Applied Energy 164 (2016) 133–139
Contents lists available at ScienceDirect
Applied Energy
journal homepage: www.elsevier.com/locate/apenergy
An economic comparison of battery energy storage to conventional
energy efficiency technologies in Colorado manufacturing facilities
Kalen Nataf 1, Thomas H. Bradley ⇑
Colorado State University, Department of Mechanical Engineering, Campus Delivery 1374, Fort Collins, CO 80523-1374, United States
h i g h l i g h t s
Energy storage’s and efficiency technologies’ economic payback is compared.
Conventional efficiency technologies have shorter payback for the customers studied.
Hypothetical incentives can lower the payback periods of battery energy storage.
a r t i c l e
i n f o
Article history:
Received 24 August 2015
Received in revised form 11 November 2015
Accepted 26 November 2015
Available online 17 December 2015
Keywords:
Battery storage
Demand reduction
Energy efficiency
Industrial
a b s t r a c t
Battery energy storage (BES) is one of a set of technologies that can be considered to reduce electrical
loads, and to realize economic value for industrial customers. To directly compare the energy savings
and economic effectiveness of BES to more conventional energy efficiency technologies, this study collected detailed information regarding the electrical loads associated with four Colorado manufacturing
facilities. These datasets were used to generate a set of three scenarios for each manufacturer: implementation of a BES system, implementation of a set of conventional energy efficiency recommendations, and
the implementation of both BES and conventional energy efficiency technologies. Evaluating these scenarios’ economic payback period allows for a direct comparison between the cost-effectiveness of energy
efficiency technologies and that of BES, demonstrates the costs and benefits of implementing both BES
and energy efficiency technologies, and characterizes the effectiveness of potential incentives in improving economic payback. For all of the manufacturing facilities modeled, results demonstrate that BES is the
least cost-effective among the energy efficiency technologies considered, but that simultaneous implementation of both BES and energy efficiency technologies has a negligible effect on the BES payback period. Incentives are demonstrated to be required for BES to achieve near-term payback period parity with
more conventional energy efficiency technologies.
Ó 2015 Elsevier Ltd. All rights reserved.
1. Introduction
The rate structures that are commonly in place for small to
medium-sized manufacturers can provide incentives for energy
efficiency, peak demand reductions, and load shifting. For Colorado
manufacturing industrial facilities, the costs of electrical energy
service is broken up into three or four scalable costs: a connection
cost (with a price measured in $ per connection), an energy cost
(with a price measured in $ per kilowatt-hour), a peak demand cost
(with a price measured in $ per kilowatt), and a coincidental
⇑ Corresponding author. Tel.: +1 (970) 491 3539.
E-mail addresses: [email protected] (K. Nataf), Thomas.Bradley@
colostate.edu (T.H. Bradley).
1
Tel.: +1 (720) 212 9743.
http://dx.doi.org/10.1016/j.apenergy.2015.11.102
0306-2619/Ó 2015 Elsevier Ltd. All rights reserved.
demand cost (with price also measured in $ per kilowatt). Utility
companies calculate peak demand in kilowatts (kW) by measuring
fifteen minute intervals of the user’s energy consumption, in units
of kilowatt hours (kW h), and dividing it by the amount of hours in
that interval, 0.25 h [1]. They then sort these demand calculations
for an entire month, and the maximum demand calculated is the
peak demand. This means a single fifteen minute period in a month
dictates the cost a customer pays for peak demand for the entire
month. Coincident demand is the customer’s electric demand during the one hour each month, called the peak hour, where the
wholesale electricity generation and transmission provider supplied the highest load [2]. The high price of these demand costs
generally incents consumers to minimize both their peak loads
and their loads during the typical peak hour (e.g. early evening in
the winter and mid-afternoon in the summer) [27].
134
K. Nataf, T.H. Bradley / Applied Energy 164 (2016) 133–139
The most common methods of reducing peak demand and loads
during the peak hour are implementing new or retrofit energy efficient technologies and load shifting techniques. As advancements
in technology are made, new products become available that can
either replace or be added to an existing piece of equipment to
improve its energy efficiency (e.g. variable frequency drives,
high-efficiency motors, light emitting diode (LED) lighting). These
energy efficient technologies effectively reduce the total peak
and peak hour loads by reducing the load of that equipment during
its hours of operation [3,4]. Manufacturers implement these energy
efficiency technologies preferentially because of their costeffectiveness and the presence of utility incentives [5]. Load shifting is another preferred technique for reducing demand charges
[6,7]. Load shifting lowers the peak and peak hour loads by moving
the time or intensity of operation of individual equipment to lower
cost times of the day. This can be done a variety of ways including
technologically (e.g. thermal storage [8], demand controllers [7])
and manually through equipment operational scheduling and
employee training. Although this method has the potential to be
more successful at lowering peak and peak hour loads than energy
efficient technologies, it is dependent on the individual equipment
loads that make up the peak and peak hour total load [8]. Furthermore, it is much harder to implement this method in manufacturing facilities, as shifting equipment loads can interfere with the
manufacturing process and cut into the productivity, and thereby
the profits, of that facility [7–11].
Behind-the-meter (BTM) battery energy storage (BES) systems
aim to reduce electricity costs by providing a way to redistribute
the peak and peak hour loads without the productivity losses
that might be associated with standard load shifting techniques.
Although there has been research on BES in several applications
such as electricity grid frequency regulation [12,13] and renewable generation storage [13–16], the business case for near-term
BES is not well-defined. BES technology is presently being
pushed to market through mandates and incentives. As an example, in California, the California Public Utilities Commission
(CPUC) has set a procurement target of 1.325 GW of energy storage (including BTM) by 2020, for installation no later than 2024
[20]. Several studies have shown BTM BES to have moderate
potential for lowering the peak demand of a facility and thereby
the associated costs in a behind-the-meter (or customer-sited)
application [12–19]. However, these studies have not sought to
consider BTM BES as one of a set of energy cost saving technologies that could be implemented separately or together, and have
not been able to consider BES’s role among the set of energy and
cost saving technologies. If energy efficient technologies and load
shifting techniques were to be applied at the same facility as a
BTM BES system, the value of BES in reducing energy costs
may be misestimated by the models currently proposed in
literature.
With the pathway toward commercialization of BTM BES
already in progress, the energy cost savings associated with BES
should be assessed relative to other energy efficiency technologies and more common load shifting techniques. To perform this
evaluation, this study measures a broad set of loads in four case
study industrial facilities, and proposes the implementation of
both BES and more conventional energy efficient technologies
for each facility. The electricity cost savings and payback period
associated with each of these technologies can thereby be directly
compared to that of BES. This analysis allows us to understand
the implications of installing conventional energy cost reduction
technologies prior to BES, directly compares the costeffectiveness of these technologies to BES, and provides insight
as to what the utility incentives in Colorado would need to be
to bring BES into payback parity with other energy cost saving
technologies.
2. Materials and methods
2.1. Summary
This research is based on four case studies of different types of
manufacturing facilities located in Colorado, USA. In each of these
facilities, the electrical loads associated with large operational
equipment were data-logged and the total electrical load of the
facility was collected. Other observational information that could
shed insight on how energy was being used was also gathered.
Using the information gathered, the estimated costs and savings
of applicable new and retrofit energy efficient technologies were
calculated using Department of Energy (DOE) and CSU Industrial
Assessment Center (IAC) toolsets. To assess the cost-effectiveness
of BES for each case study, the National Renewable Energy Lab’s
(NREL) Battery Lifetime Analysis and Simulation Tool for Behindthe-Meter applications (BLAST-BTM) was used to estimate the
costs and savings at each facility both before and after the estimated savings were applied to the collected load profile. The
cost-effectiveness for all of these technologies are compared using
the metric of payback period, which has shown to be a primary
consideration in an industrial facility’s decision to implement
energy efficiency measures [21–25].
2.2. Data gathering and acquisition
A walk-through of the facility was performed at each manufacturing plant to gain an understanding of the manufacturing process, and to locate large energy uses in the facility. While on the
walk-through, specific information was gathered that would later
assist in estimating the savings of energy efficient technology
including hours of operation, lighting counts, nameplate information on outdated or energy inefficient equipment, air leak identification, and more. One to two pieces of equipment that represented
a significant contribution to the facilities’ total load were then
data-logged using HOBOware data loggers (Onset Computer Corporation, Bourne, MA) and set to record for two week periods. This
data collection period was chosen due to time constraints determined by the plants and the storage capacity of the data loggers.
Data were collected on intervals varying from ten seconds to one
minute, and were averaged in MATLAB to fifteen minute intervals
as this is the collection period considered by the utility for peak
demand and coincident demand cost calculations. The total load
of each facility over the same two week period was also collected
either using a power quality analyzer in three minute intervals
and averaged to fifteen minutes or in fifteen minute intervals in
datasets provided by the utility. The benefit of logging the individual loads of large equipment is that it adds depth to our understanding of the electrical loads by showing how each load
interacts and contributes to the total load of the facility, especially
during peak and peak hour loads (Fig. 1). Additionally, when calculating the energy savings that might be associated with retrofit
technology, we can apply that savings only to periods when that
machine is on instead of applying it over the total load profile during the estimated hours of operation. The information obtained in
the walk-through and data collection was then used to identify
applicable new and retrofit energy efficient technologies and to
estimate the associated costs and savings.
2.3. Modeling of energy efficiency technologies
IAC toolsets developed over the 31 years of Colorado State
University’s IAC program, validated and approved by the DOE IAC
program management, were used to estimate the associated
energy savings of implementing energy efficient new and retrofit
K. Nataf, T.H. Bradley / Applied Energy 164 (2016) 133–139
700
Facility Total Load
Air Compressor Load
Refrigeration Load
600
Load [kW]
500
400
300
200
100
0
0
2
4
6
8
10
12
14
16
18
20
Time [Days]
Fig. 1. Subsample of Plant 4 load profiles generated by measuring electrical loads.
technologies. The inputs for these calculations are collected during
the walk-through process and are specific to the equipment
installed in each plant. The demand and usage saving costs were
calculated using prorated rates (shown in Table 1) based on the
previous year’s rate adjustments, seasonal variation, and taxes
from the site-specific utility. The capital costs associated with
these implementations are determined from the purchase prices
at online retailers such as Grainger.com, and from quotes given
by vendors of each technology. Where applicable, in-house labor
costs are estimated at $30/h for a duration of time determined
through interviews with plant staff or technology vendors in accordance with best-practices at the IAC. Discounted payback periods
are then computed for each piece of equipment that the new or
retrofit efficient technology is applied. The discount rate used for
this research was 2.4%, equal to the sum of the average annual United States inflation rate over 2014 (1.6%), and the average estimated annual growth rate of industrial electricity costs through
2040 (0.8%) as provided by the Energy Information Agency [26].
Beyond calculating payback period of these technologies to compare to that of BES, the estimated savings were applied to the original load profile to analyze the change in feasibility of BES at each
plant.
2.4. Modeling of BES
BES function, optimization and economics, were modeled in
NREL’s BLAST-BTM model. BLAST-BTM is a MATLAB-based simulation that uses 1 year’s history of the load in a facility to estimate
the potential cost and savings of an optimal BES system, determined by an incremental rate of return (IRR) analysis. Given the
power of the BES system (P), and the energy of the BES system
(E), the modeled upfront costs of a BES system design (C, in 2015
USD) were defined as [24],
C ¼ ð500 PÞ þ ð500 EÞ þ 10; 000
The BLAST-BTM simulation assumes perfect 48 h load forecasting and negligible battery degradation over the life of the system.
For a given battery sizing and input load profile, BLAST-BTM seeks
to minimize the peak net demand over a 1 year time frame as measured over 15-min intervals. The calculation of battery commands
to achieve this end is divided into two time scales. First, a 15-min
net demand target is optimized using a constrained SQP method.
The optimal net demand target represents the lowest level to
which the peak net meter load can be reduced with the support
135
of the battery. Second, faster battery commands are computed
and implemented every 1 min within each 15-min interval to best
achieve the interval load target, taking into account hardware limitations and fluctuations in net demand. This procedure defines an
optimal battery dispatch strategy, which is repeated across a span
of battery sizes to define a battery size optimized for IRR.
The outputs of the BLAST-BTM simulation were used to compute a discounted payback for the BES installation. The comparisons in this research are therefore based on comparisons of
payback periods of conventional energy efficient technologies to
the payback period of BES systems. For those plants that were
instrumented for 1 year, the entire 1 year dataset was input to
BLAST-BTM. For those plants that were instrumented for less than
1 year, the dataset was concatenated until 365 days of continuous
data were present. The rate structure of each facility, adjusting for
taxes, was input for each facility’s load profile, respectively. Both
the original total load profile of the facility and the ‘‘efficient” profile after applying the savings from applicable energy efficient new
and retrofit technology were run through BLAST-BTM (see Fig. 2).
The difference provides a metric of the costs of applying energy
efficient technologies on the economic value of BES, for each facility. An increase in payback period demonstrates that the mechanisms by which BES is achieving energy cost savings are being
usurped by the more conventional energy cost saving technologies.
3. Results
The analysis of each case study focused on comparing the payback of applicable energy efficient technologies to BES, analyzing
the effects of implementing those identified energy efficient technologies on the feasibility of implementing BES, and what a potential incentive program, based on current and real incentive
programs that are currently unavailable in Colorado, might look
like.
3.1. Direct comparison of energy efficient technologies to BES
Relevant and realizable energy efficient retrofit and new technologies were identified during the walk-throughs of each plant.
Although there is overlap in the recommendation opportunities
found at each facility, the applicability of these technologies are
not pre-determined prior to the walk-through and do not have
the same outcomes in terms of cost and savings among the facilities. For each recommendation at each plant and each recommendation for the entire study the aggregate number of times that it
was implemented, its payback period, and modeled load reduction
are presented in Table 1. Due to privacy concerns, the case studies
are referred to as Plant 1, Plant 2, Plant 3, and Plant 4 corresponding to the order in which the energy audits were performed.
Only six different energy efficiency recommendations could be
made, however several were made at more than one facility, and
were recommended on several different pieces of equipment at
the same facility. Of the energy efficient technologies recommended to the four plants in this study, repairing air leaks in the
compressed air lines and upgrading to efficient lighting were the
most common recommendations, as they were implementable at
all four plants (Table 1). In total, 129 energy cost saving recommendations were made (Table 1). The most frequent recommendation was replacing the inefficient fan motors on evaporator units
with electronically commutated motors, which was recommended
for implementation on 60 different motors. However, of that recommendation, 90% were applied at Plant 4. This was a large meat
packaging plant where the entire production space was cooled to
less than 34 F. Upgrading the plants’ current lighting with up-todate efficient lighting had the largest effect on reducing the total
136
K. Nataf, T.H. Bradley / Applied Energy 164 (2016) 133–139
Table 1
Quantity, Payback, and Load Reduction of Energy Cost Saving Technologies Recommended for Each Colorado Manufacturing Plant (NA = Technology was not applicable).
Plant 1
Plant 2
Plant 3
Plant 4
Prorated utility rates
$17.55
$0.04456
/kW
/kW h
$7.55
$0.04456
/kW
/kW h
$16.00
$0.0575
/kW
/kW h
$18.80
$0.0440
/kW
/kW h
New and retrofit
technologies
#
Payback
period
(years)
Electrical load
reduction
(kW)
#
Payback
period
(years)
Electrical load
reduction
(kW)
#
Payback
period
(years)
Electrical load
reduction
(kW)
#
Payback
period
(years)
Electrical load
reduction
(kW)
High efficiency
motors
Efficient lighting
EC fan motors on
evaporators
Repair air leaks
Controls on air
cooled chiller
Replace compressed
air nozzles with
blowers
NA
NA
NA
2
19.64
5.3
4
9.2
1.5
NA
NA
NA
1
6
2.85
3.94
2.0
1.1
6
NA
4.20
NA
23.5
NA
5
NA
2.6
NA
13.3
NA
3
54
0.87
2.38
71.0
6.9
22
1
0.64
1.70
6.3
9.1
9
NA
0.21
NA
9.0
NA
1
NA
0.4
NA
0.3
NA
13
NA
0.04
NA
20.2
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
2
0.35
25.0
Totals
30
1.60
18.5
17
3.90
37.7
10
3.11
15.1
72
1.83
123.1
Plant 3, and Plant 4 were 22.4 years, 34.8 years, 15.6 years, and
10.1 years, respectively.
180
160
3.2. Effects of energy efficient technologies on BES valuation
Electrical Power [kW]
140
120
100
80
60
40
Net Demand before Battery
Net Demand after Battery
Net Demand Target
20
0
0
50
100
150
200
250
300
350
Time [days]
Fig. 2. BLAST-BTM simulation for one year load history of Plant 1. The BLAST-BTM
derived optimal net demand target is plotted along with facility net demand with
and without BES. The BES is controlled to enable the facility to not surpass its net
demand target.
load at Plants 2, 3 and 4 and totaled to 110 kW of total reductions.
Repairing compressed air leaks had the lowest total payback period
of all the recommendations at 0.2 years due to the low cost of these
recommendations. All of the results in Table 1 were calculated for
comparison with the BES results for each plant.
The BLAST-BTM simulation was then used to calculate the associated costs and savings of the collected load profile at each facility.
These four plants include a small plant with a load of 100 kW up to
a large plant reaching peak loads of 650 kW. Each plant had a different electric utility and, therefore, a different rate structure.
Demand rates varied among the facilities from Fort Collins Utility’s
E300 rate ($5.90/kW for peak demand) to Xcel Energy’s Primary
General rate (where $18.795/kW represents a weighted demand
charge to model both coincident and peak demand). Before consideration of any energy efficiency technologies, each of the plants’
utility bills were made up of approximately 50% demand charges,
and 50% electricity charges. For each of these inputs of load profile
and utility rates, BLAST-BTM outputs the characteristics of an optimized BES system design including hardware cost, annual savings,
annual operation and maintenance (O&M) costs, and IRR [24].
Using the hardware cost and annual savings less the O&M costs,
the optimized BES discounted payback period for Plant 1, Plant 2,
To analyze the effects of implementing the other energy cost
savings recommendations on BES valuation, the estimated cost
savings from the facility specific recommendations described in
Section 3.1 were applied to each facility’s corresponding load profile, as illustrated in Fig. 3. If the load reduction was being applied
to a data-logged machine, then it was only applied when a load
was being drawn by that equipment. If the load reduction was
on a non-logged piece of equipment, then the hours of operation
were estimated from interviews with plant staff, and the load
reduction was applied over that time period. These new load profiles therefore model the load of the facility including the energy
savings that would be associated with the implementation of the
energy efficiency technologies. These new load profiles, labeled
‘‘efficient load profiles” in Fig. 3, were then run back through
BLAST-BTM to determine whether or not the economic value available to the energy efficiency technologies and the economic value
available to the BES are mutually exclusive.
Overall, there were negligible changes in payback periods
between the original load profile and the efficient load profile
(Table 2). Although the savings due to the implementation of BES
decrease in all scenarios, the optimal battery energy, and therefore
the BES costs also decrease. The net change in payback period of
less than 0.5 years was judged to be negligible relative to modeling
and economic uncertainties.
3.3. Effects of a hypothetical incentive programs on BES
implementation
An important consideration for understanding the results of this
study is that the purchase of most of the energy efficient technologies considered in this research are incented by the utilities in Colorado, whereas BES is not. However, in other states, notably
California, BES installation is incented by the utilities. Therefore,
two hypothetical incentive programs were considered for the case
studies in this research, to demonstrate the changes to BES valuation a hypothetical Colorado BES incentive program would provide.
The first hypothetical incentive program is based on the realworld incentives offered for BES through Southern California Edison’s (SCE) Self Generation Incentive Program, and is labeled as
the SCE Incentive model. This program offers incentives up to a
137
K. Nataf, T.H. Bradley / Applied Energy 164 (2016) 133–139
demand cost savings for the Colorado simulations were higher relative to the costs of the systems.
800
Facility Total Load
Facility Total Load (Efficient Load Profile)
Air Compressor Load
Air Compressor Load with VFD
Refrigeration Load
Refrigeration Load with Motor/Control Improvements
700
600
4. Discussion
4.1. Comparison of BES to other energy cost saving technologies for
Colorado manufacturers
Load [kW]
500
400
300
200
100
0
0
1
2
3
4
5
6
7
Time [Days]
Fig. 3. Subsample of Plant 4 load profiles before (measured) and after applying
energy efficiency technologies (simulated).
maximum of 60% of the project’s cost. This includes $1.46 per
installed watt of battery storage for the two hour rated capacity
of the system and an additional 20% of the system cost if the system is manufactured by a California supplier [29]. A second hypothetical set of incentives are based on the Xcel Energy custom
rebate program that offers $0.40/W saved by a technology up to
60% of the project’s cost [5], and is labeled as the Colorado Incentive model. It should be noted that not all of the case study locations are in Xcel service territory and therefore are only
hypothetically able to realize Xcel incentives.
The BLAST-BTM simulation has inputs for maximum fractional
incentive, incentive offered in dollar per installed kW, and minimum hours at rated capacity needed for incentive. Thus the simulations were rerun using each plant’s original load profile and the
characteristics of the first hypothetical incentive program’s criterion. Since the second hypothetical incentive program is based
on kW saved and not kW installed, the simulations of the original
load profile from Section 3.1 were analyzed for the maximum peak
reduction for each case study and the incentives were retroactively
calculated and applied. The results from these incentive programs
are shown in Table 3. Payback periods from the SCE Incentive simulations are less than those of Colorado Incentive simulations by
between 2 and 9 years. In the SCE simulations, the battery size
was increased to a size where the two hour rated energy of the battery was high enough to get the full 60% maximum fractional
incentive offered by the utility. The demand reduction fraction relative to the battery sizes were higher for the Colorado Incentive
simulations than for those of the SCE simulations, meaning the
Although the economic value of BES storage has been analyzed
in previous research, no other research has directly compared BES
and other energy cost saving technologies. Neubauer and Simpson
(2015), looking at 98 facilities, showed median paybacks of large
BES systems to be around ten years, with small BES systems reaching as low as three years [15]. These simulations used load data
from EnerNOC’s online database, did not correlate loads with electricity prices, considered only California electricity costs and incentives, and did not have the depth of information required to look at
BES as one of a set of energy cost saving technologies that could be
applied to each facility.
For this study, we can make a more direct comparison between
the payback period of individual energy efficiency implementations and BES implementation. Fig. 4 shows a box plot of the payback periods of the energy cost saving recommendations that had
more than five datum points and the BLAST-BTM optimized BES
systems in each facility. These are not the plant-aggregated payback periods shown in Table 1, but rather the individual paybacks
for each piece of equipment the recommendation was applied too.
Additionally, this figure does not represent outliers (determined by
1.5 times the inter quartile range), thereby removing the paybacks
of six air leak outliers. Rating the energy efficiency technologies
from longest to shortest median payback period and including
those not shown in Fig. 4, are BES, high-efficiency motor replacement, efficient lighting, replacing evaporator fan motors with electronically commutated (EC) motors, adding floating head pressure
controls to the air cooled chiller, replacing compressed air nozzles
with portable blowers, and repairing compressed air leaks. The
maximum BES payback period is almost fifteen years higher than
the maximum payback period of any other energy cost saving recommendation. The minimum payback period for BES is higher than
the median of all other recommendations except for the recommendation to replace conventional shaded pole motors with high
efficiency motors. These results demonstrate that BES is the least
cost-effective option of those considered in this study.
4.2. Comparison of BES payback period to nominal energy efficiency
payback periods
Under the conditions considered for these manufacturers, BES
are unlikely to be implemented in practice. In this study, the
minimum payback period observed for BES was 10 years. Of
the top fifty implemented IAC recommendations ever made,
Table 2
Comparison of BLAST-BTM simulation results before and after applying conventional energy efficiency technologies.
Plant
Load profile input in BLAST-BTM
simulation
Battery size
(kW h)
Inverter size
(kW)
BES system
cost
BES-derived annual
savings
Annual O&M costs
for BES
BES payback period
(years)
Plant
1
Original load profile
Efficient load profile
22
19
19
18
$30,328
$28,426
$2469
$2336
$703
$684
22.4
22.46
Plant
2
Original load profile
Efficient load profile
24
18
32
28
$38,032
$33,015
$2403
$2152
$780
$730
34.85
34.35
Plant
3
Original load profile
Efficient load profile
23
21
16
15
$29,387
$28,351
$2992
$2874
$694
$684
15.58
15.69
Plant
4
Original load profile
Efficient load profile
20
16
29
27
$34,393
$31,903
$4625
$4385
$744
$719
10.09
9.88
138
K. Nataf, T.H. Bradley / Applied Energy 164 (2016) 133–139
Table 3
Effects of Hypothetical Incentive Programs on Simulated BES Value and Payback Periods.
Plant
Load profile input in
BLAST-BTM simulation
Battery size
(kW h)
Inverter
size (kW)
BES
system
cost
BES-derived
annual savings
Annual O&M
costs for BES
Achieved load
reduction (kW)
Incentive
Payback
period
(years)
Plant 1
SCE incentive
Colorado incentive
28
22
21
19
$34,469
$30,328
$2730
$2469
$745
$703
20.6
22.8
$20,681
$9112
7.7
14.3
Plant 2
SCE incentive
Colorado incentive
43
24
40
32
$51,441
$38,032
$3063
$2403
$914
$780
40.3
31.5
$30,864
$12,602
11.0
19.9
Plant 3
SCE incentive
Colorado incentive
25
23
16
16
$30,892
$29,387
$3106
$2992
$709
$694
16.2
15.6
$18,535
$6233
5.6
11.7
Plant 4
SCE incentive
Colorado incentive
38
20
35
29
$46,657
$34,393
$5372
$4625
$867
$744
35.0
28.9
$27,959
$11,564
4.4
6.4
40
25
30
Payback Period [years]
Payback Period [years]
35
25
20
15
10
20
15
10
5
5
0
BES
Motors
Lighting
EC Fan
Motors
Air Leaks
0
Motors
Fig. 4. Box plot comparing discounted payback periods for energy cost saving
recommendations.
BES
with
Colorado
Incentives
BES
with
SCE
Incentives
Lighting
EC Fan Air Leaks
Motors
Fig. 5. Box plot comparing energy cost saving recommendations to BES with
hypothetical incentives applied.
recommended at least 10 times, the longest payback was 4.7 years
[30]. Thus, the range of BES payback for these industrial facilities is
well above the 5 year payback period that the IAC program has
shown to have high implementation success. Additionally, the
results of this study’s economic simulation are based on idealized
conditions including perfect load forecasting, zero battery degradation, and long-term utility rate stability. Payback period for the BES
systems would increase were more real-world conditions able to
be considered.
demonstrate that incentives can be a means to make BES payback
period competitive with conventional energy efficiency technologies. The effect of our set of hypothetical incentives is considered
in Fig. 5. In general, both sets of incentives improve the payback
period competitiveness of BES, but neither is able to reach the
threshold of a five year payback period that would warrant BES’s
consideration as an industrially-relevant energy cost saving
technology.
4.3. Means to improve BES competitiveness for Colorado
manufacturers
5. Conclusion
There are several trends that could improve the cost competitiveness of BES in Colorado and other locations in the future. First,
with lithium-ion battery costs forecasted to decrease in time, BES
technology may become less capital cost intensive [28,31]. The
results of this study demonstrate that implementing conventional
energy efficiency technologies would have negligible changes on
the valuation of BES. Thus, a less capital cost intensive system
would be cost competitive with conventional energy efficiency
technologies. Second, this study was unable to compare BES to
the economic costs and benefits of conventional load shifting techniques. These techniques were not found to be applicable in any of
the facilities or case studies. This perhaps demonstrates how difficult it is to implement these techniques in manufacturing plants
without affecting production, and adds to the argument that BES
has a place in BTM demand reduction applications. Without the
negative implications of productivity loss, BES may be a
production-independent means to achieve peak load reduction in
manufacturing plants. Finally, the results presented in this study
This study collected detailed information on electricity loads in
a set of four Colorado industrial facilities, which allowed for battery storage to be considered part of a set of energy cost saving
technologies that may or may not be implemented together. This
allowed for a direct comparison between the cost-effectiveness
of these technologies and that of BES, showed the implications of
installing these technologies prior to BES, and provided information on what incentives utilities could provide to make BES a more
attractive, energy cost saving technology. Of the six different cost
saving recommendations performed 129 times during the four
case studies, only one, standard-efficiency motor replacement with
high-efficiency motors, had a median payback period within the
range of BES implementation costs. This demonstrates that BES
was the least economical option to reduce energy costs for these
manufacturing facilities. The implementation of BES after the estimated savings of retrofit and cost saving technologies were applied
to the originally collected loads showed negligible changes in
discounted payback period, demonstrating that implementing
K. Nataf, T.H. Bradley / Applied Energy 164 (2016) 133–139
conventional energy efficiency technologies does not decrease the
valuation of BES. Adding hypothetical incentives, in general, lowers
the paybacks of BES but does not move BES below the five year
payback period threshold. Together these results demonstrate that
BES is not currently competitive with other energy cost saving
technologies. As battery prices decrease, BES technology may
become more cost competitive and would then have potential to
be implemented into the market as it reduces energy costs and
can be used in combination with more conventional energy cost
saving technologies.
Acknowledgements
The authors acknowledge the technical contributions of
Dr. Jeremy Neubauer of the National Renewable Energy Lab, Mike
Kostrzewa of the Colorado State University Industrial Assessment
Center.
References
[1] Electric Power Research Institute. A system for understanding retail
electric rate structures. Technical update 1021962; 2011. Retrieved
from <http://www.epri.com/abstracts/Pages/ProductAbstract.aspx?ProductId=
000000000001021962>.
[2] Liu Z, Wierman A, Chen Y, Razon B, Chen N. Data center demand response:
avoiding the coincident peak via workload shifting and local generation.
Perform Eval 2013;70(10):770–91.
[3] Rohdin P, Thollander P, Solding P. Barriers to and drivers for energy efficiency
in the Swedish foundry industry. Energy Policy 2007;35(1):672–7.
[4] Park CW, Kwon KS, Kim WB, Min BK, Park SJ, Sung IH, et al. Energy
consumption reduction technology in manufacturing—a selective review of
policies, standards, and research. Int J Prec Eng Manuf 2009;10(5):151–73.
[5] Business programs summary; 2014. Retrieved from <http://www.xcelenergy.
com/staticfiles/xe/Marketing/Files/CO-Business-Programs-List.pdf>.
[6] Middelberg A, Zhang J, Xia X. An optimal control model for load shifting–with
application in the energy management of a colliery. Appl Energy 2009;86
(7):1266–73.
[7] Di Giorgio A, Pimpinella L. An event driven smart home controller enabling
consumer economic saving and automated demand side management. Appl
Energy 2012;96:92–103.
[8] Patteeuw D, Bruninx K, Arteconi A, Delarue E, D’haeseleer W, Helsen L.
Integrated modeling of active demand response with electric heating systems
coupled to thermal energy storage systems. Appl Energy 2015;151:306–19.
[9] Zhang Y, Wang Z, Zhou G. Determinants and implications of employee
electricity saving habit: an empirical study in China. Appl Energy
2013;112:1529–35.
[10] Strbac G. Demand side management: benefits and challenges. Energy policy
2008;36(12):4419–26.
[11] Herrmann C, Thiede S. Process chain simulation to foster energy efficiency in
manufacturing. CIRP J Manuf Sci Technol 2009;1(4):221–9.
[12] Beer S, Gomez T, Dallinger D, Momber I, Marnay C, Stadler M, et al. An
economic analysis of used electric vehicle batteries integrated into
commercial building microgrids. IEEE Trans Smart Grid 2012;3(1):517–25.
[13] Eyer J, Corey G. Energy storage for the electricity grid: benefits and
market potential assessment guide. Sandia National Laboratories technical
report SAND2010-0815; 2010. Retrieved from <http://www.sandia.gov/ess/
publications/SAND2010-0815.pdf>.
139
[14] Hanna R, Kleissi J, Nottrott A, Ferry M. Energy dispatch schedule optimization
for demand charge reduction using a photovoltaic-battery storage system with
solar forecasting. Sol Energy 2014;103(1):269–87.
[15] Neubauer J, Simpson M. On the deployment of behind-the-meter energy
storage for demand charge reduction. National Renewable Energy Lab
technical report TP-5400-63162; 2015. Retrieved from <http://www.nrel.gov/
docs/fy15osti/63162.pdf>.
[16] Ekren O, Ekren BY, Ozerdem B. Break-even analysis and size optimization of a
PV/wind hybrid energy conversion system with battery storage–a case study.
Appl Energy 2009;86(7):1043–54.
[17] Neubauer J, Pesaran A. The ability of battery second use strategies to impact
plug-in electric vehicle prices and serve utility energy storage applications. J
Power Sources 2011;196(23):10351–8.
[18] Bradbury K, Pratson L, Patiño-Echeverri D. Economic viability of energy storage
systems based on price arbitrage potential in real-time US electricity markets.
Appl Energy 2014;114:512–9.
[19] Paulus M, Borggrefe F. The potential of demand-side management in energyintensive industries for electricity markets in Germany. Appl Energy 2011;88
(2):432–41.
[20] Public Utilities Commission of California. Decision 13-10-040; 2013. Retrieved
from <https://www.sce.com/wps/wcm/connect/435ea164-60d5-433f-90bcb76119ede661/R1012007_StorageOIR_D1310040_AdoptingEnergyStorage
ProcurementFrameworkandDesignProgram.pdf?MOD=AJPERES>.
[21] Mulder G, Six D, Claessens B, Broes T, Omar N, Van Mierlo J. The dimensioning
of PV-battery systems depending on the incentive and selling price conditions.
Appl Energy 2015;111:1126–35.
[22] Sutherland RJ. Market barriers to energy-efficiency investments. Energy J
1991:15–34.
[23] Muller MR, Barnish T, Polomski PP. On decision making following an industrial
energy audit. In: Proceedings of the 17th national industrial energy technology
conference; 1995. p. 150–5.
[24] Neubauer J. Battery lifetime analysis and simulation tool (BLAST)
documentation. National Renewable Energy Lab technical report TP-540063162; 2014. Retrieved from <http://www.nrel.gov/docs/fy15osti/63246.pdf>.
[25] Therkelsen P, McKane A. Implementation and rejection of industrial steam
system energy efficiency measures. Energy Policy 2013;57:318–28.
[26] Energy Information Agency. Energy prices by sector and source. United States;
2013. Retrieved from <http://www.eia.gov/oiaf/aeo/tablebrowser/#release=
AEO2014ER&subject=0-AEO2014ER&table=3-AEO2014ERÒion=1-0&cases=
full2013-d102312a,ref2014er-d102413a>.
[27] Spees K, Lave LB. Demand response and electricity market efficiency.
Electricity J 2007;20(3):69–85.
[28] Neubauer J, Simpson M. Optimal sizing of energy storage and photovoltaic
power systems for demand charge mitigation. In: Electrical energy storage
applications and technologies (EESAT) conference, 20–23 October 2013, San
Diego, California; 2013.
[29] Southern California Edison. Self-generation incentive program; 2015.
Retrieved from <https://www.sce.com/wps/portal/home/business/generatingyour-own-power/incentive-program/!ut/p/b1/hc_NCoJAFAXgZ2nRMueoZbac
CZvGpKEfzGYTFjYJ5oRFvn4TtCmozu7Ad-EeokhGVJ3fS53fSlPn1bOrYOeGnE7F
CkIyEUCwMR1LGQN9WLC1AF9C8e9-Q9Q7CdkotGSWRG7icc7xCfhy6EHEaZRI
5nroey8w4oimsbRgvfAh_AXmK0p9IHiBH0_GROnK7O3gDbN2kHYR6Oc2Wu_
9UBPVFMeiKRrnZK43krVt62hjdFU4B3Mml3OGUvTURHc6D5AO-78!/dl4/d5/
L2dBISEvZ0FBIS9nQSEh/?from=sgip>.
[30] Industrial Assessment Centers Database – IAC top ten; 2015. Retrieved from
<http://iac.rutgers.edu/database/topten/?ARC=&SIC=&NAICS=&YEAR_limit=%
3C%3D&YEAR=2015&STATE=&CENTER=&SHOW_TOP=50&SORTBY=0&MIN_
RECC=10&generate=Generate>.
[31] Neubauer JS, Pesaran A, Williams B, Ferry M, Eyer J. A techno-economic
analysis of PEV battery second use: repurposed-battery selling price and
commercial and industrial end-user value. SAE technical report no. 2012-010349; 2012. Retrieved from <http://www.nrel.gov/docs/fy12osti/53799.pdf>.