20200827Revised Comments.pdf

EDWARD J. JEWELL
DEPUTY ATTORNEY GENERAL
IDAHO PUBLIC UTILITIES COMMISSION
PO BOX 83720
BOISE, IDAHO 83720-007 4
(208) 334-O3t4
IDAHO BAR NO. 10446
IN THE MATTER OF IDAHO POWER'S
PETITION TO DETERMINE THE PROJECT
ELIGIBILITY CAP FOR PT]BLISHED AVOIDED
COST RATES AND THE APPROPRIATE
CONTRACT LENGTH FOR ENERGY
STORAGE QUALIFYING FACILITIBS
Jl r: fi l: !'. /t ilt'il'^.'\ei..l9 !**L;
:i:* Ar.jr e? P* 5: 09
,i-. . lJ t.'.-t;Lit
: : ii.i i ;a;: tif;},llr{iS$IOH
Street Address for Express Mail
11331 W. Chinden Blvd.
Bldg.8, Ste. 201-A
Attorney for the Commission Staff
BEFORE THE IDAHO PUBLIC UTILITIES COMMISSION
)
)
)
)
)
)
)
CASE NO. IPC.E.2O.O2
REVISED COMMENTS OF THE
COMMISSION STAFF
The Staff of the Idaho Public Utilities Commission, by and through their attorney of
record Edward Jewell, comments as follows on Idaho Power Company's Petition.
BACKGROUND
On January 2l,202o,Idaho Power Company ("Company") filed a petition requesting the
Commission determine avoided cost rates, contract terms, and conditions applicable to energy
storage qualifying facilities ("QF" or "QFs") under the Public Utility Regulatory Policies Act of
1978 ("PURPA").
PURPA, and Federal Energy Regulatory Commission ("FERC") rules require state
commissions to establish published avoided cost rates for QFs with a nameplate capacity of 100
kilowatts ("kW") or less. 18 C.F.R. * 292.3O4(cXl). The state commissions, in their discretion,
may establish published avoided cost rates for QFs greater than 100 kW. 18 C.F.R.
* 292.304(c)(2). State commissions may differentiate among QFs using various technologies
based on the supply characteristics of the resource. 18 C.F.R. E 292.304(cX3Xii).
STAI]F COMMENTS AUGUST 2I ,20201
In Idaho, this Commission has established a 100 kW project eligibility cap for wind QFs
and solar QFs to receive published avoided cost rates. Order Nos. 32262 at 8,32697 at 13. All
other QF types have a 10 average Megawatt ("aMW") project eligibility cap for published
avoided cost rates. Order No. 32697 at 14. Published avoided cost rates in Idaho are calculated
using a surrogate avoided resource-a combined cycle combustion turbine ("CCCT")-that is
assumed to be the Company's marginal resource IOOVo of the time ("SAR Method" or
"Published Rate Method"). See Order No. 32697 at 17. Negotiated rates, which are available for
QFs above the project eligibility cap, are calculated by the Incremental Cost Integrated Resource
Plan Method ("ICIRP Method"). The ICIRP Method calculates the marginal value of energy on
the Company's system on an hourly basis given the Company's actual resource stack. See Order
No.32697 at20-21.
In addition to eligibility for published avoided cost rates, the project eligibility cap
determines the length of contract for which a QF is eligible. Those above the project eligibility
cap in Idaho are entitled to two-year contracts. Order No. 33357 at25. Those below the project
eligibility cap in Idaho are entitled to 20-year contracts. See Order No. 33253 at 4.
In Order No. 33785, the Commission determined five energy storage QFs were entitled to
the terms and conditions available to solar QFs because the generation profiles of those QFs
aligned closely to the generation profiles of solar QFs, and based on the Commission's
interpretation of Luz Development and Finance Corporation, 51 FERC 1[61,078 (1990). Order
No. 33785 at ll-12. See also Order No. 33858 at 3.
On January 17,2020, the United States District Court for the District of Idaho issued a
Memorandum Decision and Order in Franklin Energy Storage One et al. v. Kjellander et al.,
Case No. 1:18-cv-00236-REB, holding that the Commission's decision in Order No. 33785
"established an implementation plan that impermissibly classified the QF status of Plaintiffs'
energy storage facilities that are certified under [PURPA] as energy storage facilities."
Memorandum Decision at37. "Classifying such facilities as 'solar QFs' is outside the
Commissioners' authority as state regulators and therefore in violation of federal law." Id.
While finding that the Commission could not treat these energy storage QFs as solar QFs, the
Court specifically declined "to order [the Commission] to require utilities under their jurisdiction
to afford energy storage QFs all rights and privileges afforded to 'other QFs' under the IPUC's
PURPA implementation plan." Id.
2STAFF COMMENTS AUGUST 27,2020
In response to the Memorandum Decision, and to two energy storage QF applications
received by Idaho Power immediately following the Memorandum Decision,Idaho Power filed
this petition requesting the Commission determine the proper avoided cost rates and contract
terms applicable to energy storage QFs. Petition at 5. Idaho Power requests the Commission
establish a 100 kW eligibility cap for energy storage QFs to receive published avoided cost rates
and 20-year contracts, the minimum project eligibility cap allowed by FERC rules. Idaho Power
requests any energy storage QF above the eligibility cap receive avoided cost rates calculated by
the ICIRP Method and be eligible for two-year contracts. Petition at 2.
On July 16,2020, Commission Staff filed a Request for Public Input and Initial
Comments. Public Input was received from Renewable Northwest,Idaho Conservation League
("ICL"), and Clenera, LLC.
Staff now files these revised comments reflecting public input received and further
analysis completed.
STAFF REVIEW
I. Introduction and scope.
Idaho Power requests "a determination from the Commission that energy storage QFs up
to a maximum nameplate capacity of 100 kW are entitled to and eligible for published avoided
cost rates and a 20-year maximum contract term and that energy storage QFs over 100 kW are
entitled to and eligible for negotiated avoided cost rates determined by the TICIRP Methodl and a
maximum contract term of two-years." Petition at 11. The Company's petition responds to an
immediate need to determine the terms for energy storage. Staff recommends the scope of this
docket be limited to addressing Idaho Power's specific request and determine how battery
storage QFs fit within the existing PURPA framework in Idaho. Staff also recommends a later
docket or dockets to align battery storage QF terms and conditions for Idaho's other electric
utilities and to address potential enhancements to Idaho's avoided cost methodologies that would
allow for greater granularity in avoided cost rates and compensation of ancillary services.
Staff believes the scope of this docket is properly limited to battery storage rather than
the more generic concept of energy storage. Staff believes it is appropriate to limit the
discussion to battery storage because there is a wide variety of characteristics of energy storage
technologies, and battery storage is currently the industry leader in energy storage technologies.
Staff also recognizes there is a diversity of characteristics within battery storage based on
STAFF COMMENTS AUGUST 27 ,20203
different battery chemistries and designs, but Staff believes it is reasonable to distinguish battery
storage QFs as their own category within Idaho's PURPA framework because they are
reasonably distinct from other types of energy storage.
Staff recommends opening a multi-utility generic case after the conclusion of this case to
determine and implement the rights and privileges of battery storage QFs for all three of Idaho's
major electric investor-owned utilities. This approach would respect previous Commission
statements encouraging alignment of PURPA implementation among Idaho's utilities. E g.
Order No. 29880 (stating, "It is reasonable for QFs to expect that the contract requirements of
Idaho's regulated electric utilities will be similar and that a QF will not be disadvantaged by
choosing to sell to one utility rather than another.") Second, it would allow further refinement of
Idaho's avoided cost methodologies to more accurately value energy and capacity based on the
time of production, and account for ancillary services provided by battery QFs. Finally, it would
allow for examination of potential updates to Idaho's avoided cost methodologies based on
FERC Order No. 872, Qualifying Facility Rates and Requirements Implementation Issues Under
the Public Utility Regulatory Policies Act of 1978, issued July 16,2020. Public input generally
reflected commenting parties' desire to continue the investigation into battery storage QFs in
Idaho.
II. Analvsis approach.
Staff s main objective for its analysis in this case was to determine where battery storage
QFs fit within Idaho's existing implementation of PURPA. A flowchart illustrating Idaho's
current PURPA implementation is provided below.
STAI]F COMMENTS AUGUST 27 ,20204
Wind or
Sol
kw
All
Other
<= 100 kW <= 10 aMW > 10 aMW
Wind
Solar
Custom for
each QF
Non-seasonal Hydro
Seasonal Hydro
Fossil-fueled Cogen
Other - includes:
- Biomass
- Biogas
- LandfillGas
- Geothermal
- Waste-to-energy
- Non-fossil fueled Cogen
Custom for
each QF
To meet this objective, Staff compared the characteristics of battery storage QFs with the
characteristics of wind, solar, and other types of QFs and considered relevant factors important to
each ofthe four key questions Staff addressed in its analysis:
1. Do the unique characteristics of battery storage justify its own category?
2. What is the appropriate method to value the output from battery storage QFs
among Idaho's presently established methods?
3. What is the appropriate project eligibility cap for a battery storage QF to be
entitled to published avoided cost rates?
4. What is the appropriate contract length for battery storage QFs above and below
the project eligibility cap?
Staff did not approach the case as a strict dichotomy between whether battery storage QFs should
receive the same treatment as wind, solar or "other" QFs, instead opting to use these categories
as guideposts.
Circumstances have changed since Idaho's PURPA framework was developed.
Additionally, Staff and the Commission now have the benefit of hindsight to understand the
results of previous Commission orders relative to its objectives. By examining historical data,
STAFF COMMENTS AUGUST 27,2O2O5
QF for
Eligibility
Size of
Eligibility
cap
Size of
Eligibility
cap
Method
SAR
Method
IRP
Method
Up to 20-Yr Up to 2-Yr
Contract
Up to 20-Yr
Rate Rate Rate Rate
Staff has been able to measure the effect of the current framework and determine how current
policies are meeting Commission goals, as well as PURPA's overall objectives. Furthermore,
significant learning has taken place both in Idaho and other state jurisdictions that can shed light
on potential improvements to Idaho's current framework.
III. The Commission should establish a catesorY for batterv storase OFs.
Staff identified several key reasons why a separate category for battery storage QFs
should be established based on differences in the output characteristics of battery storage QFs as
compared to other QF resource types. Other baseload QF resource types do not have the same
limitations with respect to the duration of output, as do battery storage QFs. Battery storage QFs
are capable ofproviding a host of ancillary services that other resources are not able to provide
such as frequency regulation, variable energy firming and smoothing, black start/voltage support,
load following, reserve capacity,local capacity, and deferrals of transmission and distribution
investments. Batteries also have a steeper degradation factor than other resources, which varies
based on the cycling of the battery storage QF. Wind and solar QFs are intermittent while
battery QFs can be dispatched. Because of these reasons, Staff believes it is appropriate for the
Commission to establish a separate battery storage QF category.
IV. Staffanalvzed the relative merits of the ICIRP Method and the Published Rate
Method for valuins the output from batterv storage OFs.
To determine the reasonableness of the ICIRP and Published Rate Methods for battery
storage QFs, Staff first performed comparisons between the two methods, then analyzed the
appropriateness of ICIRP Methods and Published Rate Methods for both energy and capacity.
Both methods determine an overall avoided cost rate, composed of two separate components: an
avoided cost rate for capacity component, and an avoided cost rate for energy component. Each
component uses a separate and distinct method to derive their respective avoided cost rates.
Because they are separate, Staff performed comparisons between the two methods for each
component and evaluated the relative merits of each component of each method separately.
Additionally, if a method was not appropriate, Staff identified changes required to make them
appropriate for battery storage, as well as identified potential incremental improvements.
Through its analysis, Staff came to the following conclusions listed below.
STAFF COMMENTS AUGUST 27 ,20206
1. The ICIRP Method is overall more accurate because it is more sophisticated,
includes more variables that affect avoided cost, makes more reasonable
assumptions, and is updated more frequently than the current Published Rate
Method.
2. The ICIRP Method for determining avoided energy cost rates and payments is
reasonable for battery storage QFs but could be improved.
3. The ICIRP Method for determining avoided capacity cost rates and payments is
reasonable for battery storage QFs but could be improved. Staff recommends
capacity payments be separated from energy payments. Capacity payments should
only be made for energy delivered during peak capacity months and hours, and
only after the Company becomes capacity deficient.
4. The current Published Rate Method for determining avoided energy cost rates and
payments is reasonable for battery storage QFs under the project eligibility cap.
However, to reduce the difference between Published Rate Method and ICIRP
Method rates and to improve accuracy, Staff believes that the method should be
changed to better align to the ICIRP Method in a later docket.
5. The Published Rate Method for determining avoided capacity cost rates and
payments is not reasonable for battery storage QFs. To make it reasonable for
battery storage QFs, Staff recommends capacity payments be separated from
energy payments. Capacity payments should only be made for energy delivered
during peak capacity months and hours, and only after the Company becomes
capacity deficient. As an improvement and to reduce the difference between
Published Rate Method and ICIRP Method rates, Staff believes that the surrogate
be changed to a Simple Cycle Combustion Turbine ("SCCT"), the same used in
the ICIRP Method, and input assumptions should be updated on a more regular
basis.
More granular avoided cost rates, for both energy and capacity, should be implemented
before the time-shifting value of battery storage QFs can be fully recognized. By fine-tuning the
existing ICIRP Method of calculating avoided cost rates to more closely reflect the Company's
system needs, the Commission can align economic incentives for QFs with the Company's
needs, thereby ensuring that QFs are fairly compensated and ratepayers are indifferent. This
change could be applied to battery storage QFs in Idaho Power's service territory first in this
STAFF COMMENTS AUGUST 27 ,20207
docket, to battery storage QFs in the service territories of Idaho's other major electric utilities in
a subsequent docket, and to other QF types in a later docket, if the Commission so desires.
In a reportl conducted by Pacific Northwest National Laboratory ("PNNL") to assist
Commission Staff in this docket, PNNL identified battery storage QFs as a potential opportunity
to make PURPA more resource agnostic and more focused on the QF's contribution to the grid,
regardless of its resource type. This could be accomplished by time of delivery price signals
when energy and capacity are most valuable to the grid.
a. Staff compared the ICIRP Method and Published Rate Method.
i. Comparison between ICIRP Method and Published Rate Method for avoided
cost ofenergy.
Whereas the current Published Rate Method assumes that a CCCT natural gas plant is the
marginal cost resour ce I00Vo of the time, the ICIRP Method uses a production cost model
(AURORAxmp) that simulates the Company's operation of its entire system, capturing the
displaceable incremental cost resource at the top of the Company's resource stack for every hour
of operation throughout the term of a contract. Based on the Company's response to Staff
Production Request No. 16, Langley Gulch, the only CCCT in the Company's system, is the
marginal resource only 9.37o of the time.
As shown in Table No. 1 below, the ICIRP Method incorporates more variables and is
updated on a more regular basis. Variables such as customer load, the cost and amount of power
purchase agreements, and market electricity prices are included in the ICIRP Method and
excluded from the Published Rate Method. Additionally, in the ICIRP Method, the Company
updates the resource stack with each QF application, whereas in the Published Rate Method, the
resource stack is not updated when new QFs come online. Without a queue, all proposed QFs
are treated as if they are all the first project to receive the next indicative pricing, resulting in
inaccurate avoided cost rates because earlier projects will displace the high-valued resources in
the resource stack, and later projects will displace lower valued resources in the resource stack.
Also, if a higher-positioned QF drops out of the queue, all the QFs after that project will get
I PNIIL Report is provided for public use and can be found at
https:/&uc.idaho.gov/Fileroom/PublicFiles/ELEC/IPC/IPCE2002/CaseFiles/20200827PNNL7c20Report.docx
Staff acknowledges the research performed by PNNL, but the findings and opinions expressed are those of the authors
and do not necessarily reflect the official policy or positions of Idaho Public Utilities Commission and Staff.
8STAFF COMMENTS AUGUST 2] ,2020
updated indicative pricing to ensure avoided costs are accurate. Staff believes that because of
these factors, the ICIRP Method produces more accurate results than the Published Rate Method
for determining avoided energy costs.
Table No. 1: ICIRP and Published Rate Methods Avoided Energy Inputs Comparison
Staff quantified the difference in ICIRP Method rates and Published Rate Method rates.
To perform its analysis, Staff created "reference" projects with l2-month by 24-hour generation
profiles for each QF type by averaging actual generation data from up to 6 actual QFs in
operation and under contract with the Company. To establish a relative comparison, Staff
normalized each reference project to be equivalent to a l0 MW nameplate project and asked the
Company to develop the energy and capacity avoided cost ICIRP Method rates across a 2O-year
period. After taking the net present value ("NPV") over 20 years, Staff compared these rates
with the NPV of these same reference projects over the same 20 years but using published rates.
Table No. 2 below illustrates the difference between the Published Rate Method avoided
energy cost rates and ICIRP Method avoided energy cost rates.
STAI]F COMMENTS AUGUST 27 ,20209
Avoided Enersv ICIRPMethod Published Rate Method
Method Custom based on generation profile of each
project and the marginal cost resource at
top of the generation stack in each hour
across contract term.
Assumes CCCT is
marginal cost resource
being avoided lolVo of
the time.
Natural Gas Price Updated every year using IRP forecast Updated every year using
Mountain Region EIAr
forecast
Performance of
Generation Resources
Updated each IRP every two years Heat Rate of CCCT
based on 2008 NWPP2
Market Electricity
Prices
WECC3 market prices generated each
AURORATn for each contract.
Not used
Power Purchase
Agreements
New, terminated, or expired contracts are
updated on a continuous basis.
Not used
QF Queue QF application queue is maintained real-
time and is included in the resource stack
in AURORA.
Not used
Forecasted Customer
Load
Updated annually in October Not used
I EIA - Energy Information Administration
2 NWPP - Northwest Power Plan3WECC - Western Energy Coordinating Council
Table No. 2: Published Rate Method and ICIRP Method Avoided Energy Cost Rates.
As can be seen from Table No. 2, published avoided energy cost rates range from
$5.45ny[Wh higher to $9.08AvtWh higher. Staff believes that aligning the source of the natural
gas forecast for both methods may reduce the difference between the two methods but would not
address the flawed assumption that a CCCT is always the marginal resource on the Company's
system. Staff did not have a battery storage QF generation profile to determine the likely ICIRP
Method rates for abattery storage QF because there is not historical data for battery storage eFs
in Idaho. Staff believes that a battery storage QF would be able to take some actions to tailor its
generation profile to capture additional value by producing at more high load hours, and
therefore would likely receive somewhat higher ICIRP Method rates than those reflected above.
ii. Comparison between ICIRP Method and Published Rate Method for
Avoided Cost of Capacity.
The factors used to derive avoided capacity rates for the ICIRP Method and the published
Rate Method are compared in Table No. 3 below.
Table No.3: ICIRP and Published Rate Methods Avoided Capacity Inputs Comparison.
Reference Project Avoided Energy Cost Differences
Levelized over 20years without integration charges
Rates authorized December 2019
Negotiated Rate
ICIRP Method (S/rvrwnl
(5/rvtwnl (s/MWh)
d Rate
Method Difference
(s/MWh)% Difference
Wind (10 MW nameplate)
Solar (10 MW nameplate)
Non-seasonal Hydro (10 MW nameplate)
Seasonal Hydro (10 MW nameplate)
Other (10 MW nameplate)
35.15
33.28
35.00
31.52
34.97
40.60
40.50
40.60
40.50
40.50
ss.4s
57.32
ss.so
ss.08
$s.ss
t5.5%
22.O%
t6.wo
28.8%
L6.7Yo
Avoided Capacity Cost ICIRPMethod Published Rate Method
Method Based on fixed cost of SCCT
surrogate
Based on fixed cost of CCCT
surrogate
Cost of surrogate Updated every two years
with IRP acknowledgement
Set in GNR-E-08-02
Capacity Factor Custom using QF's
generation profile
Set in GNR-E-I1-03
Capacity Contribution at
Peak
Custom using QF's
generation profile
Set in GNR-E-I1-03
Deficit Date Updated every two years
using IRP load/resource
balance
Updated every two years using
IRP load/resource balance
STAFF COMMENTS 10 AUGUST 27 ,2020
Table No. 4 below illustrates the difference between the Published Rate Method and the
ICIRP Method avoided capacity cost rates.
Table No.4: Published Rate Method and ICIRP Method Avoided Capacity Rates.
Examining the difference in avoided capacity cost rates shows that there are large
differences ranging from a negative difference of $2.02lIWWh for wind and a positive difference
of $16.64llvtWh for seasonal hydro. Both the Published Rate Method and the ICIRp Method use
a surrogate resource for calculating the avoided cost of capacity. However, this is where the
similarities between the methods diverge. The Published Rate Method uses a CCCT plant for its
surrogate, which has capital and fixed costs that are significantly more expensive than a SCCT
natural gas plant on a dollar per Megawatt ("MW'') basis. At the time that the Commission
reaffirmed the use of a CCCT as the surrogate for published rates, the Commission stated,
We further find it reasonable to continue to utilize a combined-cycle
combustion turbine (cccr) surrogate as the basis for all
calculations in the SAR model. The SAR Methodology is intended
to represent a surrogate base load natural gas resource. Simple cycle
combustion turbines (SCCT) are primarily utilized for meeting autility's peak loads; CCCTs provide base load energy. The
proposals of some of the parties to use an SCCT for calculating
capacity value and a CCCT to compute energy value would create avery awkward and not representative surrogate resource.
Consequently, we decline to utilize a SCCT.
Order No.32697 at 17.
However, Staff believes that continuing to use a CCCT as a surrogate for both capacity
and energy avoided cost has downsides and should be re-examined for several reasons. First, as
shown by Staff s analysis and discussed later, using different surrogates contributes to
differences in avoided cost rates between the Published Rate Method and ICIRP Method rates,
contributing to the problem of disaggregation. Second, Staff has shown that a CCCT as a
STAFF COMMENTS 1l AUGUST 27 ,2020
Reference Project Avoided Capacity Cost Differences
Levelized over 20 years without integration charges
Rates authorized December 2019
Negotiated Rate
ICIRP Method (S/MWh)
(S/tvtwtrl
I publ ish"d
lsurrogrte
lrslrurwt't
Rate
Method Difference
(s/MWh)% Difference
Wind (10 MW nameplate)
Solar (10 MW nameplate)
Non-seasonal Hydro (10 MW nameplate)
Seasonal Hydro (10 MW nameplate)
Other lOMW nameplate)
3.84
15.92
7.74
tL.82
4.90
L.82
18.55
L7.Ot
28.46
12.58
(Sz.oz1
$2.73
5g.zt
s16.54
s7.68
-52.5%
L7.L%
L79.8%
L40.4%
t56.8Yo
sulrogate may no longer be a good proxy for the avoided cost of energy in the Company's
system based on Langley Gulch being a marginal resource less than I07o ofthe time. Third,
since the value of a CCCT can provide both capacity and energy, it doesn't make sense to use the
full investment cost of a CCCT for the avoided cost of capacity when much of the value from a
CCCT is used to provide energy. Finally, in Order No.3269l,when determining the appropriate
basis for determining the avoided cost of capacity for the ICIRP Method, the Commission stated,
We further find that a simple-cycle combustion turbine (SCCT) is
the most appropriate basis for computing capacity value for all
resource types. SCCT's are added to a utility's resource portfolio to
satisfy capacity needs. Because energy and capacity are being
calculated separately, it is reasonable to use a SCCT because it
represents the lowest cost, nearly capacity only resource.
Id. at 19.
The other factor causing differences between the Published Rate Method and the ICIRP
Method is the frequency of updates for key inputs used in the calculations. The annual capacity
factor and capacity contribution at peak used in the Published Rate Method are based on amounts
set in Case No. GNR-E-I1-03, which occurred approximately nine years ago. This is compared
to the ICIRP Method, which uses annual capacity factors and capacity contribution at peak that
are unique for each project and are determined from the generation profile provided by the
project developer.
Staff believes developing an Idaho Power specific method of calculating and paying
capacity costs based on output actually delivered during peak times, similar to the Duke Energy
method, and applying this method to QFs below and above the project eligibility cap would be
an improvement to both methods. Alternatively, Staff believes that the difference in the two sets
of avoided capacity cost rates could be reduced by using the same surrogate and by updating the
assumptions used in calculating published rates on a more frequent basis.
b. Appropriateness of the ICIRP Method for energy and capacity.
As mentioned earlier, Staff believes the ICIRP Method provides more accurate avoided
cost rates than does the current Published Rate Method, but there is room to strengthen the
ability of the ICIRP Method energy and capacity methods to differentiate the value of energy
based on when it is produced.
STAI]F COMMENTS t2 AUGUST 27 ,2020
i. Avoided energy value using the ICIRP Method.
Under the ICIRP Method, the avoided cost of energy is determined by the displacement
of the Company's marginal resource each hour in the Company's ICIRP model given a l2-month
by Z4-hour generation profile supplied by the developer regardless of the type of QF. Staff
believes that this is a conceptually sound method of deriving an avoided cost rate because the
model reflects the Company's entire portfolio of resources, and numerous forecasts and other
inputs that are vetted through a public process. The resulting accuracy of the overall process
depends on two things. First, if the generation profile submitted by the QF is accurate, the
avoided cost will be as accurate as the ICIRP model. Second, the accountability of the QF to
provide generation according to its forecast is based upon contract provisions that assess a price
adjustment when actual output from the QF deviates from the output generation profile. If these
provisions are sufficient and enforced, Staffbelieves that the avoided cost ofenergy paid to the
QF will reflect the marginal value of energy on the Company's system, thereby holding
customers reasonably indifferent while equitably compensating the QF developer.
ii. Avoided capacity value using the ICIRP Method.
The ICIRP Method for determining the avoided cost of capacity looks at the amount of
generation forecasted to be delivered by the QF from 3 to 7 pm during the month of July, as
determined by an output generation profile submitted by the QF. The avoided cost is based on
the investment costs of an SCCT surrogate gas plant. Capacity payments begin based on the
capacity deficiency date authorized at the time that the first contract is signed. Once capacity
payments begin to be paid, they are spread across every unit of QF production, regardless of
whether those units of production are supplied to the utility during months or times of the day
when the utility needs incremental capacity.
Staff believes certain features of the ICIRP Method for calculating capacity are adequate,
and others should be improved. Staff believes a SCCT is a reasonable surrogate for capacity
costs because it is a more reasonable proxy resource in terms of cost to meet a capacity
deficiency. Staff believes the Company's definition of system peak, from 3 to 7 pm during the
month of July, adequately reflects the Company's system and is determined through the public
IRP process. Staff believes it is reasonable for QFs to not be paid capacity payments until the
utility's first capacity deficiency date, because prior to this date, the Commission has determined
that the QF is not avoiding incremental capacity cost. Staff is concerned that the practice of
STAFF COMMENTS 13 AUGUST 27 ,2O2O
paying for capacity during all months and hours of QF production does not adequately
incentivize a QF to actually contribute to the Company's capacity needs. Therefore, eFs that do
not actually contribute to the Company avoiding capacity costs are over-compensated for their
production, and QFs that do contribute to the Company avoiding capacity costs are under-
compensated for their production under the current method.
Staff believes that a method used by Duke Energy would be a significant improvement
over the current method for calculating capacity. Similar to the ICIRP Method, Duke Energy
uses a SCCT as a surrogate resource to determine capacity costs. However, Duke Energy only
pays the QF avoided capacity cost payments if the QF generates during specific pre-determined
months of the year and hours of the day that reflect the Company's peak hours. If the QF
generates during those hours, it gets capacity payments. If it does not generate during those
hours, it does not get capacity payments. This method provides a price incentive to align utility
costs and needs with QF output and ensures the QF provides the avoided cost capacity benefit in
return for their compensation. Staff recommends that the Commission adopt this method of
calculating avoided capacity costs for QFs that use the ICIRP Method and the published Rate
Method in this case. Staff recommends the Commission direct the Company to make a
compliance filing proposing the Company's implementation of this method within 30 days and
provide Staff and the public the opportunity to review the filing before implementation.
Doing so would help alleviate the concerns identified by PNNL and public input that
there is not sufficient granularity in the existing rates to adequately incent the development of
battery storage QFs. Public comments reflected an acknowledgment that current avoided cost
methodologies must be improved upon in order to accurately capture the value of services
provided by battery storage QFs to the grid. ICL states that intra-hour compensation needs to be
defined and a wider array of benefits unique to battery storage need to be recognized. ICL
Comments at2-3. Renewable Northwest details the types of grid benefits battery storage eFs
are capable of providing given their ability to react to signals in a fraction of a second, much
faster than conventional thermal plants. Renewable Northwest Comments at 5. Although
shifting to an Idaho Power specific method of calculating capacity costs similar to the Duke
Energy method would more accurately value the time difference of capacity, it will still be
necessary to address the additional ancillary services provided by battery storage QFs in further
docket or dockets.
STAFF COMMENTS t4 AUGUST 27,2020
c. Appropriateness of the Published Rate Method.
The Commission currently applies the Published Rate Method to all QFs below their
respective resource-type project eligibility caps.
i. Avoided energy value using the Published Rate Method.
The method for determining the avoided cost of energy for published rates is based on
three things: (1) the actual amount of QF output during high and low load hours and in different
seasons; (2) the forecasted cost of natural gas; and (3) the heat rate of a CCCT surrogate natural
gas plant. Because none of the characteristics associated with the type of QF is taken into
account, the Published Rate Method energy avoided cost rate is agnostic to the type of QF.
Since it can be generically applied to all QF types, the Published Rate Method can be equally
applied to battery storage QFs. However, the Published Rate Method assumes that a CCCT is
the marginal resource at all times, which Staff maintains is a flawed assumption.
ln Response to Staff Production Request No. 15, the Company calculated that the actual
percentage of time the Langley Gulch CCCT plant was the marginal resource in 2018 and2019
was 10.2 Vo andT.3Vo,respectively. Because the ICIRP Method uses a modeled result, Staff also
requested the percentage of time the ICIRP model reflects Langley Gulch as the marginal
resource over a two-year timeframe. In the Company's Response to Staff Production Request
No. 16, the Company stated that the ICIRP model run over this period showed that Langley
Gulch was the marginal resource 9.3Vo of the time. The similarity between these two results
confirms that the ICIRP Method more closely reflects reality at least regarding how often a
CCCT plant is the marginal resource. Therefore, Staff believes that using a CCCT as the
marginal resource L00Vo of the time assumed in the Published Rate Method is highly dubious.
Despite its flawed assumption that a CCCT is the marginal resource lO07o of the time on
the Company's system, Staff believes the Published Rate Method is still appropriate to determine
the avoided energy value for battery storage QFs under the project eligibility cap. Doing so
aligns with the Commission's treatment of other resources that use the Published Rate Method
for published avoided cost rates. Staff believes that examining whether to move away from the
use of a surrogate resource for published avoided cost energy rates for battery storage QFs
should be examined in the next docket for all Idaho electric utilities.
STAI]F COMMENTS 15 AUGUST 27 ,2020
ii. Avoided capacity value using the Published Rate Method.
Staff does not believe that the current Published Rate Method for establishing avoided
capacity rates for battery storage is appropriate and should belong in its own category under
published rates because it will require different treatment.
The Published Rate Method to calculate avoided capacity costs takes into account three
factors:(l) theinflationrateandrateofreturnadjustedfixedcostofaCCCT;(2)thecapacity
contribution at peak of resources included in the "other" category to determine how much of the
fixed cost of a CCCT will be avoided on a $ per kW basis; and (3) the annual capacity factor
based on historical generation of resources included in the "other" category which is used to
determine a $/kWh rate. In addition, capacity payments do not begin until the first capacity
deficiency date passed that is established at the time of establishing the QF's first contract.
Staff identified two key reasons why battery storage needs to belong in a separate
category based on differences in the output characteristics that determine avoided capacity cost
compared to other resource types in the "other" category. First, resources classified in the
"other" category do not have the same limitations with respect to the duration of output which is
limited for battery storage QFs. Because of limited duration, the output of an energy storage QF
will likely be manipulated to maximize revenue for the developer without proper price signals to
provide output during peak periods to avoid capacity cost. Second, battery storage suffers
degradation and will likely not provide the same amount of capacity when the utility becomes
capacity deficient compared to when batteries are new. Certain types of battery storage degrade
faster with deep-cycling, which will likely occur if developers offset generation to maximize
output during high load hours. Both of these reasons have implications from an accountability
standpoint if the utility and customers are to receive the amount of avoided capacity benefit in
return for the amount of compensation paid to the developer. To receive the amount of avoided
capacity, batteries need to be replaced on a regular basis or adjustments made to capacity factors
and the capacity contribution at peak depending on how often the battery storage QF has been
duty-cycled over the duration of a contract. Even if batteries are maintained to provide the same
amount of capacity when new, the release of energy from the QF needs to be managed to ensure
it provides capacity benefit during system peak periods. Public comments support Staff s
contentions with degradation of IUVo in the first three years of a battery's life and with deep
discharge, the amount of charge and discharge cycles can be reduced to one third of Li-Ion
battery's normal life.
STAI]F COMMENTS t6 AUGUST 27 ,2020
To address issues identified above related to accountability and assurance that the eF
provides avoided capacity cost benefits commensurate with its capacity payments, Staff proposes
the Company develop an Idaho Power version of the Duke Energy method similar to Staff's
recommendation to update the ICIRP Method for calculating avoided capacity costs. The
method would provide the proper price signals to ensure that QFs release energy from their
battery storage project during peak hours. Ifa QF releases energy outside ofthe hours
designated as peak periods, it would not receive any compensation for avoided capacity cost.
The method would still withhold capacity payments until the capacity deficiency date established
in the QF's initial contract.
In addition, Staff maintains the CCCT is not a reasonable surrogate because it is not used
primarily used as a capacity resource. In revising the method, Staff advocates revising the
capacity surrogate from a CCCT to a SCCT to match what is used in the ICIRp Method.
Staff believes that by establishing and publishing the capacity rates and applicable times
in advance, and allowing a QF to lock-in the capacity deficit date, including applicable rates and
times, upon executing its first contract or establishing a legally enforceable obligation with the
utility, that the Commission would comply with FERC rules implementing PURpA and statutory
requirements to provide published avoided cost rates known at the time of contracting, and
would align with the treatment of other QFs.
v. Appropriate Published Rate Elieibilitv Can for Enerev storase.
Staff evaluated the four factors listed below to determine if the Company's proposed
100kW published rate eligibility cap for energy storage QFs is appropriate or if it should be set
higher.
a. The incentive for projects over the eligibility cap to disaggregate to qualify for
Published Rate Method rates because of differences between Published Rate Method
rates and ICIRP Method rates;
b. The minimum module size of battery storage projects that might allow developers to
disaggregate large projects into smaller projects;
c. The difference in the amount of time, effort, and money required to develop an ICIRp
Method contract versus a Published Rate Method contract that can act as a
disincentive for legitimately smaller projects to be developed; and
STAFF COMMENTS 17 AUGUST 27 ,2020
d. The impact on customers and QF developers due to an increase in the number of
projects required to use more accurate ICIRP Method rates from a lower eligibility
cap.
After consideration of all four factors, Staff believes that until the Commission can
address incentives for large battery storage QFs to disaggregate into smaller QFs, and because
battery storage QFs, can be easily disaggregated due to the minimum size of a module of
capacity, the eligibility cap should be set at a size that prevents disaggregation. This would
likely prevent battery storage QFs from qualifying for less accurate and higher avoided cost rates
that would be unfair to customers and prevent similar disaggregation that occurred with large
wind projects prior to the Commission setting a 100 kW eligibility cap on December 14,2010.
In addition, Staff determined that the potential dampening impacts of contracting
differences between the Published Rate Method and the ICIRP Method for projects under
10 aMW was inconclusive. By requiring more projects to use the more accurate ICIRP Method,
the Commission will continue to ensure that customers are indifferent to QF purchases because
they will be paying no more and no less than the Company's marginal cost of energy. See 18
c.F.R. $ 2e2s0a@)Q).
a. Published Rate Method rates are higher than ICIRP Method rates, which creates an
economic incentive for projects to disaggregate.
Staff believes that the current avoided cost framework, using Published Rate Method
rates for published avoided cost rates, and ICIRP Method rates for QFs above the project
eligibility cap, creates an inherent incentive for QFs to disaggregate to receive the higher
Published Rate Method rates. As stated earlier, Staff developed "reference" projects for each of
the different QF types and then compared the resulting rates between the two established
methods.
Table No.5: Published Rate Method and ICIRP Method Avoided Cost Rates.
STAFF COMMENTS 18 AUGUST 27 ,2020
Reference Project Avoided Cost Differences
tevelized over 20 yearc without integration charges
Rates authorized December 2019
Negotiated Rate
ICIRP Method (S/MWh)
(S/MWh)
lPublished Rate
lsurrogate Method
Itstrurwt'r
Differe nce
(s/MWht % Difference
Wind (10 MW nameplate)
Solar (10 MW nameplate)
Non-seasonal Hydro (10 MW nameplate)
Seasonal Hydro (10 MW nameplate)
Other (10 MW nameplate)
SEa.ss
s4s.20
*2.74
Sqs.a+
s39.87
iqz.qs
Ss9.2s
5s7.62
Sss.oe
ss3.18
53.u
Sro.os
Sr+.s8
$2s.72
s13.31
8.8%
20.4%
34.8%
59.4%
33.4%
As shown in Table No. 5 above, overall published rates are approximately 8.8Vo higher
for wind, 17 .l7o for solar, and almost 60Vo higher for seasonal hydro. These rates are a
composite of a separate rate for the avoided cost of energy and the avoided cost of capacity.
Staff believes it may be appropriate for the Commission, in the next docket, to look at
using a modified version of the ICIRP Method to set published rates as a way to collapse the
distinction between projects above and below the project eligibility cap, use the same input
assumptions, and to encourage the use of more accurate avoided cost rates for all project types.
b. Battery storage QF technology allows for potential disaggregation.
Staff believes that battery storage QFs have a minimum capacity module size that allow
QF developers to disaggregate their projects to qualify for higher published rates. The ability to
disaggregate was the main factor the Commission used to justify setting the eligibility cap at the
minimum FERC cap of l00kw for wind and solar. In Order No. 32176, the Commission said,
A 100 MW wind farm or solar project can be broken up into 10
aMW pieces in order to obtain multiple published rate contracts, i.e.,
disaggregation. When a 100 MW wind or solar project is
disaggregated, we find the SAR Methodology no longer produces a
rate that accurately reflects the value of the energy to the utility. A
100 NfW project is not even eligible under PURPA nor is a utility
bound to purchase power from a 100 NfW facility under PURPA's
"must purchase" provision. 18 C.F.R. $ 292.204(a). Therefore, to
prevent large projects from disaggregating in order to not only
become eligible under PURPA but also obtain published avoided
rates, and based on the unique characteristics of wind and solar
resources to disaggregate, we find that the eligibility cap for
published avoided cost rate contracts for wind and solar projects
shall be set at 100 kW or less.
Order No. 32176 at 13.
To determine if energy storage projects fall under the same criteria, Staff evaluated
energy storage on whether or not it can be disaggregated_into smaller projects by evaluating the
same type of characteristics that allow wind and solar to be easily disaggregated. To prevent
disaggregation using the eligibility cap, it is first necessary to determine the minimum size of a
module of battery storage. The cap would then need to be set lower than the minimum size of a
module of battery storage.
From a QF developer's standpoint, if the incremental cost of separating modules of
capacity of a large project into several smaller projects is less than the incremental revenue that
STAFF COMMENTS 19 AUGUST 27 ,2020
can be eamed through higher published rates, the QF developer may choose to disaggregate. To
make this evaluation, the developer would need to account for the incremental cost of separate
contracts, additional meters, additional transformers, and other interconnection costs associated
with establishing separate projects and weigh those costs against the economies of scale
associated with different module sizes for their project.
Although there are several different battery types used in utility-scale energy storage,
Li-Ion batteries have become predominant in the industry. A Li-Ion storage cell necessary to
store energy is only slightly larger than an AA-battery and is rated at less than 5 volts. To
achieve the amount of energy and capacity needed to be useful to a utility, thousands of these
batteries are combined in different combinations and configurations to achieve a desired result.
Because the energy stored in batteries is stored as direct current, to be fed into the
utility's grid, it must be converted into alternating current through an inverter and stepped up in
voltage to match the voltage rating of the utility interconnection. Staff reviewed several different
module sizes and standard configurations offered by utility-scale battery storage solution
providers. Because a module of battery storage is upwardly scalable from the size of a single
battery, and because the size of inverters for utility application are more limited, Staff based the
size range of a module of capacity based on inverters, which range from about 70 kW to 4 MW.
As discussed above, a battery storage QF developer would need to examine the cost tradeoffs of
earning more revenue through higher published rates against the incremental cost associated with
creating multiple projects. A 100 kW published rate eligibility cap would increase these
incremental costs making disaggregating a large facility into smaller projects less economically
viable, and would protect customers from paying potentially higher, less accurate published
rates. Public input concurred with Staff's assessment that battery storage QFs are modular and
limited to the inverter sizes.
Public comments from Clenera,LLC verified the range of Staff s inverter size and that
Li-Ion batteries are currently the predominate technology. Comments from Renewable
Northwest verified that battery storage is modular and flexible making them comparatively easier
to disaggregate and that disaggregation is limited by the inverter capacity of the battery storage
QF.
STAFF COMMENTS 20 AUGUST 27 ,2020
c. Staff reviewed the impact of a previous project eligibility cap decrease.
Staff did not find strong evidence that more onerous requirements to develop an ICIRP
Method contract discourage QFs between 100kW and lOaMW from being developed. Staff did
not see a significant decrease in the number of projects in this size range that were developed
after the Commission set the eligibility cap to 100kW for wind and solar. In Order No.32262,
the Commission acknowledged that setting the eligibility cap to 100kW for wind and solar might
have an impact on small wind and solar QFs. To understand the potential impact on battery
storage QF development by setting an eligibility cap at the 100kW minimum versus 10 aMW,
Staff analyzed: (1) the differences in process requirements between the two methods to establish
a contract; and (2) the change in the number of wind and solar projects that were developed as a
result of lowering the eligibility cap to 100 kW.
Staff analyzedthe process QF developers are required to use for negotiated rate contracts
versus published rate contracts. Both types of contracts use ScheduleT3, which outline the steps
needed to establish a contract. Staff identified two differences that could make the process more
difficult and impose barriers to development of QFs above the project eligibility cap. First, the
Company is required to provide indicative pricing within 10 days of published rate contracts and
20 days for QFs above the project eligibility cap. Since published rates are readily available
online, ICIRP Method contracts may experience up to a20-day delay. Second, projects above
the project eligibility cap require an accurate generation profile used to determine contract rates,
while published rate contracts use rates that are generic for a given QF type. The development of
a generation profile for projects above the project eligibility cap require the developer to conduct
a study, which adds additional time and expense not necessarily incurred by a developer who
qualifies for published rates.
To determine the effect on project development of lowering the eligibility cap to 100 kW
for projects that were under 10 aMW, Staff requested a list of all wind and solar projects
approved between February 20,2008 and December 14,2010 when the eligibility cap was set to
10 aMW through Order No. 30488, as well as projects approved on or after December 14,2010
when the eligibility cap for wind and solar was set at 100 kW through Order No. 32176. After
adjusting for projects that were disaggregated, and recognizing the limited amount of data, the
analysis suggests that that lowering the eligibility cap to 100 kW likely did not impact the
amount of QF development for wind and solar projects under 10 aMW.
STAFF COMMENTS 2t AUGUST 27 ,2020
d. Impacts to customers and QF developers.
Staff considered the potential impact on customers and QF developers due to more
projects using more accurate ICIRP Method rates that would result from a lower eligibility cap.
As discussed earlier, Staff concluded that the ICIRP Method is likely to produce a more accurate
rate than the Published Rate Method. If the Commission sets a project eligibility cap close to or
at the FERC 100 kW minimum rather than at l0alvtw or higher, the lower eligibility cap would
increase the proportion of projects using more accurate rates.
Although allowing developers to qualify for higher and less accurate avoided cost rates
through a higher eligibility cap would allow more energy storage QF projects to be economically
viable to developers, the Commission has been clear on the issue of accuracy of rates and its
impact on retail customers as well as developers. In Order No.32262, the Commission stated,
"Rates should not be set to allow QFs to be cost effective. PURPA entitles QFs to arate
equivalent to the utility's avoided cost, a rate that holds utility customers harmless - not arate
at which a project may be viable. 18 C.F.R. 5 292.3O4(a)(2)." Order No.32262 at8.
In addition, the Commission also stated in Order No.32262 when reducing the eligibility
cap to 100kW:
While we recognizethe impact that this decision will have on small
wind and solar projects, it would be erroneous, and illegal pursuant
to PURPA, for this Commission to allow large projects to obtain a
rate that is not an accurate reflection of the utility's avoided cost for
the purchase of the QF generation." Id, citing Rosebud Enterprises
v.Idaho PUC, 128Idaho 609,623,917 P.2d766,780 (1996), citing
Connecticut Light & Power Co.,70 FERC 6I,012 (1995), reconsid.
denied,71 FERC 61,035 (1995).
Public input received generally advocates for a higher project eligibility cap. Renewable
Northwest details technological and economic reasons why a developer would not disaggregate
their project to meet a lower project eligibility cap and Clenera states that a project eligibility cap
of 10 to 20 MW would achieve meaningful cost savings through integration of components.
VI. Contract length analvsis.
As described above, Staff believes that the ICIRP Method more accurately values
avoided cost rates. Neither the Company nor any of the three groups who responded to Staff s
Request for Public Input disputed this point. If battery storage QFs are assigned avoided costs
STAITF COMMENTS 22 AUGUST 2] ,2020
rates under the ICIRP Method, Staff believes that the more accurate pricing method - which
more closely aligns with how the Company values all other resources - mitigates concerns that
contract lengths exceeding two years harm customers.
All three groups who responded to Staff's Request for Public Input confirmed that a
payback period is critically important to secure project financing. No new QFs have signed a
two-year contract since the Commission made the change to two-year contracts in 2015.
Therefore, Staff believes the payback period is a very important aspect to consider when
determining contract length for battery storage QFs receiving ICIRP Method rates. In addition to
considering the payback period, Staff also evaluated the expected life of battery storage QFs and
the contract terms in surrounding states as points of reference. Based on consideration of all
these factors, a contract term length between 10 and 15 years is reasonable. Staff recommends
the Commission establish a lO-year contract term for battery storage QFs above the project
eligibility cap.
a. Results of quantitative analysis of battery storage QF costs and expected return.
Staff worked with PNNL to quantitatively analyze contract lengths for battery QF
projects. This analysis took the costs for a range of battery storage projects and optimized the
generation profile of the project to maximize revenue under a simplified version of the
Company's ICIRP Method. The revenue produced by optimizing the generation profile was then
compared against the total cost of the project to determine a reasonable payback period.
PNNL analyzed Lithium-ion battery QF projects of 20 MW, 50 MW, and 80 MW
powered by fixed tilt and single-axis solar photovoltaics. Battery cost assumptions were taken
from the "Energy Storage Technology and Cost Characterization Report" prepared by multiple
national laboratories. Solar PV cost assumptions were taken from U.S. Solar Photovoltaic
System Cost Benchmark: Q1 2018 prepared by the National Renewable Energy Laboratory.
Costs were estimated at $1130/kW DC for the solar PV component and $1,4461kW for the
battery component. Fixed O&M costs were estimated at $10.40/kW-yr for PV and $9lkW-yr for
the Lithium-ion battery. PNNL then optimizedthe generation profiles of the reference projects
to maximize revenue using the Company's solar monthly ICIRP Method rates and calculated that
the project would not pay back in a meaningful time frame.
PNNL noted that a lithium-ion battery has an 86Vo rowdtrip efficiency, which may be a
disincentive to developers to utilize the technology, unless there are rates that are
STAFF COMMENTS 23 AUGUST 27 ,2020
correspondingly higher. Current ICIRP Method rates are not sufficiently time differentiated to
take advantage of the benefits that battery storage can provide.
While this is a reasonable approach to evaluating the payback period, Staff cautions that
the availability of data to PNNL makes this analysis a rough-cut approach and several
considerations should be made clear. Staff clarifies that these considerations are limited to the
payback period analysis - they do not impact the calculation of QF avoided cost rates that Staff
recommends be addressed in the follow-on docket.
The first consideration regarding the payback analysis is that because the Company does
not yet have battery storage avoided cost rates, PNNL used the Company's monthly solar ICIRP
Method rates to determine the battery QF's revenue stream. However, this approach likely
decreased revenue for the battery QF project because solar QFs cannot be dispatched to align
with the Company's peak and therefore does not capture the full value.
A second consideration is that PNNL only had access to monthly, rather than hourly or
sub-hourly, avoided cost data from the Company for its analysis. With more granular avoided
cost data, PNNL could more closely match the generation profile of the battery QF project to the
Company's system needs, which would likely increase revenue for the project. Staff believes it
would be beneficial for PNNL, through Staff, to provide the Company with a variety of battery
storage QF generation profiles to be to modeled in the ICIRP Method to more accurately
determine the revenue stream of these projects.
The third consideration is that because the Company does not yet have battery storage
avoided cost rates, there is no value in these avoided rates to reflect the ancillary services these
projects may be able to provide. Staff notes that all three considerations listed here served to
decrease the hypothetical revenue stream calculated by PNNL.
Staff also notes that while PURPA requires the Commission to encourage the
development of QFs, PURPA does not require the Commission to make economically infeasible
projects feasible, either through rates or contract terms. Instead, avoided cost rates are to be set
at the Company's marginal resource value, as described above, and contract terms "should be
long enough to allow QFs reasonable opportunities to attract capital from potential investors."
Windham Solar LLC and Allco Finance Limited,l57 FERC [161134 (2016).
STAFF COMMENTS 24 AUGUST 27 ,2020
b. Expected Life of Battery Storage QFs.
Another important aspect Staff considered in its contract length analysis is the
the expected life of a battery storage system. Staff's research and work with PNNL indicates
that this is approximately 10 years, although it can be longer or shorter depending on how
frequently the battery storage project is cycled. Renewable Northwest stated battery storage
projects have a calendar life (i.e. inactive or under minimal use) of 11- 15 years that decreases as
the battery is cycled. Clenera stated the expected useful life is 20 - 25 years before accounting
for degradation.
Although the life of the original battery cells may be limited to l0 - 15 years, Staff
understands that some battery storage PPAs now include capacity refills. This means that when
a developer signs a.PPA with a utility, the developer commits to producing a certain amount of
energy for the entire duration of the contract (i.e.2O years). As the battery performance
degrades, the developer simply replaces or adds additional battery cells to the original project to
meet the generation production required by the contract. The developer accomplishes this by
building extra rack space into the original project - which remains empty in the early years of the
project - to hold the additional battery cells that will supplement, or refill, the original battery
capacity as it degrades in later years. Maintaining battery storage system production with
capacity refills is about 2O - 30Vo of the original project cost, which makes it a much less
expensive alternative than building an entirely new battery storage project at the end of the
original project's useful life. Staff believes that the ability of a developer to extend the life of the
project may also assist with the QF's ability to secure financing.
c. Contract Lengths in Surrounding States.
As discussed in Staff s Initial Comments, Staff reviewed QF contract lengths in several
surrounding states to provide context for a similar decision in this case. Staff s research found
that surrounding states have contract lengths of 10, 12,15, and20 years for various technologies
and that those contract lengths were chosen because they provided an adequate opportunity for
QFs to attract financing. This aligns with the recommendation from Renewable Northwest of 10
to 15 years and Clenera's recommendation for 15 to 20 years. ICL stated that it does not know
the correct contract length, but that five years is the minimum that should be considered.
STAFF COMMENTS 25 AUGUST 27 ,2020
d. Staff conclusion on contract lengths.
Staff's quantitative analysis on contract lengths indicates that no duration of contract
length will be sufficient to attract financing for the majority of battery storage QFs under a
constrained view of the current ICIRP Method rates. However, Staff believes by feeding an
actual battery storage QF generation profile into the ICIRP model, taking further steps to
improve the granularity of the ICIRP model, and accounting for the possibility of battery storage
QFs to extend the lives of their QFs, a 10 to 15 year contract length would be sufficient to allow
QFs a reasonable opportunity to attract financing while maintaining ratepayer indifference.
Staff notes that this aligns with the shorter end of the contract length spectrum offered in
surrounding states.
STAFF RECOMMENDATIONS
o Staff recommends the Commission establish a 100 kW project eligibility cap for
battery storage QFs in order to ensure that as many battery storage QFs as possible
use the more accurate ICIRP Method of calculating avoided cost rates.
o Staff recommends a l0-year contract term length for battery storage QFs above the
project eligibility cap, and continue to give QFs below the project eligibility cap 20-
year contracts.
o Staff recommends the Commission order the Company to make a compliance filing
within 30 days detailing how it will implement an Idaho Power specific version of the
Duke Energy Method for calculating capacity payments for battery storage QFs
above and below the project eligibility cap, which compensates capacity only for
energy actually delivered during specific peak hours.
o Staff recommends the Commission direct Avista, Rocky Mountain Power, and Idaho
Power to file a battery storage QF case to:
a. align battery storage QF treatment across Idaho's major electric utilities;
b. modify the ICIRP Method to provide greater time-differentiated granularity of
avoided energy costs;
c. determine compensation of ancillary services, and
d. refine implementation of Idaho-specific capacity costs similar to the Duke
Energy Method.
STAFF COMMENTS 26 AUGUST 2] ,2020
Respectfully submitted this 27h day of August 2020.
F> 1.*t{
Edward l.{ry}tt
Deputy Attorney General
i :umisc/commentslipc.e20.2ejrfyytnc conrments
STAFF COMMENTS 27 AUGUST 27,2O2O
CERTIFICATE OF SERVICE
I HEREBY CERTIFY THAT I HAVE THIS 27th DAY OF AUGUST 2020,
SERVED THE FOREGOING REVISED COMMBNTS OF THE COMMISSION
STAFF, IN CASE NO. IPC-E-20-02, BY E-MAILING A COpy THEREOF, TO T[rE
FOLLOWING:
DONOVAN E WALKER
REGULATORY DOCKETS
IDAHO POWER COMPANY
PO BOX 70
BOrSE rD 83707-0070
E-MAIL: [e-mail]
dockets @ idahopower.com
S Y
CERTIFICATE OF SERVICE