, though, installing a battery that can be charged via rooftop solar may make storage economic for commercial customers. Solar is usually less expensive than grid energy in Hawaii, and installing a storage system together with solar would enable the owner to take the 30% ITC on the storage system.

Economics of Energy Storage in Iowa

MidAmerican has a rate structure completely based on demand. That means that every kilowatt decrease in demand affects the entire bill, not just the demand charges, explaining the significant increase in IRR associated with an optimized battery for a profile with multiple shorter peaks. However, because the rates are relatively low, the savings is not large enough to provide as high of a return as in some other states.

In Des Moines, MISO, like NYISO and CAISO, has a frequency response market. However, our system cannot participate because the minimum size requirement for energy storage resources is 1 MW. MidAmerican, like HECO, does not purchase capacity.

Conclusions

Four major factors influence where energy storage can be economic: demand charges, energy rates, grid services opportunities, and incentives. However, a customer’s load profile is also a significant factor. Finally, it is possible to optimize returns by smartly designing the battery system, perhaps using optimization software.

Woodlawn would be happy to assist you with any questions about energy storage. Please contact Micah Sussman or Josh Lutton.

Energy Storage 201: Commercial BTM Energy Storage