GSA_Storage_Analysis_NREL_memo.pdf

The National Renewable Energy Laboratory (NREL) evaluated two GSA buildings for opportunities for cost-effective solar PV and storage. The buildings were narrowed down by

GSA based on the following criteria: 1) demand over 2,000 kW, 2) savings potential over

2,227,728 kWh and 3) lowest annual consumption with high peak demand. For each building, GSA provided 15-minute interval data and information about the rate tariff.

The economic viability of PV and storage was evaluated using a tool called REopt by estimating the utility cost savings the technologies could provide and comparing those savings to the estimated installation cost and annual recurring ownership costs of each. Projects were evaluated on net present value (NPV) which is the present value of the savings (or costs if negative) realized by the project over the 20 year lifecycle. While batteries can derive value from multiple revenue streams including ancillary services, demand response programs and capacity markets, only demand management (reducing peak demand charges on utility bill) and energy arbitrage

(buying energy in off-peak hours, consuming during peak hours) were considered in this analysis because they are the most commonly available revenue streams.

Building #1 is located in Hawaii and is served by Hawaiian Electric Company Schedule P, Large

Power Service. The energy charge is $0.14/kWh with no seasonal or time of use adjustments.

The demand charge is $24.34/kW and is determined each month by the highest of: 1) the maximum demand of that month and 2) the mean of the maximum demand for the current month and the greatest maximum demand for the preceding eleven months. The total annual electricity use is 11,495,937 kWh with a maximum15-minute peak of 2,021 kW and an average 15-minute peak of 1,312 kW.

NREL compared the business as usual case with four alternative scenarios shown in Table 1. In scenario 1, PV alone was evaluated assuming 86,000 ft of roof space would be available; in this scenario a 330 kW system (the maximum size that fits in the available space) would lower the 20 year cost of energy by $605,000 by providing about $34,000 in demand savings and about

$70,000 in energy savings each year. The combination of PV and storage was evaluated in scenarios 2 and 3 using two different assumptions for space available: 86,000 ft of roof space, and unlimited space available. In both scenarios 2 and 3, PV and storage appear cost effective.

When 86,000 ft of roof space is available, the 330 kW PV system is still recommended along with a 92 kW-253 kWh battery. These two technologies would provide about $56,000 in demand savings and about $70,000 in energy savings each year for a total savings of $697,000 over the

REopt is an energy-planning platform that can be used to evaluate renewable energy (RE) options, estimate costs, and suggest a mix of RE technologies, storage and conventional generation to meet defined assumptions and constraints. The best mix of RE technologies depends on several factors, including utility rates; RE resources;

technology cost and performance; state, utility, and federal incentives; and economic parameters (discount rate, inflation rates). Each of these factors was considered in this analysis. For more information about REopt, visit http://www.nrel.gov/tech_deployment/tools_reopt.html http://www.nrel.gov/tech_deployment/tools_reopt.html

20 year analysis period. When space is not limited, a 2,450 kW PV system is recommended along with a 289 kW-802 kWh battery. This combination of technologies would provide about

$153,000 in demand savings and $509,000 in energy savings each year and reduce energy costs by over $2 million over the 20 year analysis period. Finally storage alone was evaluated. In this scenario a much smaller storage system (29 kW-63 kWh) could lower annual demand costs by about $8,000 and provide $27,000 in savings over the 20 year analysis period.

Table 1. Results for building 1, located in Hawaii

0. Base

Case

1. PV Only

2. PV +

Storage

3. PV +

Storage no land limit

4. Storage Only

PV size (kW) 0 330 330 2,450 0

PV cost ($) $0 $993,000 $993,000 $7,362,000 $0

PV energy production (kWh)

0 485,433 485,433 3,600,272 0

PV energy export (kWh) - 21 0 64,838 0

Battery size (kWh) 0 0 253 802 63

Battery size (kW) 0 0 92 289 29

Battery cost ($) $0 $0 $218,500 $690,000 $60,500

Demand costs ($) $545,000 $511,000 $489,000 $392,000 $537,000

Energy costs ($) $1,666,000 $1,596,000 $1,596,000 $1,157,000 $1,666,000

Demand savings ($) $0 $34,000 $56,000 $153,000 $8,000

Energy savings ($) $0 $70,000 $70,000 $509,000 $0

Percent RE (%) 0% 4% 4% 31% 0%

Net present value ($) $0 $605,000 $697,000 $2,063,000 $27,000

Building #2 is located in Washington DC area and is served by PEPCO (distribution and riders) and WLG Energy. The combined energy charge including riders is $0.071/kWh with no seasonal or time of use adjustments. The demand charge is $6.16/kW and is determined each month by the highest 30-minute demand for the current month. The total annual electricity use is 4,547,474 kWh with a maximum15-minute peak of 1,226 kW and an average 15-minute peak of 519 kW.

NREL compared the business as usual case with three alternative scenarios shown in Table 2. All three scenarios assume enough land or roof space to host a 1 MW PV system would be available.

In scenario 1, PV and storage were evaluated without the value of the solar renewable energy certificate (SREC) included. In this scenario, neither PV nor storage appear to be cost effective.

When the value of the SREC is included ($175/MWh for 10 years) in scenario 2, a 1 MW PV system appears cost effective and would provide $1.6 million in savings over the 20 year analysis period by lowering demand costs by $6,000 and energy costs by $104,000 each year.

Finally, in scenario 3, storage costs were lowered until storage was economically viable. This happened when storage costs were lowered by 40% from $1,000/kW-$500/kWh to $600/kW-

$300/kWh. At these lower costs a small storage system can provide an additional $2,000 in annual demand savings.

Table 2. Results for building #2, located in Washington DC

Scenario

0. Base Case 1. PV + Storage without SREC

2. PV + Storage with SREC

3. PV + Storage with SREC & lower storage costs

PV size (kW) 0 0 1,000 1,000

PV cost ($) $0 $0 $2,540,000 $2,540,000

Battery Size (kW) 0 0 0 28

Battery Size (kWh) 0 0 0 35

PV energy production (kWh)

0 0 1,261,157 1,261,157

Demand costs ($) $55,000 $55,000 $49,000 $47,000

Energy costs ($) $399,000 $399,000 $295,000 $295,000

Demand savings ($) $0 $0 $6,000 $8,000

Energy savings ($) $0 $0 $104,000 $104,000

Percent RE (%) 0% 0% 28% 28%

Net present value ($) $0 $0 $1,577,000 $1,583,000

Table 3 provides additional analysis assumptions used for this screening.

Table 3. Additional analysis assumptions

Technologies Solar PV and storage

Scenario Minimize lifecycle cost (cost-effective projects)

Ownership model Third party (PPA)

Analysis period 20 years

Discount Rate 3.1% Government/10% developer

Utility Escalation rate 0.1% for HI per NIST

0.51% for DC

Inflation rate 0.1% general inflation rate per NIST

Incentives PV: 30% ITC; 5 year MACRS Battery: 30% ITC; 5 year MACRS (no charging from grid)

Net metering limit None

Electricity over net metering limit $0/kWh

Interconnection limit None

Technology costs

PV: $3.00/W

Battery: $500/kWh; $1,000/kW ($200/kWh; $200/kWh for replacement in year 10)

Lavappa, Priya D., and Joshua D. Kneifel. 2015. Energy Price Indices and Discount Factors for Life-Cycle Cost

Analysis – 2015: Annual Supplement to NIST Handbook 135. National Institute of Standards and Technology.

NISTIR 85-3273-30. http://dx.doi.org/10.6028/NIST.IR.85-3273-30.

IRS rules state battery may capture the percentage of the ITC equal to the percentage the battery is charged from

RE, with a minimum 75% RE requirement. This analysis requires the battery to be charged exclusively by PV and therefore captures the full value of ITC and MACRS.

http://dx.doi.org/10.6028/NIST.IR.85-3273-30

Technology resource NREL GIS data

Analysis time step 15-minute