2018 Varible Energy Resource Integration Analysis Report.pdf

38ffi*.
REC T IVE D
2OIB JUL 3I PH L: 32
An IDACORP Company
DONOVAN E. WALKER
Lead Gounsel
[e-mail] C
SSION
July 31 ,2018
VIA HAND DELIVERY
Diane M. Hanian, Secretary
ldaho Public Utilities Commission
47 2 W est Wash ington Street
Boise, ldaho 83702
ldaho Power Company's 2018 Variable Energy Resource lntegration
Analysis Report
Dear Ms. Hanian
Enclosed please find four (4) copies of ldaho Power Company's ("!daho Power")
2018 Variable Energy Resource ("VER') lntegration Analysis Report. This report was
conducted and filed in compliance with the Public Utility Commission of Oregon's Order
Nos. 17-075 and 17-223 from Case No. UM 1793. This report is being filed with the
Public Utility Commission of Oregon on today's date. Michael Eldred, Mike Louis, and
Yao Yin from the ldaho Public Utilities Commission ("Commission") Staff participated in
the Technical Review Committee that assisted in the study analysis contained in this
report.
ldaho Power is providing these informational copies for the Commission's
records and convenience. Please contact me at [phone] if you have any
questions.
Donovan E. Wal
DEW:csb
Enclosurescc: Karl T. Klein, Commission - w/encl.
Terri Carlock, Commission - w/encl.
Michael Darrington, ldaho Power - w/o encl. (via e-mail)
1221 W. ldaho 5t. (83702)
PO. Box 70
Boise, lD 83707
Re
Si
C
RA O IIVED sIm.Pll L: 33
An IDACORP Company
",'l L)
Variable Energy Resource (VER)
lntegration Analysis
July 2018
O 20l8Idaho Power
ldaho Power Company VER lntegration Analysis
TagLe oF CoNTENTS
Table of Contents
List of Tables ......
List of Figures.....
l. Introduction...
2. Technical Review Committee.
3. 2018 Wind Integration Study.........
3.1. Background ..
3.1.1. Wind and ldaho Power's System.....
3.1.2. Dispatchable Generating Capacity...
3.1 .3. 2017 Operations Issues
3.1.4. Issues Not Addressed by the Study...
................. I
................ ii
............... iv
.................1
2
2
2
5
6
6
8
3.1.4.1. Day-Ahead Uncertainty ............. ............8
3.1.4.2. Cycling Costs (Variable Operation and Maintenance Costs).........................8
3.1.4.3. Sub-Hourly Costs of Responding to Variability............... ...........9
3.1.4.4. Reserve Violation Impacts on Integration Costs..... .....................9
3.2. Study Design ......................9
3.3. Regulating Reserve Calculations and Other Operating Reserves.... ............10
3.3.1. Area Control Error .....................10
3.3.2. NERC BAL Standard
3.3.3. Estimation of RegUp/RegDn for Wind.......... ..................11
3.3.4. Estimation of RegUp/RegDn for Load.......... ..................13
3.3.5. Estimation of RegUp/RegDn for Load Netted with Wind. .................15
3.3.5.1 . Diversity Benefit...... .........15
3.3.5.2. Contingency Reserve ............... I 5
3.3.5.3. Estimation of RegUp/RegDn for Alternative Wind Buildouts .....................16
...................1 83.4. System Modeling
Page i
I
VER lntegration Analysis ldaho Power Company
3.5. Modeling Results
3.5.1. Cost Results for Simulation at Current Wind Buildout
3.5.2. Simulated Dispatch of Reserve-Providing Resources .,
3.5.3. Cost Results for Simulations at Alternative Wind Buildouts.............
3.5.4. Incremental Integration Costs.........
3.5.5. Hydro Condition Sensitivity Analysis ........
3.5.6. Regulating Reserve Violations
4. Energy Imbalance Market and VER Integration
5. Unified Wind and Solar Integration Costs
6. System Limits and Maximum VER Buildout....
7. Conclusions.................
Lrsr oF TABLES
Table 1
2007 WIS results using historical Mid-C prices as benchmark
Table 2
20 I 3 WIS integration costs ($/IVIWh) ...............
Table 3
2018 WIS results
Table 4
Number of reserve violations..
Table 5
RegUp and RegDn percentages for wind reserves based on 2HA wind forecast...........
Table 6
Winter, spring, fall
Table 7
Summer
Table 8
Derived RegUp and RegDn percentages for BA load reserves based on 2HA load
forecast.....
l8
r8
l9
...,...,.,.....3
...............3
.23
.24
.26
.26
.28
.29
.34
.35
3
l3
,4
t4
Page ii
Table 12
Total output by fuel type.............,,21
Table 9
Allocation factors for netted load and wind
Table 10
Increase in standard deviation...
Table l1
Estimated integration costs for the current wind buildout...........
Table 13
Estimated production costs for alternative wind buildouts....
Table 16
Hydro condition sensitivity analysis results........
Table 17
Number of reserve violations, load net wind scenario......
Table 18
Total MWh of violations, load net wind scenario
Table 19
Max MW of violations, load net wind scenario
Table 20
Monthly standard deviation of l0-minute changes, load alone time series, and load net
wind and solar time series
Table 21
Integration cost comparison of 727 MW wind, 1,000 MW of wind, and 727 MW wind
plus 289 MW solar
Table22
AURORA reserve violations count by scenario
Table 23
AURORA reserve violations maximum MW by scenario
Table24
Future integration cost recommendation for incremental VER project additions
.......... I 5
t7
....1 9
....23
Table 14
Incremental integration cost for 727 MW to 800 MW of nameplate wind ,,,,,,...24
Table 15
Summary integration costs and incremental integration costs per MWh with reserve
violations.. .................25
26
.27
.27
.....29
.32
.27
.34
.33
.36
ldaho Power Company VER I ntegration Analysis
Page iii
VER lntegration Analysis ldaho Power Company
LIST OF FIGURES
Figure 1
Wind resources on Idaho Power's system
Figure 2
Load and net load after VERs
5
7
Figure 3
Twenty-minute ramping of 2HA forecast BA load ll
Figure 4
Standard deviation of the 10-minute time-step wind production data for summer 2017 ...........16
Figure 5
Load alone-generation from units providing reserves..... ...................20
Figure 6
Load net wind-generation from units providing reserves ..................20
Figure 7
Load alone-RegUp
Figure 8
Load net wind-RegUp ..............
.21
.22
Figure 9
Load alone-RegDn ....22
Figure 10
Load net wind-RegDn '.,,23
Figure 11
Histograms of l0-minute changes for March 201 8, load alone time series and load net
wind and solar time series .............30
Figure 12
Monthly contributions of load, wind, and solar to the standard deviation of l0-minute
time series ..3 I
Page iv
ldaho Power Company VER lntegration Analysis
1. lNrnoDucnoN
This report summarizes the actions taken in compliance with the Public Utility Commission of
Oregon's (OPUC) Order Nos. I 7-075 and 17-223 from Case No. UM 1793. The OPUC's final
orders from UM 1793 adopted Idaho Power's 2016 Solar Integration Study and approved the
implementation of solar integration charges based on that study. The OPUC also directed
Idaho Power to work with a Technical Review Committee (TRC), similar to what was done with
the 2016 Solar Integration Study, to conduct a new wind integration study, evaluate potential
impacts of participation in the Energy Imbalance Market (EIM) on integration costs,
and evaluate whether to conduct a joint wind and solar integration cost study. In clarifying
its direction to Idaho Power, the OPUC stated:
At page 7 of our order, [Order No. l7-075] we affirmed our intent that integration
studies, as well as the additional factor of EIM participation, should be addressed
in the annual IRP update and IRP acknowledgement processes. We therefore order
Idaho Power, as soon as the 2017 IRP had been filed, to work with the TRC to
conduct a new wind integration study, perform an analysis of the impact of
participation in the EIM and thoroughly evaluate whether to conduct a joint wind
and solar integration cost study. We also ordered the company to, as part of this
assessment, examine different methods for allocating jointly determined costs
between wind and solar and to submit a study report and recommendation to us no
later than April 30, 2018, well ahead of the beginning of the 2019 IRP. Order No.
17-223. [The April 30,2018, deadline was extended to July 3 l, 201 8.] I
Idaho Power initiated the study process by organizing a TRC as summarized below. Idaho Power
held regular meetings and communications with the TRC to receive and incorporate feedback
throughout the process. A comprehensive wind integration study was conducted, the results of
which are set fomh in this report. Additionally, Idaho Power examined a combined integration
approach for both wind and solar, as well as conducting some initial evaluation of the differences
participation in the EIM may make on such determinations.
This report concludes that the varied analyses of wind, solar, load, EIM, and reserves indicate a
unified variable energy resources (VER) integration analysis approach may be the best way to
assess costs for additional increments of variable and intermittent generation resources, like wind
and solar, going forward from current levels of penetration. However, the analysis also indicates
Idaho Power's system is nearing a point where the current configuration can no longer integrate
additional VERs. Additional investigation is warranted into the combined effect of wind and
solar, in a unified VER integration analysis and cost impact determination, along with the
potential effects of participation in the EIM and its unique requirements, attributes, costs,
and benefits.
Page 1
I Order No. l8-130.
VER lntegration Analysis ldaho Power Company
2. TecnNrcAL Reuew ColtturrrEE
Idaho Power greatly appreciates the involvement of the TRC members:
. MichaelEldred, Mike Louis, and Yao Yin-Idaho Public Utilities Commission (IPUC)
o Kurt Myers-Idaho National Laboratory
o Ben Kujala-Northwest Power and Conservation Council
o CameronYourkowski-RenewableNorthwest
o Brian Johnson Ph.D., P.E.-University of Idaho
o Brittany Andrus and Jean-Pierre Batmale-OPUC
The TRC provided important guidance in the design and vetting of the approach Idaho Power
adopted in re-evaluating the cost of integrating wind and solar resources. Specifically,
TRC discussions led to Idaho Power adopting the North American Electric Reliability
Corporation (NERC) Real Power Balancing Control Performance standard (NERC BAL
standard)2 for defining the reserves; sharing the diversity benefits ofreduced reserves across load
and wind and solar generation; and using incremental standard deviation methods to calculate the
contributions of load, wind, and solar to their netted variability. These three study improvements
are foundational to the fair treatment of variability in a generation resource on Idaho Power's
system.
3.2018 Wrno lrurecmnoN SrUDY
3.1. Background
The 2018 Wind Integration Study (WIS) is the third Idaho Power study evaluating the impact of
increased variability on the cost of power supply operations. The first two studies were
completed in 2007 and 2013. The 2007 WIS evaluated three hydro condition years and four wind
levels using meteorological simulations of 300, 600, 900, and 1,200 megawatts (MW).
The matrix of the results from the 2007 WIS is shown in Table l.
2 Standard BAL-001-2-Real Power Balancing Control Performance:
nerc.conr/palStand/Rel i abi I ity%20Standards/BAL-00 I -2.pdf.
Page 2
ldaho Power Company VER lntegration Analysis
Table 1
2007 WIS results using historical Mid-C prices as benchmark
Study Year Penetration Level (MW) Cost Per MWh* Wind
1 998
1 998
1 998
1 998
2000
2000
2000
2000
2005
2005
2005
2005
300
600
900
1,200
300
600
900
1,200
300
600
900
'1,200
$3.1 9
$4.73
$6.06
$6.92
$21.89
$30.30
$39.06
$39.40
$10.69
$9.32
$10.58
$8.12
"Megawaft-hour
The second WIS was completed in 2013 and evaluated integration costs at wind levels of 800,
1,000, and 1,200 MW using three hydro condition years. Idaho Power had 678 MW of wind
generation on-line as of January 2013. The results from the 2013 WIS are shown in Table 2.
Table 2
2013 WIS integration costs ($/MWh)
Water Condition 800 Mw 1,000 Mw 1,200 Mw
Average (2009)
Low (2004)
High (2006)
Average
$7.1 8
$7.26
$9.73
$8.06
$11.94
$12.44
$14.79
$r3.06
$1 8.1 5
$1 8.1 5
$20.73
$19.01
The 201 8 WIS evaluates integration costs at wind levels of 300, 500, 727,800, 900, I ,000,
and 1,100 MW using a median, low, and high hydro forecast. Actual Idaho Power system wind
production data from 727 MW of nameplate capacity was used to develop the 300-, 500-, 800-,
900-, 1,000-, and 1,100-MW levels. The results from the 2018 WIS are shown in tables 3 and 4.
Table 3 provides the integration charges for varying levels of nameplate wind capacity. As noted
in Table 3, increasing levels of wind capacity result in times where the model of Idaho Power's
system cannot meet its reserve obligations, which would indicate times when generation
curtailment would likely be required to meet compliance. Table 4 indicates the number of
reserye violations for varying levels of wind capacity.
Table 3
2018 WIS results
Wind Nameplate (MW)
lntegration Charge ($/MWhf
300
$2.29
500
$2.88
727
$4.52
800
$4.88
900
$5.56
1,000
$5.96
1,100
$5.1 7
*Costs included in the lntegralion Charge do not include mitigation for periods when the requested operating reserves vvere unable
to be provided by the model.
Page 3
Table 4
Number of reserve violations
VER lntegration Analysis ldaho Power Company
Nameplate
(Mw)
Regulation Up
(RegUp)
Total Reserve
Violations
Percent of
Hours
Regulation
Down (RegDn) Spin NonSpin
300
s00
727
800
900
1,000
1 ,'100
1
7
52
133
522
988
1,736
3
14
76
232
799
1,434
2,434
<0.1o/o
0.2o/o
0.9o/o
2.6Yo
9.1o/o
16.4o/o
27.8o/o
2
1
23
91
255
435
690
6
1
I
22
11
8
The 2018 WIS was initiated in compliance with OPUC Order No. l7-075. In Order No. l7-075,
the OPUC allowed Idaho Power to adopt their 2016 Solar Integration Study and directed the
company to file amendments to Schedule 85 setting forth the incremental costs of integrating
solar and wind generation into its operations. The order further directed the company to do
the following:
Conduct a new WIS and improve the wind integration cost methodology.
Evaluate the quantitative benefits of participation in the western EIM on the costs of
integrating variable resources into its operations.
Establish a TRC that, along with the company, will assess the feasibility of estimating the
unified costs of integrating wind and solar into its system and evaluate methods for
sharing those estimated costs between wind and solar resources.
Submit the updated solar integration study, new wind integration study, and assessment
ofjoint integration cost study to the OPUC.
The 2018 WIS focuses on the first item from the order. The objective of an integration study is to
investigate the operational impacts of the variability and uncertainty of intermittent resources,
like wind and solar, on the electric power grid, and to estimate the costs incurred to integrate these
types of generation resources. The estimation of these integration costs is used by the company
to facilitate a comparative evaluation of intermittent generation resources to other resource
options during the company's integrated resource planning process. Integration costs are also
used as a cost offset to the avoided cost price paid for must-take generation from qualifying
facilities as defined under the Public Utility Regulatory Policies Act of 1978 (PURPA).
The new WIS incorporates several changes and improvements from previous studies.
The primary changes and improvements are as follows:
Using actual observed Idaho Power system wind dataa
a
a Using the AURORA market model to simulate Idaho Power's system
Using the actual two-hour ahead (2HA) load and wind forecasts available to operations
a
a
a
a
Page 4
ldaho Power Company VER lntegration Analysis
a
a
Basing the hydro conditions under which the integration costs are determined on
50-percent exceedance, I O-percent exceedance, and 9O-percent exceedance
hydro conditions
Deriving operating reserves from application of the NERC BAL standard.
Each of these changes will be discussed in more detail later in this report.
3,1.1. Wind and ldaho Power's Sysfem
The amount of wind connected to Idaho Power's system has grown considerably over the past
decade, and at the time of the 2018 WIS analysis, totaled 727 MW of nameplate capacity.3
The most rapid growth occurred during 20l l and 2012, during which nearly 500 MW of capacity
were added. See Figure 1 for a graphical depiction of Idaho Power's wind resource additions
over the years.
800
6 @ @ tsts m (, oo o o -d N N o o rf I 66 @ @ N N (Eooooooooo:dHd::dooooooooooooooooooooooo oooNNNNdNNNNNNNNNNNNNNNNNNNNN\\\\\\\\\\\\\\\\\\\\\\\\\\\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \
Figure 1
Wind resources on ldaho Power's system
3 Idaho Power is required to sell renewable energy credits (REC) associated with the wind production
from the wind projects connected to its system. Thus, while the company has enabled the development
of this wind capacity through the enerry sales agreement process, it cannot explicitly represent the
output from the wind projects under contract as energy delivered to customers.
700
6m
500
lmo
300
200
100
3
=rF
Srr
=z
0+
Page 5
VER lntegration Analysis ldaho Power Company
3.1.2. Dispatchable Generating Capacity
Dispatchable generating capacity owned and operated by Idaho Power is critical to system
reliability. This generating capacity has long been used to follow customer load ramps. With the
growth in VERs over the past decade, dispatchable generating capacity is increasingly used to
provide the regulating reserves necessary for balancing VER output. For the 2018 WIS
modeling, Idaho Power designated a blend of resources capable of providing regulating reserves
to respond to intra-hour ramping needs. The blend of regulating reserve resources for the WIS
modeling totals 1,365 MW of nameplate capacity and consists of coal-fired generation
(Jim Bridger Plant), natural gas-fired generation (Langley Gulch Plant), and hydro generation
(Brownlee, Oxbow, and Hells Canyon plants).
3.1.3. 2017 Operations ,ssues
In addition to the influx of wind generation, from 2016 to 2017 approximately 289 MW of
PURPA solar were added to Idaho Power's system, resulting in a total of 1,016 MW of VER
projects. Prior to 2016, Idaho Power had limited operational experience with utility-scale solar
projects. VER curtailments in2017 exceeded all previous years' curtailments combined.
Multiple factors-the addition of non-dispatchable, must-take generation resources; relatively
flat load growth; high spring hydro conditions; and a low-priced energy market in the West-
contributed to the increased number of curtailment events in Idaho's balancing area (BA).
VER projects are curtailed when Idaho's BA is unable to maintain sufficient dispatchable
generation resources to respond to contingencies and provide regulating reseryes to respond to
changes in load and non-dispatchable generation. High river conditions with dam operating
restrictions and flood-control target levels will not allow Idaho Power's dispatchable resources,
such as hydro units, to reduce generation when VERs generate above forecast levels.
Low market prices make keeping thermal resources on-line and spinning to respond when VERs
generate below their forecast or down ramp unexpectedly very expensive. Additionally, other
possible reliability events, such as a line outage, require some dispatchable unit generation be
maintained in reserve to respond for reliability, public safety, or the protection of Idaho Power or
public equipment.
Page 6
ldaho Power Company VER lntegration Analysis
3
=
r,700
1,500
1,300
1,100
900
700
500
1,429 $W
MW
r}}
**r1.
300
1nA67 ON 4&nfi76:$
-
TOTAL AREA LOAP
-LoAD
NETSOTAR
lQn$l 12,0A 4r2tr0t, t&00 ,1/J/20,17 0:00
LOAD NET IVINO ANO SOLAR
-LOAD
NETWIND ANO SOLAR AND COGEN
Figure 2
Load and net load after VERs
Figure 2 illustrates an instance of the impact of VER (wind, solar) and cogeneration netted
against load, resulting in an integration problem. The top line (green) in Figure 2 represents the
total area load profile for a day in April 2017. The second line (blue) is the load netted with solar
generation, and the third line (yellow) is the load netted with solar and wind generation. The
fourth line (red) is the load netted with solar, wind, and cogeneration. The vertical reference line
indicates a point during the middle of the day when the total area load was 1,429 MW, yet the
load when netted with must-take resources was only 370 MW. Figure 2 does not include the
dispatchable hydro and coal generation required to be operated for environmental and
flood-control requirements and for contingency events and balancing.
As part of the Northwest Power Pool (NWPP), Idaho Power meets contingency reserve
requirements by maintaining capacity in reserve equal to 3 percent of load and 3 percent of
generation at all times. During the hour in the case shown in Figure 2, if there were no energy
imports or exports, the contingency reserve requirement would be 86 MW ([oad of 1,429 MW x
0.031 + [generation of 1,429 MW x 0.031 : 86 MW). The portion of the load available to be
served by conventional generation is 370 MW. The conventional generation would be required to
carry 23 percent of its output as contingency reserve (86 MW370 MW :23 percent).
Further complicating this operating scenario is the need for regulating reserves (up and down),
river-flow minimum constraints, adverse environmental effects of spill, and a lack of positive
PageT
VER lntegration Analysis ldaho Power Company
unit controllability in the operating range required to maintain the balance of load and generation
in these conditions.
When overgeneration conditions exist as described above, Idaho Power must export the excess
generation through off-system sales or curtailthe contributing VER generation if no market
exists or transmission constraints prohibit further energy exports.
3.1.4. lssues Not Addressed by the Study
The 2018 WIS focused on impacts and costs associated with errors in ZHA wind production
forecasts and the regulating reserves needed to respond to the forecast errors without
compromising compliance with the NERC reliability standard. The production cost modeling
performed for the study indicates higher production costs as a direct consequence of having to
carry the incremental, wind-caused regulating reserves. In this section, Idaho Power identifies
other impacts and costs associated with wind integration beyond the relatively narrow focus of
the 2018 WIS.
3.1.4.1. Day-Ahead Uncertainty
Idaho Power, similar to other regional BAs, performs day-ahead generation scheduling.a In the
2018 WIS, Idaho Power did not include impacts and costs associated with building readiness into
day-ahead generation scheduling to cover day-ahead uncertainty in wind production.
Idaho Power recognizes that capacity held in reserve to cover day-ahead uncertainty does not
necessarily provide response capability as readily as capacity held in reserve to cover 2HA
unceftainty. Nevertheless, the day-ahead forecasting of wind plant production, particularly the
timing of ramping events, can be problematic, and substantial and costly intra-day modifications
to day-ahead generation scheduling may be necessary.
3.1.4.2. Cycling Costs (Variable Operation and Maintenance Costs)
As noted earlier in this section, the 2018 WIS focused on the higher production costs associated
with having to carry incremental regulating reserves to cover erors in 2HA wind production
forecasts. The hourly production cost modeling performed for the study simulates the scheduling
of the incremental regulating reserves, but the actual intra-hour deployment of these reserves is
not simulated. In contrast to contingency reserves, which are deployed only in response to
relatively infrequent system disturbances (i.e., contingency events), regulating reseryes are
frequently deployed. The deployment of regulating reserves leads to a substantial increase in
intra-hour cycling of dispatchable hydro and thermal generating units, which is likely to cause an
increase in maintenance costs. Idaho Power has not estimated the increased maintenance costs
forthe 2018 WIS.
a Day-ahead scheduling is performed at intervals ranging from one day prior to several days prior for
weekends and holidays. For example, day-ahead scheduling for Sunday and Monday of a given week is
typically performed on Friday morning of the preceding week.
Page 8
ldaho Power Com VER I Anal S
3.1.4.3. Sub-Hourly Costs of Responding to Variability
The cost of deploying reserves to respond to intra-hour variability is not captured in the
integration analysis.
3.1.4.4. Reserve Violation lmpacts on lntegration Gosts
AURORA does not include a cost for reserve violations in the total portfolio cost.
Integration costs for intermittent resources are driven by a BA's need to carry incremental
operating reserves. Thus, to fully account for integration costs, production cost simulations
should reflect the necessary operating reserve requirements. If the production cost simulations
are unable to meet operating reserve requirements, as observed for the WIS, the production cost
simulations are not fully accounting for integration costs.
3.2. Study Design
Idaho Power designed the 2018 WIS with the objective of isolating the effects of integrating
wind generation in the operations modeling. Idaho Power used a common study design to meet
this objective, simulating system operations for a test year under the following two scenarios:
Load-alone share scenario: Base scenario for which the system is not burdened with
regulating reserves associated with wind and instead only has regulating reserves
associated with load's share of the total regulating reserves.s
Load net wind scenario: Test scenario for which the system is burdened with regulating
reserves associated with the netted load and wind time series.
A critical feature of this design is to hold equivalent model parameters and inputs between the
two scenarios, except for the regulating reserves. The incremental regulating reserves built into
the load net wind scenario simulations necessarily result in higher production costs for the
system, a cost difference that can be attributed to wind integration.
Idaho Power estimated the regulating reserves associated with the current buildout of wind
connected to its system ,727 MW of nameplate capacity. The company performed simulations
under the above-described two-scenario study design for the current buildout case.
Alternative buildouts, larger and smaller than the current buildout, were also simulated,
where the regulating reserves for the alternative buildouts were estimated based on determined
relationships between wind variability and installed wind nameplate capacity. The analysis to
determine the relationships befween wind variability and installed capacity are described later in
s Diversity benefit when neffing load and wind results in a total regulating reserye requirement less than
the sum of the separate regulating reserve requirements. Because of the diversity benefit, the total
regulating reserve requirement for the netted load and wind time series is found asX%o x (load
regulating reserve) + Xyo x (wind regulating reserve), where X is less than 100. The regulating reserves
for the load-share alone simulations were the first term only, yoTo x (load regulating reserve).
Both terms of the total regulating reserve formula were used for the load net wind simulations.
a
o
Page 9
VER lntegration Analysis ldaho Power Company
this report. The following alternative buildouts were simulated, defined in terms of installed
nameplate capacity:
a 300 Mw
a
a
o
o
a
500 Mw
1,000 Mw
1,100 Mw
800 Mw
900 Mw
The test year selected by Idaho Power for the study was 2017. Median hydro conditions for the
Snake River Basin and regionally for the Columbia River system were used for the simulations.
To investigate the effect of Snake River hydro conditions on the cost of providing regulating
reserves, sensitivity analyses were performed using very low (9O-percent exceedance) and very
high (1O-percent exceedance) hydro conditions.
3.3. Regulating Reselve Calculations and
Other Operating Reselves
3.3.1. Area Control Error
In performing the analysis to estimate regulating reserve requirements, Idaho Power analyzed
time-synchronous 1-minute time-step data for wind production and BA load from December
2016 through November 2017 . The actual wind and BA load data were compared to their
respective 2HA forecasts, where the 2HA forecast is a prediction of the hourly average.
The 2HA forecast assumption is predicated on the system's need to have adequate resources
available. 2HA has been determined as a reasonable amount of time for system Load Serving
Operations to schedule or procure resources in an economic fashion for the study reserve
calculations. For both wind and BA load, the 2HA forecast was assumed to transition in a linear
fashion over the 20-minute period centered on the top of a given hour. Figure 3 illustrates the
20-minute ramping of 2HA forecast BA load from 2,850 MW for the hour l4:00 to 15:00 to
3,000 MW for the hour l5:00 to l6:00.
Page 10
The hourly wind production profile used for the WIS simulations was identical for the two
scenarios (load-alone share and load net wind) and was also the same as that used for the
regulating reserve analysis described in the following sections. The referenced wind production
profile was for the l2-month period from December 201 6 through November 2017 . To simulate
a calendar year (i.e., January through December), wind data for December 2016 were appended
to the profile after November 2017.
3r00
3050
ilo00
2950
UM
285t
!
6
2800
14:00 ,.4:10 L4:2O 14:30 14:40 14:50 15:O0 15:10 l5;2O 15;30 15:40 15;50 16:@
Figure 3
Twenty-minute ramping of 2HA forecast BA load
The area control error (ACE) was then calculated from these data as the difference befween an
actual l-minute observation and its coffesponding2HA forecast:
ACE : Observed L-minute observationfor load (or wind) - 2HA hourly average forecast for
load (or wind)
3.3,2, ,VERC BAL Standard
The 1-minute ACE data for the l2-month period were analyzed to estimate the amount of
bidirectional regulating reserve that would have been necessary to comply with the NERC BAL
standard. Under this standard, non-zero ACE can be held for up to 29 consecutive clock minutes,
or non-zero ACE can be maintained for 30 consecutive minutes or longer provided ACE is
below the BA ACE limit (BAAL). For the study, Idaho Power assumed a BAAL of 0 MW;
the implication of this assumption is the analysis assumed ACE needed to be brought to 0 MW
for at least I minute for every 3O-minute interval. In other words, the company's analysis of the
historical load and wind data derived the amount of regulating reserve resulting in no
occurrences of 30 consecutive clock minutes of non-zero ACE for the l2-month period.
3.3.3. Estimation of RegUp/RegDn for Wind
Idaho Power is using the terms RegUp and RegDn for the bidirectional regulating reserve
necessary for balancing wind and load. RegUp is generating capacity that can be ramped up
intra-hour to respond to ACE undersupply conditions (load exceeding supply), and RegDn is
generating capacity that can be similarly ramped down to respond to ACE oversupply conditions
(supply exceeding load).
RegUp and RegDn for wind were expressed as a function of the 2HA wind forecast.
Specifically, RegUp was expressed as a percentage of the 2HA forecast:
RegUp MW : RegUpo% x 2HA windforecast
ldaho Power Company VER lntegration Analysis
Page 1 1
VER lntegration Analysis ldaho Power Company
RegDn was expressed as a percentage of the total nameplate wind capacity above the 2HA
wind forecast:
RegDn MW : RegDn?6 x (total nameplate wind capacity - 2HA windforecast)
Idaho Power estimated the amount of RegUp and RegDn for wind by iterative methods.
Under this approach, differing values for RegUp% and RegDn%owere evaluated for the
l2-month historical data period (December 2016 through November 2017) until compliance with
the NERC BAL standard was achieved. The determined reserve percentages accounted for the
nameplate potential from the 2HA forecast to the bounds of possible generation, from 0 to
727 MW . To capture seasonal effects and effects at different wind levels, the wind data were
binned first by season, then by 2HA forecast. Seasons were defined as follows:
o Winter: December, January, February
. Spring: March, April, May
o Summer: June, July, August
o Fall : September, October, November
The binning by 2HA forecast was defined as follows:
o Bin I ) 2HA wind forecast < 143 MW
o Bin 2 ) 143 MW < 2HA wind forecast < 321 MW
o Bin 3 ) 321MW < 2HA wind forecast < 536 MW
o Bin 4 > 2HA wind forecast > 536 MW
The 2HA forecast for RegUp and RegDn regulating reserve requirements for wind are provided
in Table 5. The equations for RegUp and RegDn providing the application of RegUp%o and
RegDn% are provided earlier in this section.
Table 5
RegUp and RegDn percentages for wind reserves based on 2HA wind forecast
Fall
RegUp% RegDn%
100Yo
8Oo/o
760/o
39o/o
Bin
1
2
3
4
66%
65%
75%
43o/o
Winter Spring Summer
RegUp% RegDn%RegUp% RegDn%RegUp% RegDn%
1O0o/o
86%
55%
49o/o
28%
51%
65%
34o/o
100o/o
94o/o
7 1o/o
43%
62%
79%
81o/o
690/o
100%
93o/o
68%
59%
48o/o
7 5o/o
85o/o
82Yo
Page 12
ldaho Power Company VER lntegration Analysis
3.3.4. Estimation of RegUplRegDn for Load
RegUp and RegDn for BA load were both expressed as a percentage of the 2HA forecast for
BA load:
RegUp: RegUpo% x 2HA BA loadforecast
RegDn: RegDno% x 2HA BA loadforecast
Similar to wind, Idaho Power estimated the amount of RegUp and RegDn for BA load by
trial-and-error methods. Under this approach, differing amounts of RegUp and RegDn were
evaluated for the l2-month historical data period (December 2016 through November 2017)
until compliance with the NERC BAL standard was achieved. Seasonal binning identical to that
used for wind was applied to the BA load data. The BA load data were also binned based on time
of day (TOD), where TOD binning for summer differed from non-summer seasons. The different
TOD binning for summer reflects the unique shape of summer loads relative to the other seasons.
The TOD bins were defined as follows:
Table 6
Winter, spring, fall
Hour Start Hour End BA Load Bin
0:00
"t :00
2:O0
3:00
4:00
5:00
6:00
7:00
8:00
9:00
10:00
11:00
12:00
'13:00
14:00
15:00
16:00
17:00
18:00
19:00
20:00
2100
22:00
23:00
1:00
2:00
3:00
4:00
5:00
6:00
7:00
8:00
9:00
10:00
11:00
12:00
13:00
14:00
't5:00
16:00
17:00
18:00
19:00
20:00
21:00
22:00
23:00
0:00
1
1
1
1
2
2
2
3
3
3
4
4
4
1
1
1
2
2
3
3
3
3
4
4
Page 13
VER lntegration Analysis ldaho Power Company
Table 7
Summer
Hour Start Hour End BA Load Bin
0:00
1:00
2:00
3:00
4:00
5:00
6:00
7:00
8:00
9:00
10:00
11:00
12:00
13:00
14:00
15:00
16:00
17 00
18:00
19:00
20:00
21 00
22:00
23:00
1:00
2:00
3:00
4:00
5:00
6:00
7:00
8:00
9:00
10:00
1'l:00
12:00
13:00
'14:00
15:00
16:00
17:00
18:00
'19:00
20:00
21:00
22:OO
23:00
0:00
1
2
2
2
2
2
2
3
3
3
3
3
3
4
4
4
4
4
4
4
4
The derived RegUp and RegDn percentages for BA load are provided in Table 8. The equations
for RegUp and RegDn providing the application of the RegUp% and RegDno/o are provided
earlier in this section.
Table 8
Derived RegUp and RegDn percentages for BA load reserves based on 2HA load forecast
Fall
Bin RegUp% RegDn%
1
2
3
4
8.0%
7.5%
9.9o/o
7.3%
10.60/o
8.9o/o
8.5%
7.1o/o
Winter Spring Summer
RegUp% RegDn%RegUp% RegDn%RegUp% RegDn%
4.9o/o
9.3%
9.5%
7.9o/o
9.1o/o
6.8%
5.8%
6.9%
8.1%
6.8o/o
9.9%
8.3%
10.50/o
11.3o/o
6.7%
7.0o/o
7.9%
8.1o/o
9.7%
6.2%
11.5%
6.0%
9.8o/o
13.3o/o
Page 14
ldaho Power Company VER lntegration Analysis
Table 9
Allocation factors for netted load and wind
RegUp RegDn
Winter
Spring
Summer
Fall
86.0%
84.60/o
92.6Yo
81 .00/o
78.4o/o
78.3%
70.5o/o
83.1o/o
As an example, the company's analysis found that the sum of 86 percent of each of
load-associated RegUp and wind-associated RegUp readies the system during the winter to
comply with the NERC BAL standard from an undersupply perspective, and 78.4 percent of
each respective RegDn similarly readies the system to comply from an oversupply perspective
3.3.5.1. DiversityBenefit
The allocation factors provided in the previous section are related to the diversity benefit;
load and wind are relatively uncorrelated, and consequently the errors in their 2HA forecasts do
not always augment each other (i.e., errors for each can be partially offsetting). Past studies have
credited this benefit entirely to the wind resource. For this study, Idaho Power is sharing this
diversity benefit between the load and wind elements of the load and resource balance.
Therefore, for the example described above, Idaho Power's simulation of production costs for a
system only needing readiness to respond to load variability and uncertainty carries only
86 percent of the load-associated RegUp and78.4 percent of the load-associated RegDn.
Under this method, load benefits from its diversity with wind, just as wind benefits from its
diversity with load.
3.3.5.2. ContingencyResen/e
For the production cost simulations, Idaho Power assumed a contingency reserve obligation
equal to 6 percent of system load, with at least half of the obligation required to be provided by
resources synchronized to the grid (Spin) and the remainder to be provided by resources capable
of responding within l0 minutes (NonSpin).This level of contingency reserve approximates
relatively well the current NWPP reserve-sharing contingency reserve obligation of 3 percent of
load and 3 percent ofgeneration. The level also reflects the need to set aside generating capacity
for operating reserve requirements to comply with disturbance control standards and control
performance standards. Contingency reserves remained constant for all simulations.
Page 15
3.3.5. Estimation of RegUplRegDn for Load Netted with Wind
When netting load and wind, Idaho Power found compliance with the NERC BAL standard does
not require the full arithmetic addition of the respective load and wind regulating reserve levels.
Idaho Power proportionally adjusted the respective load and wind regulating reserve levels
downward until compliance with the NERC BAL standard was achieved. Idaho Power referred
to the adjusted levels as allocation factors. The company found the following seasonal
allocation factors:
VER lntegration Analysis
3.3.5.3. Estimation of RegUp/RegDn for Alternative Wind Buildouts
Idaho Power analyzed 10-minute time-step wind production data to estimate the effect of
geographic dispersion on wind variability. The objective of this analysis is to characterize the
variability associated with alternative wind buildouts having differing geographic dispersion
from the current wind buildout of 727 MW of nameplate capacity. To estimate the effect of
geographic dispersion on wind variability, the company calculated the standard deviation of
progressively larger buildouts :
Buildout I : Fossil Gulch-Total nameplate : 10.5 MW
Buildout 2: Fossil Gulch plus Elkhom-Total nameplate: 111.2 MW
Buildout l6: Fossil Gulch plus Elkhorn plus ... plus Huntington-Total nameplate :
726.9 MW
Figure 4 is the standard deviation of the l0-minute time-step wind production data for summer
2017 (JunrAugust) of the l6 progressively larger buildouts plotted as a function of the buildout
nameplate capacity. The standard deviation increases with increased nameplate capacity;
however, the increase in standard deviation is proportionally slightly less than the increase in
nameplate capacity. This is likely the product of geographic dispersion occurring as wind
capacity is added to a buildout. Based on the analysis of the summer wind production data,
Idaho Power estimates that for every 1 percent increase in nameplate capacity there is an
approximately 0.93 percent increase in standard deviation.
y = -1E-05x2 + 0.2116x + 1.8071
o
a
a
160.0
140.0
,t20.0
100.0
80.0
60.0
40.0
20.0
oEEo
tr = 0.9978
100.0 200.0 300.0 400.0
IrtW buildout
500.0 600.0
Figure 4
Standard deviation of the 10-minute time-step wind production data for summer 2017
700.0 800.0
Page 16
ldaho Power Company
ldaho Power Company VER lntegration Analysis
Table 10 provides the increase in standard deviation found to occur with increased nameplate
capacity for all seasons.
Table 10
lncrease in standard deviation
% lncrease lnstalled MW % lncrease Standard Deviation (Variability)
Winter
Spring
Summer
Fall
1%
1%
1%
1o/o
0.99%
0.920k
0.93%
0.94%
Idaho Power used the analysis of standard deviation versus wind buildout to estimate the
regulating reserve requirement of alternative buildout futures. Under this approach, an increase
in standard deviation is considered to bring about proportionally an equivalent increase in
regulating reserve requirements. As an example, expanding the current buildout of 727 MW of
wind generation by 10 percent results in a new buildout of 800 MW of nameplate capacity.
Based on the standard deviation analysis, Idaho Power estimates the regulating reserve
requirements for the new (800 MW) buildout would seasonally increase by the
following percentages:
Winter: 9.9 percent
Spring: 9.2 percent
Summer: 9.3 percent
Fall:9.4 percent
Idaho Power analyzed alternative buildout futures ranging up as well as down from the current
buildout. Lower alternative buildouts at 300 and 500 MW were analyzed primarily to develop a
trend of integration costs as a function of buildout. Identifying a trend in changing costs at
different wind MW levels is informative to predicting costs at higher levels not studied.
The following altemative buildout futures were analyzed (defined in terms of nameplate capacity
and percent lesser or greater than the current buildout):
300 MW (59 percent decrease from the current buildout)
500 MW (31 percent decrease from the current buildout)
727 MW (current buildout)
800 MW (10 percent increase from the current buildout)
900 MW (24 percent increase from the current buildout)
1,000 MW (38 percent increase from the current buildout)
1,100 MW (51 percent increase from the current buildout)
a
a
a
a
a
a
a
a
a
a
o
Page 17
VER lntegration Analysis ldaho Power Company
While the short-term variability of the aggregate time series, as evaluated by the standard
deviation statistic, decreases with expanded buildout and associated increased geographic
dispersion, the longer-term average energy production is assumed to simply scale with expanded
buildout. For example, the 800-MW buildout, which constitutes an expansion of l0 percent from
the current buildout, was modeled as also having l0 percent more energy production.
3.4, System Modeling
The company used the AURORA model to perform the operational analysis and determine the
reserve component of the integration costs for the wind integration study. AURORA determines
the total portfolio cost for Idaho Power's system using Idaho Power's system resources and
market purchases and sales. The AURORA model is the same model the company uses for its
integrated resource plan (IRP), PURPA pricing, regulatory filings, and other types of operational
modeling and analyses.
The AURORA setup for the 2018 WIS includes the assumptions from the 2017 IRP updated to
include the actual load, wind, and solar production observed during the study period.
AURORA also incorporates the operational and contingency reserves in the form of hourly
inputs for RegUp, RegDn, Spin, and NonSpin. A total generating resource nameplate capacity of
1,365 MW was designated to provide reserves for RegUp, RegDn, and Spin, and an additional
444 MW of capacity were designated to provide reserves for NonSpin. The total 1,809 MW of
reserve carrying capacity included hydro, coal, and natural gas generation.
3.5. Modeling Results
3.5.1. Cost Resulfs for Simulation at Current Wind Buildout
To e stimate the costs of integrating wind, the company used a comparison of annual production
costs between two scenarios having different regulating reserves requirements, where the
difference in regulating reserves is related to wind's variability and uncertainty. The production
cost difference between scenarios was divided by the annual MWh of wind generation to yield
an estimated integration cost expressed per MWh of wind generation. The integration cost
calculation is summarized as follows:
Load-alone share scenario: Base scenario for which the system is not burdened with
regulating reserves associated with wind and instead only has regulating reserves
associated with load's share of the total regulating reserves.
Load net wind scenario: Test scenario for which the system is burdened with regulating
reserves associated with the netted load and wind time series.
The wind integration cost is the cost difference of the two scenarios divided by the MWh of wind
generation, where the quantity and shape of wind generation was the same in both scenarios:
Wind integration cost : (Load net wind scenario production cost - Load-share alone scenario
production cost) + Wind generation in MWh
Page 18
a
a
ldaho Power Company VER lntegration Analysis
The estimated integration costs for the current wind buildout are provided in Table 1 I
Table 11
Estimated integration costs for the current wind buildout
AURORA 727 MW simulation Production costs
Load-share alone simulation
Load net wind simulation
lncremental cost
$428,220,656
$436,434,800
$8,214,144
Wind MWh
Cost per wind MWh
1 ,815,626
$4.52
3.5.2, Simulated Dispatch of Reserve-Providing Resources
The differing production costs between the paired simulations (load-alone share and load net
wind) are a consequence of the differing dispatch of resources designated as capable of providing
regulating reserves. For the load net wind simulations, the reserve-providing resources are
dispatched less optimally to ready those resources to respond to the greater variability and
uncertainty of the load net wind time series. Figures 5 through l0 illustrate AURORA's
simulated operation of reserve-providing resources under the two scenarios.
Figures 5 and 6 illustrate the total hourly generation for the reserve-providing resources for the
two scenarios. These graphs illustrate that for the load-alone share simulation, Bridger is
dispatched during most of the year, except April, May, June, and October. In contrast, the load
net wind simulation dispatches Bridger in all months throughout the year.
Page 19
coCL
Eoo
Lo
=otLoEoE'
.oo
oc
c.9o
o)oc
trtu
9-zt- tLoz
6t-zl- Ltoz
t,t-zl- Ltozldzr- /t1z
tdzr- Lt1z9{ tt- Ltoz
6l- Lr- Ltoz
t t- t t- LlozLd tL- Ltoz
to- tt- Ltoz9{ot- Ltozo{ot- Ltoztt-ot- Ltoz
lc_oL- Ltoz
tc-ot- Ltozs-6dLtoz
6t-60- Lt1z
Et-60- Ltoz
Ld6dLtoz
L0-60-/ Lozg{eo LLozo-sdLtoz,r-ac- Lt1z
80-80-/ r0z
L|SA-LnZ
e{zL-LLoz
z{zr.-Ltoz
Lt-zt- Ltoz
I t-zL- Ltozgdzt-Ltoz
6{tL-LLoz
cdLt-LLoz
LL-I'_ LLOZ
zt- r l- Ltoz
9t,.r-LLOZ
ra0s{o
6 r.-0
e r-o
80-0zfot-Ltoz9-60-Ltoz0{60-Lroz
tt-6c- LtozBf6d/roz
c0-60-/ r0zs.go-Ltozz{BdLtoz
9t-80- Lloz
ot-gc-t-toz
,c-ec-Lo{LdLr{Ldt
8t-LC-Lzt-LdL
90-/0-/0t90-zsdedL
6 t-90-r.
0 r-90-/LdedL
r0-90-2e.strLtasdL
I L-90-Z
,- LLOZ
.-Ltoz
,- Ltoz
,- LLOZ
,_LLOZ
_.:
(Fo
r-c.lr--
B
(0o
c)
o0
63Lo0(')
'=
E()t')
th()o
ort()ko
OFd)=Z,Y0)o=Ed.ED=
L
Pe
o.Yo_(Bah
=6)o0Ec)o_c,! -j
.o6E3o)=
oi?toIr
=;<06 .n(l,E E)
=p =ED q oI)
I!J L
t0ztlztoz gtgz Etjz d]toz 6
toz gIOZ Ctoz -5LOZ I
l0Z p
LOZ ;
LOZ iLOZ I
LOZ o)
Loz 3toz El0Z x
l0z otoz 3tlz _5tlz r
LOZtoz :
toz tt0z
loz
Mt,II
oo
Looo!
E)
.=I
oLo-o=cf
Eo
c.o
GLo
Lo
Toclooo6
drEE:
oo(foooooooooooosc{o@(orifN
'9{to-t'o{LdL.,;LdL
'si- LdL'zdLdL'stso-t
odsdt'rL-gd Leos0z'zded/
'Ltgo- L'otso-t
'tt-sc_ L
eo-s0-z'zdsdL
'stvo-L'o-rdL.,1_'dL
'adttrL'zdrdL't-edL't{tdL
I r-00-/
e0-e0-z'zdedL'n.zd/
gt-zd /'zr-zdt-'strzdt
v{l0-L's{tdL'6t-tdL'tt-tdL'Lf tf L'ti-tc_L
tlztla
Loz
LOZ
Llz
toz
t0z
l0z
t0z
t0z
t0z
t0z
t0z
t0z
t0z
t0z
toztlztlz
t0z
'60-90-/ rcZ
€0-90-/ t0zL{vdLroz't{td/Loz
'gt-tc- Ltoz
ot-tdLtoz'vf?f Ltoz'6{tdLtoz't{edLnz'Lt-edLnz'zt-tILtoz
90-e0-/ toz'e{zf Ltoz
z{zdLtozgt-zo-LLoz
oL-zdLLOZ'sfzf LLoz'o{LdLLoz'v{tdLtoz
'8t- to- Ltoz'zt'td LLoz'gi- vdLvoz'tC- td Ltoz
ooooooooooooooo$a\oo(o\iN
M1,11
oNo
CDo(L
ilfl
il
?d
b
{I
NoE'o(L
'g-zt-Ltoz
'6t-zt- Ltoz'e l-zt- Ltoz
o0
oL9.
t-€E(.)(h
6
v)
qdLbI)
bI)
C)o0)Lq
0)
o
a
th
C)o
ott()tr(+-o
o
CBq
ah
E
(o
o
EE(!
0)
o
0)
dL
6
'a
aOboaic)9A:e
F,5a,-6-t ii
b0 00:- o)llL
(Fo
6,)iaB.i.a
>! i:'-E 7)r<(dE
FE e,
c.l F- Er '.E0).= >
=a)_!(\l-\ h00*j-g+>'
ol.l o XHoEtiho- -o (,E U KE'tr.= u-eJL-.lJ ts Ya5; gt
€_ruJ!sH-oLU--EE3;o00) = =EE3;FSAE:'icd(l,ci E:.o- t-,v iJ(r 0.)- tBo-c (, l--utV*LAL> b'E 8oZS.9EaPE.!
=> r}
ci- c.l€;EE;E.9i- trE€ i'Bi=Etr=.E6Lv_vElitrod(u.=ecdl E(g-0)u
-.=-cpv>!L9.o o.l *.E-€v
=OcBrL^U - r^,vYts9-;99<L6ca'- a'5ro9o
IclaEe:.9 t!*r.-
€ E E€
?.E s1O _.= rn
=PE;EB9--g8oEoc*?E
- [ic h^EEEgL--F
ootro
o
i5
!ttr
=o2!,.Eo
!,a!o
==ooo
v)()
L()
U)C)tr
bI)
(ll
bI)o
c)
bo
(0
(0I
EC)0
U)
C)
I(!
o
(Bav)
'1'
ah)o
lid
0.)
.L
C)q
c).)
3
o.
Jo
(E
o
c)
aqK}
=o!s
tr-o)rh=o-N=
-JI -s:s .olgGl (EbF FF
..4o.>oc
c.9o
o)os
d
IJJ
Loo.
Eoo
o
=o(L
o.c,os
t0z
loz
t0z
t0z
t0z
t0z
t0ztlz
t0z
t0z
t0z
LOZ
L0z
t0z
LOZ
LOZ
t0z
toz
L0z
t0z
toz
toztlz 8,toz EIOZ co
LOZ
toz -q-
LOZ E'toz -5l0z *
tlz otoz ttoz itOZ I
toz
l-Lt-Lt-L
t-L
t-L
t-L
l-L
l-L
l-L
t-L
t-L
t-L
'LO-Z
'LO-Z's-t'6ar'et-l
'LO-l'I0-t'e-0'o-o
'€ r-o
/0-0
r0-0s-6dL
6 r-60-/
e r-60-/Ld6dL
r0-60-/'9480-L
oagdLtt-gdL
80-80-/
t0-80-/
o.lE')otr
IoCNOoE
=!hlU
ltJ
toz
loztlz
toztlztlz
LOZtlz
L0z
t0z
t0z
toz
toz
toz
Loz
LOZ
toz
t0z
L0z
toz
toz
'e{LdL
'o-Lo-L'tt-LdL
'si-Lc-L'ztrttrL
'9{90-L'o{ec-L'rt-90-L'sded/
'zci-edL'L-90-L'o{sc-L
'rt-90- L'80-90-/
'zi-sc- L'e{tdL'o.to-L
'vt-vo- L'si-vc_L'zdro-L'L-€,0-L
'tdeo- L'9 L-eo-/
80-e0-/'zdEdL'r{zdL
'sl-zo- L'zL-zdL'strzi- t
re- Lo-/'s-t{L
'6 t- L0-z
€ r-10-/'Lc- tci- L'ti-tdL
so(v)d)o(o@
$@FO)l-d)O)Ot-o)SNJ---N<o
\lt@(ot\6lOOt\@@SJ--u)
o_
F.BgE 88888880t-(Olr)rt(9N-
ItAlll
I.
J
EL
c(E
CL
Eoo
Lo
=oo-oEop
g-zt-Ltoz
z-zt-Lt1z
Lt-zt- Ltoztt-zt- Llozgdzt- Ltoz6{tt-Ltoz
Ed tt- Ltoz
Lt- lt- Lloz
zL- L t- Ltoz9{ tt- LtozL{ot-Ltoz9-ot- Ltoz
6t-ot- Ltoz
t L-ot- Ltoz
8c-0t-t-toz
zd0t-Ltoz
9{6c_Ltoz
0{60-LLoz
tt-6c-LLOZ
so-60-/ r0zof6fzroz
s{sc-Ltozz{sc-LLoz
9t-80- LL\Z
ot-80- LLoz
'LtZt- L\OZ't.zt- Ltoz'9t-zt- LL\z'60-zt- Ltoz'eo-zt- Lto?,
' t. tt- Ltoz' td lt- Ltoz'9t-It-Ltoz
'6c-tt-Ltoz
' eo- I t- Ltoz's.ot-Ltoz
'zdot-Ltoz
'gt-oL- Lt\z'0t-0t- Lt\z'tc-o| Ltoz's.6dLtoz
'z{60-Ltoz
'9160-L\OZ'of6o-Ltoz
'vc-6ci'LLoz
'6.80-L\OZ'edaduoz'Lfaduoz'ttec_zloz'sfedzlozvtrei-to{Lc-LvtLo-L
et- to- tzt-Ldtgdzdz
oe-go-zs{gc-L
e l-so-z
e L-so-zttrstrL
r.o-so-zsdsdLv{so-t
t0z
i32 stoz Eloz d!
toztoz aIOZ o,,oz Eloz :toztoz Ptoz gtoz rtoz r
t0z
toz
t0z
o{Lo-Lvdto-tef n-tzf to-Lsf LdL
oe-e0-zvdso-te!-sfz
toz btlz 9toz '=
roz dl
LOZtoz At1Z o,toz Etoz J
tlztoz Itlz g
IOZ Iroz r
zv-sd L
sdeo-z
r.fso-z
s r.-so-z
sz s\-t
o r.-sfz
e L-sdz
,9o
Gc
c.9
o
o)oc
trIU
'tdrdt'tdvtrt'gt-vo-t'ot-ro-t'vtrvi- t
ofeo-z'ctto-t'u-eo-t
'zt-cdLgfefz'a{zf t'z{zdL'sfzo-L
'ot-zo- L'sfzdt
'oe-lo-z'v{r'e-t'ev-tc-t
'zv- vo- t'gdrdz'vi- r.f t
Loz
toz
toz
Loz
t0z
toz
L0z
toz
toz
toz
toz
tgz
toz
toz
toz
toz
toztlztlz
toz
CLfE"otr
IEc'=
o6oc
AEItJ
t[s[tvozldsdzlozsdvdu.ozat vo-ttoz
efvo-LtozLdvdttoz
tci-vi- tvoz
e{eo-u.oz
odeo-u.ozvftdttoz
eo-eo-z r.ozztref Lyoz
vdzdLvozet-zdu,oz
zv-zo-u,ozstrztrttoz
t dt o-ttozsdvdu.oz
at- vo- Ltoz
et-vc_ ttozttttttoztf ttttoz
88888880N(OrOt(ANFl l,ll
oooooooooooooooF-(Otr)sld)N-
NNoE)(U(Lmul
€€
!flr{rq
I
Coo)otr
Jcorogo
=!-lEltJ
I{
cf,N
o)o,oo-
t-
codo6@
N
N.c!
@q
lr)e,(Y)
c\iF-st_
N
@tr)
'ris(,,@J(Y)ot
N
@@o+rr) Gb(r).-:o)o)-
Ntr)(o+3o
to
@-
oo@ c.io)6o)o,(oN.
o)Nt- c.i5- .,)
o)@F
bo.
6s 8=
=sEt: g=c 0rx
= a=
@lr)N
$No-s
@
$$6lt-oF,-s
@
(o
@_
$os_N
@
t-(o
O)-lr)Nt.o,@
s$
s
6l
@@
(olr,@-
NrO@-(t,@
oolr,
@o@-
e
oooco
o
o)0)c
oo(o_s(f)N.s@s@
ooo)-(olr)N.oF-$@
rils@-
6t@o)_
o)lr)s@
ooo_
(f)o
(os$e
oo@-s(.)s-(O(.)se
rr)N
F$
t-o$@
olr,
f.-_
rr)o@-(.)@(f)e
DE'=.
6.9caEoEOco.=)o
s$(o-o
ryoF-so
(o
r{)(o-
o)
1r)-or{)so
(r)o-
@Flf)-
t-$sra
(f,
@o-
F-t\(e-(o(f,$@
@rr)(o-oNry@Nso
O)@$_so,$-
(f,o$@
olr)(\t-
o,o)o)-
@(oe
oco(u
o(Ub
?EO=Oco.=JO
==oC'q
Io
==oC)o-
A
==oo(rt
J
==ooo
Io
==FNNIo
==oou,
Io
==oo(r,
Io
orH
o!
a€!E\-oSEo't
tr.=GHSFolJ*'. =
*.,PG.5 e8T Eo8 ='so5SE E
=E =.SFp Iq: 6
LLoE grh_:ogE .E\.-\-aau Hq= EtL. 0) 'Eu, = 3EE E. : eE3 n iEc'i f; g,i
Coctot
I!,c
o'=-fioc
.d,EllJ
'ez-zt-il.02'z{zt-ttoz
' Lt-zl- Lloz
' I l-zt- LLoz'sdzt-u.oz'a{vt-ttoz'e{vt-ttoz
' Lt- t t- Ltoz'zt- tt- LLoz'gc-vt-ttoz't{ot-ttoz
'sdot-unz
'6t-ot-LLOZ
'€,t-oL- Ltoz'gi-ot-ttoz'ztot-ttoz'sdadil.oz'ofao-noz
'vv-ao-noz
edoo-z loz'eo-oo-z t oz'edeo-iloz'zdedilnz
Loz
toz
toz
i32 stoz Eroz d)
LOZtoz atoz Etoz E
i1l- "'toz gtlz ELoz -roz I
toz
toz
toz
toz
toz
toz
toz'a{eduoz'efto-ttoz'u.-etuoz
'zy-edu.oz
'so-eo-z 1.oz'atzo-ttoz'zdzdttoz'gt-zf ttoz'ov-zo-ttoz
'si-zi-u,oz'o{vo-ttoz'vtvo-ttoz
'ev-vo-ttoz'zt-vdttoz
'sc- vf ttoz'ti- tdttoz
'g L-eo-z
o L-eo-z'10-e0-z'o{to-t'v{tdL
'gt- to- t'zv- LO- t'si- tdt'oe-go-z's{so-t
'o t-go-z's l-90-z'ti-eo-t
't o-go-z's{sc-t'ttso-t'st-s1-t
oo-s0-z'edso-z'ttvo-t'ttto-t
'gv-vdt
'ov-vo-t'ttrvdt
88888880F-(O1r)sld)c'I-
,32o.>GC
c.oo
o)oc
ttu
CoCL
Eoo
o
=o(L
o.c,(EE'MW
Iilil
B
F.
I{
VER lntegration Analysis ldaho Power Company
The estimated production costs for the additional wind MWh included from the changing
buildout size were determined by using the same methodology used to determine average
PURPA costs. These costs are equivalent for each of the paired simulations at each
wind buildout.
Results for the wind buildouts toward the upper end of the studied range should be qualified as
likely underestimating the costs to integrate. As noted in the Regulating Reserve Violations
section later in this report, the AURORA production cost simulations for the expanded wind
buildouts identified occuffences in which the system modeling was unable to satisfy regulating
reserve requirements for load net wind scenarios. Consequently, the production costs for these
scenarios are not indicative of the full (and necessary) costs associated with regulating reserves;
AURORA does not assess a penalty or cost associated with the occuffence of unmet regulating
reserve constraints. As noted later, these occuffences may reflect the inability of the current
system of dispatchable resources to allow the integration of expanded wind buildouts without
compromising reliability. The decrease in integration costs reported in Table 13 for the
1,100-MW wind buildout relative to the 1,000-MW buildout is considered a manifestation of the
above-described inability of the model to satisfu the regulating reserve requirements in
production cost simulations and the consequential underestimation of production costs.
3.5.4. lncremental lntegrationGosfs
The integration costs provided in Table 13 for the seven analyzed buildouts are the estimated
per-MWh costs to integrate the total wind production for each of the seven buildouts.
However, the cost results can also be expressed on an incremental basis. The expression of
integration costs on an incremental cost basis is consistent with the principle of associating costs
with the causes of those costs. For example, the cost results can be used to estimate the
incremental per-MWh cost (S8.60/IMWh) associated with expanding from the current 727-l|l4W
buildout to an 800-MW buildout. This calculation is summarized in Table 14.
Table 14
lncremental integration cost for 727 MW to 800 MW of nameplate wind
Wind
Buildout
(Mw)
Annual Wind
MWh
lntegration
Cost
($/Mwh)
Total Annual
lntegration
Cost
lncremental
Gost
lncremental
MWh
lncremental
Cost
($/Mwh)
727 '1,815,626
1,991,358
$4.52
$4.88
$8,214,144
$9,725,937 $1,5'11,793 175,732
The calculated incremental integration costs for the remaining incremental buildouts and the
modeling reserve violations summary statistics are provided in Table 15.
Page 24
800 $8.60
l{)N
o)
EDo(L
(osoNsoNN$O)Ir)@ NN
trr)F@osc\r(oo)o(9@No) lr)- @_ s- o)_so)(oN6t@@
(fr0orr)$N(o@l-Nl-NNNl-S_S-N
\f(ooso@@ (r, (O l- O)od6ici;oi6io@@@@-e,@
ooN(f)oF-@N(r)@l*O)o)(oF-tsso|r,rr)ooo@sNtss$roNclN
o)@N@@(oF-c!qu?oqqq':NN<.<.Tr)TO()@eo@ae@
F\@@@@cDNrr)too- o)_ (o- (fr- o- (f)- @-o)o)rr)NN@(oo)(frF,-t\N@O)NSF-
---N6iN
(o@o(oorr)@ort*(o@
1 S- F-- cO- S- O)-(O@NNt(Ol-({)otf)ro(o(frF-
$N(\l@e@ee@
O(O$F-(Y)S(oorr)s(osrorO_ (O- F- Or- cO- ry ry(ON.rflr)St**toroNo(f)N@- (O- N_ t*- t- l'-- O-(fr@orNss@eqo e@@
o(ooo8EooNoo-
(f,tttl';lrFNlroooNoro(rrr()t-@HO-
o2
=.E=ETE
=>
toEC
=.e=H
E9'o
e,
EOOI20io3E
E
E=
Ht
E$
E-cE=9=
EEE;
6
9EEOEo(,
E
E-?6!'E er! aE a,< ao
E E'OOE
*sseE=.g6atr-
t
s=trYtr!,<c;
o
--g^E+=; E=
2
oco
(o
.9
o
L
@ooL
.E=
=E
=
oo-ooooc.9o
ct,o
.E
(Eco
EoLo
.E
E'E(uooooco
(uLE)o
.c
roi-(ooEEEG]l-o
.oa
(E
C
C
.oo
cnoc
tr
LU
C(,o-
Eo(J
c)
=oo-
oE(5E'
VER lntegration Analysis ldaho Power Company
3,5.5. Hydro Condition Sensifivity Analysis
To investigate the effect of Snake River hydro conditions on the cost of providing regulating
reserves, sensitivity analyses were performed using very high (10-percent exceedance) and very
low (90-percent exceedance) hydro conditions. The hydro condition sensitivity analysis was
performed using regulating reserves based on the current wind buildout (727 MW). As noted
earlier, the average integration cost found under a median (50-percent exceedance)
hydro condition is $4.52i\4Wh. The results of the hydro condition sensitivity analysis are
provided in Table 16.
Table 16
Hydro condition sensitivity analysis results
AURORA 7 27 -MW simulation
l0% Exceedance
Production Costs
90% Exceedance
Production Costs
Load-share alone simulation
Load net wind simulation
lncremental cost
$380,270,400
$388,658,531
$8,388,131
$480,144,781
$489,051,600
$8,906,819
Wind MWh
Cost per wind MWh
1,815,626
$4.62
't,815,626
$4.91
The results of the hydro condition sensitivity analysis do not differ substantially from the median
case result, suggesting basing integration costs on simulations using the median hydro condition
is appropriate.
3.5.6. Regulating Reserve Violations
AURORA identifies periods when the model was unable to satisfy the imposed regulating
reserve requirements, and to a lesser extent the contingency reserve requirements. Idaho Power
designates these occurrences in which AURORA's modeling of the system indicates potential
reliability issues as reserve violations. Table l7 shows the number of regulating reserve (RegUp
and RegDn) and contingency reserve (Spin and NonSpin) violations occurring under the
different wind buildout simulations. Simulations having regulating reserve for load-share alone
had no reserve violations; violations only started to occur when reserves were added for wind.
Under the current wind nameplate of 727 MW, there were 23 RegUp, 52 RegDn, and I NonSpin
violations, which means 0.9 percent of the time the model cannot meet the reserye requirements.
As more wind capacity is added, the violations increase substantially. AURORA's failure to
maintain reseryes at increasing levels of wind is a strong indication additional wind may not be
accommodated without significant changes to Idaho Power's system load and resources,
or changes to increase the control of wind during periods of low regulating reserves.
Page 26
ldaho Power Company VER lntegration Analysis
Table 17
Number of reserve violations, load net wind scenario
Nameplate
(Mw)
Regulation Up
(RegUp)
Regulation Down
(RegDn)Spin NonSpin
Total
Reserye
Violations
Percent of
Hours
300
500
727
800
900
1,000
1 ,100
2
1
23
9'1
255
435
690
1
7
52
133
522
988
1,736
6
1
I
22
11
I
3
14
76
232
799
1,434
2,434
<0.1o/o
0.2%
O.9o/o
2.60/o
9.1o/o
16.4o/o
27.8o/o
Table 18 shows the total MWh of deficiencies that occured during the simulations
Table 18
Total MWh of violations, load net wind scenario
Nameplate (MW) RegUp RegDn Spin NonSpin
21
6'1
700
2,183
18,755
58, 1 38
122,822
Table 19 shows the largest MW violation that occurred during the simulations
Table 19
Max MW of violations, load net wind scenario
Nameplate (MW) RegUp RegDn Spin NonSpin
300
500
727
800
900
1,000
1,100
3
4
297
2,325
11,075
28,323
60,1 02
0
0
0
0
0
0
0
0
69
24
103
264
117
56
300
500
727
800
900
1,000
1 ,100
2
4
44
90
152
214
260
21
26
36
54
117
179
242
0
15
22
17
21
15
17
0
0
0
0
0
0
0
Page 27
VER lntegration Analysis ldaho Power Company
4. ETEnGY IMBALANCE ManxeT AND VER !rureCRerION
The western EIM, in which Idaho Power began participating in April 2018, requires each
participant to be load-and-generation balanced going into each hour and to have sufficient
operational flexibility to respond to forecast errors and load and resource variability.
Participating in the EIM helps reduce the costs of responding to within-hour variability by
including a greater number of resource alternatives than would otherwise be available to respond
to errors and variability. The EIM is not designed to change the system reserve requirements and
flexibility needs associated with maintaining the system to comply with the NERC BAL
standard. The EIM improves Idaho Power's access to more cost-effective resources for
responding to within-hour forecast errors and variability.
Idaho Power's short experience with participating in the EIM has resulted in a couple of
observations. First, the NERC BAL standard allows for 29 minutes of system imbalance, but the
EIM wants the system balanced every l5 minutes. Consequently, the EIM has required more
frequent resource moves, either by Idaho Power resources or by other EIM participant resources
balancing the Idaho Power schedule. The frequent sub-hourly balancing has periodically exposed
Idaho Power to very high locational marginal pricing (LMP). The EIM LMP is capped at
S1,000 per MWh, which Idaho Power has experienced. Pam of the reason for the high LMP
exposure is Idaho Power's system is frequently off forecast due to the high penetration levels of
VERs with large forecast errors.
Another EIM observation is in the amount of flexible operating reserves Idaho Power is required
to maintain. Since becoming apart of the western EIM, the company has experienced an increase
in the quantity of flexible operating resources required by the EIM to pass the EIM flexibility
tests compared to the operating reserves Idaho Power maintained prior to joining the EIM.
The 2018 WIS does not include the costs of responding to within-hour variability or error,
but rather determines the cost of holding reserves to respond in the event they are needed.
Integration costs identified in the 2018 WIS are the increased opportunity costs of maintaining
adequate resources to reliably manage system 2HA forecast error and one-minute variability with
added variable generation.
The 2018 WIS determines the appropriate amount of flexibility to be held to respond to forecast
error and variability to comply with the NERC BAL standard using the December 2016 to
November 2017 ldaho Power actual system data. The NERC BAL standard RegUp and RegDn
reserves are then modeled in AURORA on a one-hour time step. The AURORA model simulates
the system operations and maintains resource availability to respond to the RegUp, RegDn,
and contingency reserves on Idaho Power's generating units. AURORA does not simulate the
sub-hourly movement of generating units to balance the system to the NERC BAL standard.
Although the reserves are determined using one-minute data, the AURORA model is set to run
on a one-hour time step. Consequently, costs of moving the units to respond to the within-hour
effors and variability are not captured in the AURORA modeling for the 2018 WIS.
As Idaho Power continues to gain experience participating in the EIM, the company will be
better able to assess the resulting impact on VER integration costs.
Page 28
ldaho Power Company VER lntegration Analysis
5. UrurnED WIND AND Solan lrurecnlnoN Cosrs
Previously, Idaho Power evaluated wind and solar integration costs in separate studies because
wind and solar have significantly different generating characteristics and therefore different
integration requirements and costs. Idaho Power has completed three wind integration studies
and two solar integration studies. The OPUC directed Idaho Power to consider a unified look at
the wind and solar integration costs. To accomplish a unified look, two analyses were completed
to evaluate wind and solar reserve characteristics and costs.
The addition of 289 MW of solar in 2017 has given Idaho Power a unique opportunity to
evaluate the differing generation and variability characteristics of wind and solar using actual
data from its system. The solar data evaluated consists of the actual solar data from May 2017 to
April 2018, which corresponds to the period when total solar equaled 289 MW for the entire
study period. The wind was evaluated during the same 12-month period.
The first evaluation was designed to compare equal quantities of nameplate wind and solar.
The output data for 289 MW of solar were used, and the output from a set of 14 wind projects
that equaled 289 MW of nameplate wind were used.
The load, wind, and solar data were analyzed to investigate the effect of wind and solar on
ramping requirements. The l0-minute changes6 for the following time series were investigated:
. Load alone
Wind
Solar
Load net wind and solar
The standard deviation for each time series was calculated. For each month, the standard
deviation of the load net wind and solar time series exceeded that of the load alone time series;
this result is an indication of the broader distribution of the l0-minute changes for the load net
wind and solar time series and reflects the increased ramping requirements brought about by
wind and solar. The monthly standard deviations for the two time series are provided in
Table 20.
Table 20
Monthly standard deviation of 10-minute changes, load alone time series, and load net wind and solar
time series
Month Load Alone Std Dev Load Net Wind and Solar Std Dev Percent lncrease Over Load Alone
a
a
a
28o/o
39o/o
57o/o
48o/o
JAN
FEB
MAR
APR
15
'15
14
15
19
21
21
23
6 For each quantity, the lO-minute change is defined as the difference between an observed value at time I
and the preceding observed value at time I - I 0 minutes.
Page 29
VER lntegration Analysis ldaho Power Company
Month Load Alone Std Dev Load Net Wind and Solar Std Dev Percent lncrease Over Load Alone
MAY
JUN
JUL
AUG
SEP
ocr
NOV
DEC
13
16
21
'19
15
14
'15
15
't9
20
23
21
'19
19
19
18
42o/o
29o/o
9Yo
12o/o
26Yo
39o/o
32o/o
15o/o
Figure 1 1 provides histograms of the lO-minute changes for the two time series for March 2018.
The histograms illustrate the broader distribution of the changes for the load net wind and solar
time series.
1,300
1,2OO
1,1@
1,000
900
I L@d , LnWnS
Strndrrd dcvirtion
Load alone = 14 MW
Load net wind and solar= 21 MW
b
0
qor
8{X'
7@
600
500
&o
300
zoo
100
-
!66
.o9to .96to
.JJII
,,$ a9 .O al
.nrs ,s r9'o ,s$
llr--.
fr $L
.$1to
19
lnto
o.rt*' r*o.."tt*o'!*
Binr - ,.Gminute changer (MW)
+ua-
tl*o n1ro' ^ro9t9oo%"'i"6**|*to"
,tot* ,"'*
Figure 11
Histograms of 10-minute changes for March 20'18, load alone time series and load net wind and solar
time series
Idaho Power then used incremental standard deviation (ISD) methods to calculate the respective
contributions of load, wind, and solar to the standard deviation of the load net wind and solar
time series.7,8 ISD methods are useful in determining the component drivers from a signalthat is
7 Bermejo, J., and L. Kirby. August 24,2076. PowerPoint presentation, Incremental Standard Deviation
Me thodol o gt BPA. bpa. gov/F i nance/RateCases/BP-
lSlbplSlCenoh2}lnput%o2}W orkshop%o2024o/o20Augusto/o2)2Ul6Yo20Final.pdf .
8 BPA. July 201 1- 2012 BPA final rate proposal, Generation Inputs Study, BP-12-FS-BPA-05.
bpa.gov/Finance/RateCases/lnactiveRateCases/BP I 2/Final%20Proposal/BP- l 2-FS-BPA-05.pdf.
Page 30
ldaho Power Company VER lntegration Analysis
the sum of several signals. For this application, the I 0-minute change in load net wind and solar
time series is the summed signal, and the component signals are the respective l0-minute
changes in load, wind, and solar. The following equation describes this application:
l}-minute L loadnetwindand solar: l)-minute A load + l0-minute Awind + l}-minute A solar
Figure l2 provides for each month the calculated respective contributions of load, wind,
and solar to the standard deviations of the totaled load net wind and solar time series. The line in
the graph is the calculated monthly standard deviation of time series of l0-minute changes in
load alone.
a
o
C
Eo
o
.9
.g
1)E
u
25.O
20.o
15.O
10.o
5.O
OCT NOV DEC
Figure 12
Monthly contributions of load, wind, and solar to the standard deviation of 10-minute time series
Under the ISD approach, the individual contributions equal the standard deviation of the total
time series. For example, the standard deviation of the time series of l0-minute changes for load
net wind and solar for March 201 8 is 2 1 MW. Figure I 2 illustrates for March 201 8 contributions
of 9 MW for load, 3 MW for wind, and 9 MW for solar; the three contributions sum to 21 MW,
which matches the calculated standard deviation of the total time series.
Figure l2 shows that the contribution of solar to the standard deviation equals or exceeds that of
wind for all months except July. This result indicates that solar, except during mid-summer,
is likely to have greater influence on ramping requirements than wind.
The within-hour variability is a key component to determining reserves, which contributes to
integration costs. Solar having alarger impact on a per-MW basis on variability and ramping
requirements is an important component in integration and a significant takeaway from
the analysis.
Page 31
FEB MAR APR MAY JUN JUt AUG SEP
r LOAD CONTRIBUTION TO NET TD DEV m wlNO CONTRISUTION TO NET STD DEV
ff SOLAR CONTRISUTION TO NET STD OEV +LOAD ATONE STO OEV
IAN
3.3
3,1 3.1
t0.:
1.7
4.t
z3
3.0
2.7
4.3
4.2
rt.5
3.5
10,iLS
VER lntegration Analysis ldaho Power Company
The second analysis evaluates the reserves calculated using the 2018 NERC BAL standard,
which is the basis for the 2018 WIS. The analysis looked at the current Idaho Power 727 MW of
wind, plus the 289 MW of solar output, to construct the reserves and compare them to the
reseryes constructed for 1,000 MW of wind used in the 2018 WIS. The aim was to evaluate the
reserves required for approximately the same amount of total energy for the two VER energy
mixes, one with 100% wind energy (1,000 MW of wind) and the second with approximately
74%owindand26Yosolar (727 MW ofwind + 289 MW of solar: 1,016 MW of wind and solar).
The VER comparison table (Table 2l ) shows the results for the base wind reserves for 727 MW
wind, 1,000 MW of wind, and727 MW of wind plus 289 MW of solar. The two wind-only
reserve scenarios do not include additional reserves for the 640,492 MWh of solar energy
provided by the 289 MW of solar included in the generation mix.
Table 21
lntegrationcostcomparisonof 727MW wind, 1,000MWof wind, and727 MWwindplus2S9MWsolar
Base Reserves for
727 MW Wind
Base Reserves for
1,000 MW Wind
Base Reserves
for Wind & Solar
Wind ttlWh
Solar lvlWh (* indicates no
additional reserves included)
MWh used in determining
reserves for VER
Wind MWh%
Solar MWh%
1,815,626
640,492-
1,815,626
lOOo/o
2,472,311
640,492-
2,472,311
100o/o
1,815,626
640,492
2,456,118
7 4o/o
26%
Total 1-year portfolio cost load
net VER
Total 1-year portfolio cost load
alone
$436,435,800
$428,220,656
$473,256,900
$458,519,656
$439,281,250
$428,191,844
Difference $8,214,144 $14,737,244 $1'1,089,406
lntegration cost per MWh in $$4.52 $5.96 $4.52
Base Reserves tor 727 MW wind
+ 289 MW solar
Base reserves for727 MW wind
$11,089,406
$8,214,144
$2,875,262
Solar MWh
$ per solar MWh
640,492
$4.49
* No additional reserves included
Table 2l provides information for several observations. First, the total portfolio cost for reserves
for 1,000 MW of wind is $14,737,24412,472,31 I MWh : $5.96 per MWh. The reserves integrate
2,472,311 MWh of wind energy. For essentially the same amount of total energy integrated,
thetotalportfoliocostforreservesforabase of 727 MWofwindand2S9 MWofsolaris
$l 1,089,40612,456,118 MWh : $4.52 per MWh. The reserves integrate 1,815,626 MWh of wind
energy and 640,492 MWh of solar energy, totaling 2,456,118 of VER energy. The portfolio cost
for the 1,000 MW of wind-alone reserves resulted in a $3,647,838 higher integration cost
Page32
Difference
ldaho Power Company VER lntegration Analysis
($14,737,244 - $l 1,089,406 : $3,647,838) compared to the portfolio cost for the combined wind
and solar reserves. We attribute this savings in the combined wind and solar portfolio to the
diversity benefit of the two resources' characteristics.
Wind is more uncertain, and solar is more variable. The sun rises and sets each day, limiting the
number of hours of uncertainty during the day for a solar resource. Wind does not have a
predictable schedule and is therefore more uncertain. Solar production is impacted by clouds,
and clouds are frequent and impact electrical production quickly. Wind does not tend to turn on
and off as quickly as solar. As the atmospheric pressure systems move throu gh an area, the wind
will rise and fall with periods of gusty, quick ramps. Although wind generation is variable,
solar is more variable than wind.
It should also be observed that the integration portfolio cost to integrate 727 MW of wind
generation is $8,214,144, integrating 1,815,626 MWh of wind energy and resultingina$.4.52
integration cost per MWh. However, this is the same integration cost as the $4.52-per MWh to
integrate 2,456,118 MWh in the base wind and solar scenario.
The difference in total cost between the two energy mixes (727 MW wind and 727 MW wind
plus 289 MW of solar) is $2,875,262.Dividing this amount by the increase of 640,492 MWh of
solar energy integrated results in an integration cost nearly equivalent to the $4.52 per MWh for
wind, at $4.49 cost per MWh for the incremental cost to integrate solar energy. The results show
the cost to integrate solar when paired with wind results in an integration cost very nearly equal
to that of integrating wind alone.
The number of violations for the 289 MW of solar with the 727 MW of wind are shown added to
the WIS violations shown inTable22.
Table 22
AURORA reserve violations count by scenario
Nameplate (MW) RegUp RegDn Spin NonSpin
Tota!
Reserue
Violations
Percent of
Hourc
300
500
727
800
900
1,000
1,100
727 Wind + 289 Solar
2
1
23
91
255
435
690
178
1
7
52
133
522
988
1,736
483
3
14
76
232
799
1,434
2,434
665
<0.1o/o
0.2%
O.9o/o
2.60/o
9.1o/o
16.4%
27.8o/o
76%
6
1
I
22
11
I
4
The number of total reserve violations is higher than with the 727 MW wind alone but is lower
than the 1,000 MW of wind. The fewer violations of wind and solar compared to the equivalent
amount of wind alone is consistent with the lower costs discussed above.
Page 33
VER lntegration Analysis ldaho Power Company
The maximum hourly violations for the 289 MW of solar with the 727 MW of wind are shown in
Table23.
Table 23
AURORA reserve violations maximum MW by scenario
Nameplate (MW) RegUp RegDn Spin NonSpin
300
500
727
800
900
1,000
1,'100
727 Wind + 289 Solar
2
4
44
90
152
214
260
131
15
22
17
21
't5
17
17
21
26
36
54
117
179
242
199
6. SYSTEM LIMITS AND Mnxruuu VER BuIIoour
Idaho Power recognizes its system has a limit to its capability to integrate VERs. Evidence from
this study of wind integration, as well as situations encountered during actual operations,
suggest the company is nearing the upper bound of this capability with the current VER buildout.
As noted in the section on2017 Operations Issues, VER curtailment has increased with the
addition of 289 MW of solar generation. Curtailments are generally linked to the inability to
provide flexible generating capacity for regulating reserve purposes during seasonal periods
marked by severe oversupply and regionally depressed wholesale electric market prices.
This inability to provide sufficient regulating reseryes is evident in practice by the periodic
VER curtailments and in model simulations by the increasing frequency of regulating reserve
violations at expanded wind buildouts.
The exhausting of the current system's operating reserves has significant implications for the
continued groMh of reserve-intensive VERs. As the wind study has alluded, the practical limit of
the current system is being encountered and is forecasted to increase in frequency with additional
VERs. Altering the current system's reserve carrying capacity and providing additionaltools to
system operators to respond to forecast error and short-term variability may be necessary.
Although beyond the scope of this study to evaluate new resources, it is anticipated that future
IRPs will include an evaluation of operationally flexible resources, such as batteries and pumped
storage, which can give operators flexibility to respond to real-time variability.
As an example, the costs for adding lithium-ion batteries to provide additional system flexibility
would add an additional S13.67 per MWh to the current level of wind integration costs.
(Using the IPC 2017 IRP resource cost assumptions and the results for the 727 MW of nameplate
wind, a 44-MW lithium-ion battery to cover the 44 MW of maximum RegUp violations would
result in a levelized annual cost of $24.8 million, which adds $13.67 per MWh to the integration
costs for reserve violation mitigation.)
Page 34
ldaho Power Company VER lntegration Analysis
Results of the 2018 WIS indicate that once wind penetration exceeds 900 MW, reserve violations
start to ramp up quickly, with violations exceeding l0 percent of all hours. That indicates an
incremental increase in wind penetration on the current system is significantly constrained past
an additional 173 MW (900 MW - 727 MW of current wind : 173 MW). Based only on the
initial evaluation conducted by this study, and because solar generation is also a variable
generation resource with similar integration costs as wind, Idaho Power proposes to define VER
integration cost tables to a buildout of 173 MW of additional nameplate capacity. The 173 MW
of additional VER results in approximately 1,190 MW of total nameplate VEP.(727 MW wind +
289 MW solar * 173 MW additionalVER: approximately 1,190 MW of totalVER) on a
system with a 3,400 MW peak and average sales of 1,755 MW. Expansion beyond this level
carries concerns that significant reliability issues will be encountered associated with the
system's inability to provide sufficient regulating reserves. VER development past this level
based on Idaho Power's current system configuration and the current state of technology for
available resources may not be possible. The company recognizes the energy industry is
experiencing a period of profound innovation, and developments such as new market tools
(e.g., EIM) or advancements in VER forecasting may enable VER buildouts beyond the 173 MW
of incremental VER penetration. However, the company also emphasizes VER capacity cannot
realistically be "un-built," and consequently, expansion beyond 173 MW of additional VER
nameplate capacity without verifying the ability to integrate such expansion would be imprudent.
7. CottcLUSIoNS
Evaluating the combined effects of wind and solar on reserve requirements and costs has been a
valuable exercise. The analysis has enhanced Idaho Power's understanding of the challenges and
complementary characteristics of combining load, wind, and solar generation. The TRC was
instrumental in providing feedback and guidance.
The results of this study and its varied analyses of wind, solar, load, EIM, and reserves indicates
a unified VER integration analysis approach may be the best way to assess costs for incremental
wind and solar. However, the analysis also indicates Idaho Power's system is nearing a
point where the current configuration can no longer integrate additional VERs.
Additional investigation is warranted into the combined effect of wind and solar, in a unified
VER integration cost analysis, along with the potential effects of participation in the EIM and its
unique requirements, attributes, costs, and benefits. The initial analysis as part of this study
points toward wind being more uncertain and solar being more variable, particularly within the
hour, which may have more identifiable impacts, implications, and/or costs as we move forward
with additional experience and history of operating as part of the EIM and its additional/varying
intra-hour requirements, timelines, and standards.
Based only on the initial evaluation conducted by this study, the cost of integratingT2T MW of
wind is equivalent to integrating72T MW of wind and289 MW of solar ($4.52 per MWh).
Therefore, the incremental integration charges that could apply to wind and solar per MWh of
output associated with incremental nameplate additions could be the same as those for
incremental wind (Table 24).Table 24 assigns a potential unified VER integration charge across
two additional tiers of incremental VER resources up to the maximum incremental addition of
173 MW, indicated by this study as what the current system configuration can integrate without
Page 35
VER lntegration Analysis ldaho Power Company
unacceptable regulating reserve violations and/or system inability to supply sufficient reserves.
However, as described in the 2017 Operational Issues section, the current quantity of variable
resources on Idaho Power's system periodically exhausts the operating reserves available.
The modeling results and number of actual wind curtailments during 2017 suggest a strong case
could be made that no additional VER resources should be put on the system to avoid periodic
reserve deficiencies.
Table24
Future integration cost recommendation for incremental VER project additions
Wind Nameplate
(MW)
Total Combined Wind 727 and Solar
289 Nameplate (MW)
VER Additions
Nameplate (MW)
lncremental Cost
per MWh
727-800
801-900
1 ,016-1 ,089
1,090-'1 , 189
0 73
173
$8.60
$1 I .1474
The quantity of additional VERs and the costs described in Table 24 are proposed based on the
AURORA modeling under median hydro conditions. It is strongly recommended that future
VER contracts include language that allows additional curtailment.
Idaho Power also believes additional VER generation development may have significantly
detrimental implications to maintaining adequate reserves.
Idaho Power's short experience with EIM will continue to be evaluated, as its impact to the VER
costs identified in this filing is not yet clear.
Page 36