Press Alt + R to read the document text or Alt + P to download or print.
This document contains no pages.
HomeMy WebLinkAbout20230501Direct Ellsworth with Exhibits.pdf
BEFORE THE IDAHO PUBLIC UTILITIES COMMISSION
IN THE MATTER OF IDAHO POWER
COMPANY’S APPLICATION FOR
AUTHORITY TO IMPLEMENT CHANGES TO
THE COMPENSATION STRUCTURE
APPLICABLE TO CUSTOMER ON-SITE
GENERATION UNDER SCHEDULES 6, 8,
AND 84 AND TO ESTABLISH AN EXPORT
CREDIT RATE METHODOLOGY
)
)
)
)
)
)
)
)
)
CASE NO. IPC-E-23-14
IDAHO POWER COMPANY
DIRECT TESTIMONY
OF
JARED L. ELLSWORTH
RECEIVED
2023 May 1, 4:44PM
IDAHO PUBLIC
UTILITIES COMMISSION
ELLSWORTH, DI 2
Idaho Power Company
Q. Please state your name, business address, and 1
present position with Idaho Power Company (“Idaho Power” or 2
“Company”). 3
A. My name is Jared L. Ellsworth and my business 4
address is 1221 West Idaho Street, Boise, Idaho 83702. I am 5
employed by Idaho Power as the Transmission, Distribution & 6
Resource Planning Director for the Planning, Engineering & 7
Construction Department. 8
Q. Please describe your educational background. 9
A. I graduated in 2004 and 2010 from the University of 10
Idaho in Moscow, Idaho, receiving a Bachelor of Science Degree 11
and Master of Engineering Degree in Electrical Engineering, 12
respectively. I am a licensed professional engineer in the State 13
of Idaho. 14
Q. Please describe your work experience with Idaho 15
Power. 16
A. In 2004, I was hired as a Distribution Planning 17
engineer in the Company’s Delivery Planning department. In 2007, 18
I moved into the System Planning department, where my principal 19
responsibilities included planning for bulk high-voltage 20
transmission and substation projects, generation interconnection 21
projects, and North American Electric Reliability Corporation’s 22
(“NERC”) reliability compliance standards. I transitioned into 23
the Transmission Policy & Development group with a similar role, 24
and in 2013, I spent a year cross-training with the Company’s 25
ELLSWORTH, DI 3
Idaho Power Company
Load Serving Operations group. In 2014, I was promoted to 1
Engineering Leader of the Transmission Policy & Development 2
department and assumed leadership of the System Planning group 3
in 2018. In early 2020, I was promoted into my current role as 4
the Transmission, Distribution and Resource Planning Director. I 5
am currently responsible for the planning of the Company’s wires 6
and resources to continue to provide customers with cost-7
effective and reliable electrical service. 8
Q. What is the scope of your testimony in this case? 9
A. I will first describe the Company’s proposed 10
methodology for establishing an Export Credit Rate (“ECR”) for 11
on-site customer generation exports. Next, I will describe the 12
Company’s proposed methods for valuation of the ECR, including 13
values for the avoided cost of energy, capacity (generation, 14
transmission, and distribution), line losses, and integration 15
costs. Finally, I will describe the Company’s proposed technical 16
requirements to support a modified project eligibility cap. 17
Q. Have you prepared any exhibits? 18
A. Yes. My testimony includes Exhibit Nos. 1-5. 19
Exhibit No. 1 is the summary pages of the workpaper supporting 20
the proposed ECR to be effective January 1, 2024, to May 31, 21
2024. Exhibit No. 2 is an Excel copy of the workpaper with 22
summary schedules and supporting data included. Exhibit No. 3 is 23
a copy of the Company’s T&D deferral calculation and Exhibit No. 24
4 is a copy of the Company’s most recent line loss study 25
ELLSWORTH, DI 4
Idaho Power Company
completed in March 2023. Exhibit No. 5 is a copy of the 1
Company’s most recent Variable Energy Resource Integration 2
study. 3
I. EXPORT CREDIT RATE VALUATION 4
Q. Is there value associated with on-site customer 5
generation exports? 6
A. Yes. As demonstrated in the Commission-acknowledged 7
October 2022 Value of Distributed Energy Resources Study (“VODER 8
Study”)1, there are several variables to consider when assessing 9
the value of on-site generation exports: 10
Avoided energy costs 11
Avoided generation costs 12
Avoided or deferred transmission and distribution 13
costs 14
Avoided line losses 15
Avoided environmental costs 16
Integration costs 17
Q. Did the Commission approve a method for valuing on-18
site generation exports? 19
A. No. The Commission found that the VODER Study was 20
completed in accordance with the Commission directives and 21
provided a basis for the Company to make recommendations in an 22
implementation case.2 The Commission directed the Company to 23
request any changes to its net metering service offering in a 24
1 See Attachment 1.
2 In the Matter of Idaho Power Company’s Application to Complete the Study
Review Phase of the Comprehensive Study of Costs and Benefit of On-Site
Customer Generation & For Authority to Implement Changes to Schedules 6, 8,
and 84, Case No. IPC-E-22-22, Order No. 35631 at 29 (Dec. 19, 2022).
ELLSWORTH, DI 5
Idaho Power Company
separate implementation case by proposing specific methods or 1
systems in support of changes to its customer generation service 2
offering. 3
Q. Has the Company developed its proposal and methods 4
for how the ECR should be valued? If so, please describe 5
generally how the Company developed its recommendation. 6
A. Yes. As articulated in Ms. Aschenbrenner’s 7
testimony, the Company identified four primary objectives as it 8
developed its overall proposal. Specific to the development of 9
the ECR, the Company sought to identify and apply methods that 10
result in a fair and accurate valuation of customers’ exported 11
energy while balancing customer understandability. The Company 12
also prioritized relying on recent data and implementing a 13
repeatable method for updating the ECR that will ensure timely 14
recognition of changing conditions on Idaho Power’s system and 15
the broader power markets. Generally, the Company relied on 16
avoided cost principles as a foundation for its proposal in this 17
case. 18
In the following portion of my testimony, I will describe 19
the methods the Company is proposing for each component of the 20
ECR. Mr. Anderson’s testimony will describe the Company’s 21
proposed annual update cycle for the ECR as well as the source 22
that will be relied on for each of the respective components of 23
the ECR. 24
Q. Please describe the term “avoided cost.” 25
ELLSWORTH, DI 6
Idaho Power Company
A. The term “avoided cost” is a commonly used term 1
which can be defined as the incremental cost that is not 2
incurred when additional output is not produced. More simply 3
stated, in the specific context of on-site customer generation 4
exports, when Idaho Power generates or purchases a kilowatt-hour 5
(“kWh”) of energy to serve customer need, there is an associated 6
cost. When, through customer action, the utility does not have 7
to serve that kWh, the avoidance of the cost associated with 8
generation or procurement of that energy is an “avoided cost.” 9
These costs are specific to costs “avoided” by the utility 10
system. As described in the VODER Study, there can be an avoided 11
cost of energy or capacity, as well as the related line losses, 12
associated with on-site customer generation exports. 13
Q. Please explain the significance of the Company’s 14
proposed measurement interval as it relates to developing the 15
ECR. 16
A. The measurement interval selected for net billing 17
is an important input to many of the components of calculating 18
the ECR under the Company’s proposed methods because it 19
determines the volume and timing of exported energy. As more 20
fully explained in Ms. Aschenbrenner’s testimony, the Company is 21
proposing to implement a real-time measurement interval. 22
Consistent with that recommendation, the Company has developed 23
the ECR valuation methodology relying on exports measured on a 24
real-time basis. It is important to note, the measurement 25
ELLSWORTH, DI 7
Idaho Power Company
interval selected to measure customer-generator exports should 1
be the same measurement interval used for the inputs in the ECR 2
calculation. A misalignment of the measurement intervals between 3
the ECR calculation and measurement of exports would result in 4
over- or under-valuing the ECR. 5
ECR Summary 6
Q. What structure is the Company proposing for the 7
ECR? 8
A. The Company is proposing a seasonal and time-9
variant ECR to compensate for energy and other elements 10
associated with avoided capacity, line losses, and integration 11
costs as I will describe in more detail. Exhibit Nos. 1 and 2 12
provide the workpaper summary and associated calculations for 13
the Company’s proposed methods. 14
Figure 1 shows a summary of the ECR components for the 15
methods proposed by the Company in this docket with an on-peak 16
and off-peak ECR. 17
// 18
ELLSWORTH, DI 8
Idaho Power Company
Figure 1 1
Proposed Export Credit Rate 2
3
If the Company’s proposal is approved as filed, the ECR 4
shown in Figure 1 will be in effect from January 1, 2024, 5
through May 31, 2024. As more fully explained in Mr. Anderson’s 6
testimony, the Company anticipates submitting a filing in April 7
of 2024 as part of a proposed annual update process (based on 8
most recently available information), to be in effect from June 9
1, 2024, through May 31, 2025. 10
Q. How are the proposed on- and off-peak periods 11
identified? 12
Season ECR
Export Profile
Volume (kWh per kW)Annual 1,465
Capacity Contribution (%)Annual 8.76%
Export Credit Rate by Component (cents/kWh)
Energy On-Peak 8.59 ¢
Including integration and losses Off-Peak 4.91 ¢
Annual* 5.16 ¢
Generation Capacity On-Peak 11.59 ¢
Off-Peak 0.00 ¢
Annual* 0.79 ¢
Transmission & Distribution Capacity On-Peak 0.25 ¢
Off-Peak 0.00 ¢
Annual* 0.02 ¢
Total On-Peak 20.42 ¢
Off-Peak 4.91 ¢
Annual* 5.96 ¢
*Annual values provided for informational purposes only and reflect seasonal
weighting for 12 months ending December 2022.
Note: On-Peak defined as June 15 - September 15, Monday - Saturday (exluding
holidays), 3pm - 11pm. All other hours defined as Off-Peak.
ELLSWORTH, DI 9
Idaho Power Company
A. The proposed on-peak hours are 3pm to 11 pm, June 1
15 through September 15, Monday through Saturday, excluding 2
holidays. As described in more detail in the avoided generation 3
capacity cost section of my testimony, these hours are those 4
currently identified as the hours of the Company’s greatest 5
system need for energy and capacity.3 6
Q. Did the Company consider differentiating seasonal 7
on- and off-peak time periods to compensate for energy and 8
capacity components of an ECR separately? 9
A. Yes. The VODER Study evaluated differentiating the 10
basis for defining seasonal on- and off-peak time periods for 11
energy separate from those hours of need for capacity and found 12
energy prices generally align with the Company’s hours of 13
capacity need.4 Additionally, the Company believes different on- 14
and off-peak periods for energy and capacity would add 15
additional complexity without commensurate benefit. As an 16
example, if different on- and off-peak times were used for 17
energy and capacity you could double the number of time periods 18
– potentially without a meaningful difference in the ECR for 19
certain time periods. For these reasons, the Company is 20
proposing a singular on- and off-peak ECR for both energy and 21
capacity components. 22
3 In the Matter of Idaho Power Company’s Application for Approval to Modify
Its Demand Response Programs, Case No. IPC-E-21-32, Application, Table 3.
4 Case No. IPC-E-22-22, October 2022 VODER Study, Appendix 4.8.
ELLSWORTH, DI 10
Idaho Power Company
Q. As illustrated in Figure 1, the Company is 1
proposing a single on- and off-peak ECR applicable to all 2
customer-generators. Did the Company consider calculating the 3
ECR by customer class? 4
A. Yes. However, as the Company explored the 5
feasibility of this type of an approach, several potential 6
issues were identified. For example, and as I will discuss when 7
describing the generation capacity component of the ECR, the 8
Company determined that quantifying class specific capacity will 9
result in over- or under- valuing capacity at a system level. 10
Additionally, the Company is generally concerned that class 11
specific ECRs will lead to customer confusion. While each 12
respective class’s export quantities and shape would be relied 13
upon in the development of the differing rate by class, that 14
nuance may not be apparent to customers, leading to confusion, 15
frustration, and potentially customer complaints from one class 16
of customers not understanding why their solar export is worth 17
less than another class’s export. 18
Avoided Energy Costs 19
Q. How is the Company proposing to value the avoided 20
cost of energy? 21
A. The Company recommends using a twelve-months ending 22
December 31 weighted average Energy Imbalance Market (“EIM”) 23
Load Aggregation Point (“ELAP”) price for the avoided energy 24
ELLSWORTH, DI 11
Idaho Power Company
component of the ECR. These prices would be weighted relative to 1
customer-generator exports over the twelve-month period. 2
Q. Did the Company consider other prices in its 3
proposed method for valuing the avoided energy component of the 4
ECR? 5
A. Yes. As discussed in the VODER Study,5 the Company 6
also considered the Intercontinental Exchange Mid-Columbia (“ICE 7
Mid-C”) and the Idaho Power Integrated Resource Plan (“IRP”) 8
forecast prices. Both methods were evaluated in the VODER Study 9
but ultimately the Company believes the ELAP prices most closely 10
meet the Company’s objectives of a value that fairly and 11
accurately reflects the value of on-site generation exports on 12
Idaho Power’s system and the broader power markets, while also 13
balancing customer understandability and a need for transparent 14
pricing. 15
Q. Did the Company consider using ELAP actual hourly 16
pricing rather than relying on a historical weighted average 17
price? 18
A. Yes. ELAP actual hourly pricing would be the most 19
accurate approach to valuing the avoided energy component for an 20
ECR. Under this method, the customer is compensated for all 21
exports at that hour's ELAP price which means that all 22
customers, irrespective of customer class, would be compensated 23
for the value of their energy at the time of day and year it is 24
5 See Attachment 1, pp 40-41.
ELLSWORTH, DI 12
Idaho Power Company
exported. This approach mitigates the risk of over- or under-1
compensating customers for their exports as there is no lag to 2
pass those market prices on to the customer generators. 3
However, dynamic pricing of this nature could be 4
challenging for some customers to understand due to ever-5
changing prices and potential volatility from one hour to the 6
next. Because actual ELAP prices are far more dynamic than a 7
fixed price - in fact they will likely vary hour-to-hour, 8
relying on this type of an input would assume customers are able 9
and willing to access real-time EIM data to respond to the price 10
signal sent in a given hour. While the Company could bill on 11
these dynamic market prices, making data available to customers 12
when reviewing their bills online would require significant 13
information technology system development and customization. In 14
the alternative, the Company considered developing a report that 15
could be generated on a monthly basis that would reconcile the 16
number of exports with the market price in each hour. However, 17
the Company does not view this as a viable approach as it would 18
require each customer to receive a 720 to 744 row reconciliation 19
report each month just to understand how the value of their 20
exports was determined. 21
Finally, this type of an approach does not provide 22
transparency for customers as there is no tariffed rate 23
associated with the exports, which could lead to customer 24
frustration or confusion. I am also not aware of any utilities 25
ELLSWORTH, DI 13
Idaho Power Company
in the country that have adopted such an elaborate structure for 1
compensating exports from customer-generators. 2
Q. Please describe why the Commission should approve 3
using the ELAP historical weighted average as the value of 4
avoided energy. 5
A. The use of a twelve-month weighted average to 6
develop the avoided energy value component of an ECR would allow 7
for ECR value(s) to be published in Idaho Power’s tariff and on 8
its website for public transparency and customer understanding. 9
This approach also provides for a repeatable method for updating 10
the ECR to achieve timely recognition of changing conditions on 11
Idaho Power’s system and the broader power markets. The Company 12
proposes to update the ECR annually, which will mitigate the lag 13
that would otherwise occur by updating less frequently, or if 14
the method relied on the average price of energy over multiple 15
years. Ultimately, the Company believes those benefits outweigh 16
the accuracy and timing of the more dynamic actual market price 17
approach. 18
Q. How does the Company propose the twelve-month 19
historical ELAP weighted average price be calculated? 20
A. The Company is proposing to calculate an on-peak 21
and off-peak value weighted with customer exports. The on-peak 22
period would align with the greatest system need.6 Currently the 23
on-peak time period for the ECR would be June 15 to September 24
6 Case No. IPC-E-21-32, Application, Table 3.
ELLSWORTH, DI 14
Idaho Power Company
15, 3-11 pm, Monday through Saturday, excluding holidays. All 1
other hours would be considered off-peak. Starting with the 2023 2
IRP and each successive IRP, the Company will evaluate and 3
update the hours of greatest system need that will inform the 4
annual update to the ECR. 5
Avoided Generation Capacity Costs 6
Q. How is the Company proposing to value the avoided 7
generation cost? 8
A. Three components collectively determine the avoided 9
cost of a generation resource: (1) contribution to capacity, (2) 10
the cost of an appropriate proxy, or alternative resource, and 11
(3) the energy generated during a given period. 12
Q. What method is the Company proposing to determine 13
the contribution to capacity? 14
A. The Company proposes to use the same method that is 15
utilized in the Company’s IRP process. At this time, that method 16
is the Effective Load Carrying Capability (“ELCC”) method – the 17
ELCC method would calculate the capacity contribution for all 18
on-site customer generation exports. 19
Q. Please describe the ELCC method and how it 20
determines the contribution of a resource to meet the Company’s 21
capacity needs. 22
A. The ELCC method looks at the equivalent capacity of 23
a given resource that can be added to or removed from the system 24
and maintain the same level of reliability. This method is more 25
ELLSWORTH, DI 15
Idaho Power Company
fully described on page 58 of the October 2022 VODER Study and 1
was also relied on as the method to value capacity of all 2
supply-side resources in the Company’s 2021 IRP. 3
Q. What other methods were considered in the 4
evaluation of capacity contribution? 5
A. Three methods were considered for the 6
quantification of capacity contribution: the ELCC method, the 7
National Renewable Energy Laboratory (“NREL”) 8,760 hour-based 8
method and a variant of the Peak Capacity Allocation Factor 9
(“PCAF”) method. These methods are more fully described on pages 10
58 to 60 of the October 2022 VODER Study. ELCC is the most 11
robust and accurate method to determine capacity contribution of 12
variable resources and it is widely utilized in the electric 13
industry as the preferred method to determine capacity 14
contribution. 15
Whereas the ELCC method calculates risk for all hours, 16
the NREL 8,760 hour-based method utilizes only the top hours of 17
a load duration curve as a proxy to determine the highest risk 18
hours. The PCAF method is based on a capacity factor during high 19
load hours and fails to account for any shift in high-risk 20
hours. Said plainly, both the NREL and the PCAF methods 21
oversimplify capacity contribution by assuming Idaho Power’s 22
resource needs align with the total system load. While that may 23
have been the case through most of the last century, the 24
development of non-dispatchable resources on the Company’s 25
ELLSWORTH, DI 16
Idaho Power Company
system has necessitated a change in how capacity needs are 1
identified and met. 2
Q. Why is a proxy, or alternative, resource utilized 3
in determining the avoided cost of a generation resource? 4
A. A proxy resource is utilized to determine the 5
equivalent capacity of the IRP-identified lowest-cost resource 6
that the on-site generation is avoiding. 7
Q. What resource does the Company propose be utilized 8
as a proxy resource? 9
A. The Company proposes to rely on the levelized fixed 10
cost associated with the least-cost dispatchable resource from 11
the Company’s most recently acknowledged IRP. In the 2021 IRP, 12
that was a simple-cycle combustion turbine (“SCCT”).7 13
Q. Currently, what is the Company’s periods of 14
greatest capacity need? 15
A. Currently, the highest Loss of Load Probability 16
(“LOLP”) hours are from 3 to 11 pm, June 15 through September 17
15, Monday through Saturday, excluding holidays. 18
Q. How are the highest LOLP hours calculated? 19
A. LOLP can be calculated by determining the 20
probability that the available generation at any given hour is 21
able to meet the net load during that same hour. The highest-22
risk hours are those which have the highest LOLP values. 23
7 2021 IRP, Appendix C at 38.
ELLSWORTH, DI 17
Idaho Power Company
Q. How do the highest LOLP hours pertain to the ELCC 1
calculation? 2
A. In general terms, the ELCC calculation is driven by 3
the quantity of generation produced during the highest-risk 4
hours. 5
Q. How is the Company proposing to compensate 6
customers for avoided generation capacity? 7
A. The Company proposes to distribute the calculated 8
avoided generation capacity value across on-site generation 9
exports during the Company’s identified period of capacity need. 10
Q. Why does the Company believe it is reasonable to 11
provide a time-variant credit for capacity? 12
A. The procurement of capacity resources is driven by 13
the identified hours of highest risk - the period that capacity 14
can be avoided. By aligning the period of capacity avoidance 15
with that of the ECR, a price signal is created that could 16
incentivize customers to invest in or optimize systems to 17
maximize output during the period of capacity avoidance. 18
Examples of a potential incentivized price signal with a time-19
variant credit for capacity include systems with optimized 20
direction of panels or installation of energy storage devices. 21
Q. Does the proposed methodology satisfy the 22
Commission’s aim of having an avoided generation capacity value 23
that accurately considers actual avoided costs?8 24
8 IPC-E-22-22, Order No. 35631 at 29.
ELLSWORTH, DI 18
Idaho Power Company
A. Yes. The proposed methodology aligns with the 1
Company's IRP process for determining greatest capacity needs. 2
Future IRPs will identify the hours of greatest capacity need, 3
which will be used to determine the capacity avoided by 4
customers with on-site generation. 5
Q. Did the Company consider incorporating its next 6
capacity deficiency date when valuing exports? 7
A. Yes, however the Company identified several issues 8
that it believes would be challenging to overcome. First, 9
relying on a capacity deficiency period would necessitate the 10
Company tracking and applying a different ECR depending on the 11
vintage of systems. Second, this type of an approach would not 12
be easily understood by customers, as it would result in 13
differing ECRs (one that excludes capacity) for a number of 14
years, and then inclusion at a future point in time. Ultimately, 15
it is the Company’s position that the capacity deficiency date 16
should be considered in most avoided cost applications; however, 17
valuing exports from customer generators inherently presents a 18
set of challenges that are not present with large, utility-scale 19
projects. 20
It is also important to note that at this point in time, 21
the Company is capacity deficient, so compensating customer 22
generators for an avoided cost of capacity is reasonable. The 23
Company is open to incorporating the capacity deficiency period 24
if this can be done in a manner that is fair for all customers, 25
ELLSWORTH, DI 19
Idaho Power Company
can be consistently applied, and does not result in unnecessary 1
customer confusion. 2
Q. Did the Company consider calculating the avoided 3
generation capacity component of the ECR by customer class? 4
A. Yes. However, the Company identified a few issues 5
related to determining the ELCC by class rather than considering 6
all on-site generators as a single resource. First, a class-7
specific ELCC determination creates a timing issue based on the 8
order in which the class ELCCs are calculated – resulting in a 9
higher value to whichever class adds capacity to the system 10
“first.” Additionally, due to the relatively small size of 11
certain customer classes, in terms of megawatts, the margin of 12
error in the ELCC calculation increased significantly, resulting 13
in an inaccurate valuation of capacity avoided and potential 14
over-payment for capacity. For these reasons, in combination 15
with my earlier comments about customer understandability, the 16
Company is not proposing to calculate the avoided generation 17
capacity component of the ECR by customer class. 18
Deferred Transmission & Distribution Capacity Costs 19
Q. What is Idaho Power proposing for the value 20
associated with avoided, or deferred, transmission and 21
distribution (“T&D”) costs? 22
A. The October 2022 VODER Study presented a method 23
that incorporates data specific to Idaho Power’s electrical 24
system to determine what transmission and distribution projects 25
ELLSWORTH, DI 20
Idaho Power Company
could be avoided or deferred and the associated value. This 1
method has been recognized as a best practice by energy industry 2
expert Kurt Strunk, managing director of NERA Economic 3
Consulting.9 As described in Mr. Anderson’s testimony, the 4
Company would plan to update these calculations it its 2024 5
annual update – after the next IRP has been filed. 6
Q. Did the Company consider other methods? 7
A. Yes. The VODER Study considered other T&D deferral 8
approximation methods that, when applied in the absence of 9
project-level data, may provide a reasonable proxy for T&D 10
deferral value.10 These “top-down” approximation methods often 11
rely on general utility information and ignore or make 12
assumptions about whether T&D investments could be deferred. 13
When project-level data is available, as it is for the Company’s 14
proposed method, it is the preferred analysis method. It 15
provides the most applicable and accurate calculation of the T&D 16
deferral value because it considers how and when T&D investments 17
are made. 18
Q. Please explain how the Company’s proposed method 19
calculates the value of deferred transmission and distribution 20
capacity from customer-generator exports. 21
A. To determine the potential value of on-site 22
generation in deferring or delaying the need for Idaho Power to 23
9 IPC-E-22-22, Idaho Power Reply Comments, Attachment 1 – Affidavit of Kurt G.
Strunk.
10 Case No. IPC-E-22-22, October 2022 VODER Study, pp. 71-72.
ELLSWORTH, DI 21
Idaho Power Company
build T&D resources, the analysis identifies local peak hours 1
for each T&D resource. Local peak hours are specific to the 2
amount and types of loads connected to individual resources. The 3
analysis incorporates 15 years of historical project data and 4
five years of forecasted project data on Idaho Power’s T&D 5
system. This data identifies the historical trends and projected 6
T&D projects and the capacity need for each project. Exhibit No. 7
3 includes the T&D deferral calculations. 8
Q. How is the Company proposing to compensate 9
customers for avoided T&D? 10
A. The Company proposes to compensate on-site customer 11
generation exports for the value of deferred T&D during the same 12
hours as described for the avoided generation capacity component 13
of the ECR. 14
Q. Did the Company consider using different hours for 15
T&D deferral value? 16
A. Yes. The avoided cost of T&D could be spread over 17
all exports in a given year; however, the Company believes it is 18
most reasonable to align with the hours of system need. While 19
not all T&D projects are deferrable by on-site customer 20
generation, a vast majority of the deferrable projects are 21
projects that would have otherwise been installed to serve 22
system need during those highest risk hours. 23
// 24
25
ELLSWORTH, DI 22
Idaho Power Company
Avoided Line Losses 1
Q. What is the Company proposing as it relates to 2
avoided line losses in the ECR? 3
A. The Company is proposing to include an adjustment 4
to the avoided energy value of the ECR to account for the 5
benefit of avoided line losses. The Company proposes to use its 6
line loss study completed in March 2023; Table 1 provides a 7
summary of the avoidable transmission and distribution line 8
losses. Exhibit No. 4 of my testimony is a copy of the Company’s 9
2023 line loss study. 10
Table 1 11
Energy Loss Coefficient Table from 2023 Line Loss Study 12
13
Q. What does an avoided line loss percentage value 14
represent? 15
A. These values represent the reduction in losses that 16
the Company avoids from a reduction in serving load due to 17
exports from customers with on-site generation. 18
Q. How were the avoided line loss values calculated? 19
A. The losses for transmission lines were obtained by 20
applying Ohm’s law to the conductors, that is, the current 21
measured at one end of the line squared times the resistance of 22
the line. The losses in the distribution system were obtained by 23
VODER
System Level Energy Loss Coefficient Peak Loss Coefficient
Transmission 1.026 1.034
Distribution Station 1.029 1.037
Distribution Primary 1.044 1.050
Distribution Secondary 1.044 1.050
ELLSWORTH, DI 23
Idaho Power Company
determining the hourly energy leaving a station in comparison 1
with the energy consumed by all customers served by that station 2
during the same hour; the difference between those two energy 3
values equates to the losses in the distribution system. The 4
transformer losses were obtained by adding the transformer core 5
losses and the transformer winding losses together. 6
Transmission line losses, distribution primary line 7
losses and transformer winding losses are the only line losses 8
that can be avoided by customers with on-site generation; the 9
sum of these three loss components equate to the hourly isolated 10
avoidable losses. 11
Q. How are the avoided line losses valued in the ECR? 12
A. The avoided line losses are applied to both energy 13
and capacity. The energy prices are multiplied by the loss 14
percentage to determine the corresponding impact of the energy 15
price due to losses. This calculation is depicted in Figure 4.21 16
on page 78 of the October 2022 VODER Study. For capacity, the 17
Company proposes a slightly different method for valuing the 18
line losses. For the capacity line losses, hourly customer-19
generator exports are scaled up to be inclusive of avoided 20
losses and then utilized in the generation capacity value 21
calculation. 22
// 23
24
25
ELLSWORTH, DI 24
Idaho Power Company
Environmental & Other Benefits 1
Q. Is Idaho Power proposing to include any value 2
associated with externalities such as local job creation, 3
avoided health costs, or other environmental benefits? 4
A. No. These externalities are just that – external to 5
the utility’s system and have not been included in the Company’s 6
proposal for implementation. Similarly, environmental benefits 7
based on non-quantifiable or speculative values are not 8
appropriate to include in the ECR. In Order No. 35631, the 9
Commission stated the following: 10
Generic conclusions and recommendations from 11
third-party studies that do not fully reflect 12
the environmental conditions and legislative 13
requirements in Idaho or the particulars of 14
the Company’s system, should not be considered 15
by the Company in its implementation 16
recommendations. Likewise, environmental 17
benefits or costs that cannot be quantified or 18
shown to affect customers’ rates, should not 19
be considered in valuing an ECR.11 20
In accordance with the Commission findings in Case No. 21
IPC-E-22-22, and as more fully described in the VODER Study,12 22
the Company has not proposed to include a value in the ECR at 23
this time. However, if environmental and/or legislative 24
requirements change, the Company anticipates initiating a docket 25
that would seek to modify the ECR methodology to include any 26
11 Order No. 35631 at 29.
12 Attachment 1, pp. 78-81.
ELLSWORTH, DI 25
Idaho Power Company
benefits or costs that can be quantified and affect customers’ 1
rates. 2
Integration Costs 3
Q. Does Idaho Power incur costs to receive excess 4
energy from on-site generation? 5
A. Yes. All Variable Energy Resources (“VERs”) such as 6
solar and wind cause incremental costs associated with 7
accommodating variable resources on the system. Examples include 8
dispatchable unit cycling from increased unit stops and starts, 9
increased load following ramping, and imperfect unit commitment 10
and dispatch. 11
Q. Is the Company proposing to include costs in the 12
ECR to account for integration costs? 13
A. Yes. 14
Q. How does the Company propose to account for 15
integration costs? 16
A. Idaho Power periodically conducts integration 17
studies based on the number of variable resources on its system. 18
The most recent integration study was completed in 2020 and 19
reflected the then-current level of intermittent generation on 20
the system. The 2020 VER Integration Study relied on a 2023 base 21
year and determined the costs to integrate additional variable 22
resources including customer generation under a variety of 23
assumed conditions on the Company’s system and the broader power 24
markets. Exhibit No. 5 includes a copy of the 2020 VER 25
ELLSWORTH, DI 26
Idaho Power Company
Integration Study. The October 2022 VODER Study13 also summarizes 1
the 2020 VER Integration study results. 2
The Company has proposed to use the 2020 VER Integration 3
Study Case Number 1 integration cost. The 2020 VER Integration 4
Study identifies an applicable solar integration rate in Case 5
Number 1 of $2.93 per megawatt-hour (“MWh”), or $0.00293 per 6
kWh. 7
Q. Please explain why the 2020 VER Integration Study 8
Case Number 1 is most appropriately applied? 9
A. Case Number 1 is directly comparable to the base 10
case (Case Number 7). While a Bridger unit is retired in Case 11
Number 1, it is also retired in the comparative Case Number 7. 12
Therefore, all of the integration costs can be attributed to the 13
difference of 251 MW of solar between cases. The identified 14
$2.93 per MWh is therefore a solar integration cost and is 15
appropriate in the purposes of the ECR. 16
Q. Did the Company consider developing a new VER 17
Integration Study in support of its request in this case? 18
A. Yes, however VER Integration Studies are complex 19
and the 2020 study is still considered to be current. While the 20
last VER study was performed by an externally contracted 21
company, the key Idaho Power personnel who are involved with the 22
development of the study are also key contributors to the IRP. 23
Similar to the IRP, the Company also solicits feedback for a VER 24
13 See Attachment 1, page 83, Table 4.10.
ELLSWORTH, DI 27
Idaho Power Company
study through a stakeholder process called a Technical Review 1
Committee. Therefore, the opportune time to complete a VER 2
Integration Study for both stakeholder engagement, and to ensure 3
adequate Company representation, is during those years between 4
IRPs. Idaho Power expects to complete its next VER Integration 5
Study, if necessary, following the completion of the 2025 IRP. 6
Q. Will the addition of battery storage impact solar 7
integration costs? 8
A. Potentially. This will be determined in the next 9
VER Integration Study. Generally new VERs will continue to 10
increase integration costs, and battery storage can potentially 11
act to counter those cost increases. The Company plans to gather 12
operational data on recent and planned near-term solar 13
additions, as well as determine the operational characteristics 14
associated with battery storage the Company is able to leverage, 15
prior to beginning the next VER Integration Study. 16
II. PROJECT ELIGIBILITY CAP 17
Q. Please explain the Company’s proposal for the 18
project eligibility cap for on-site generation systems. 19
A. As described in Mr. Anderson’s testimony, the 20
Company is not proposing to modify the 25 kW project eligibility 21
cap for Schedules 6 and 8. The Company is, however, proposing 22
that the project eligibility cap for Schedule 84 be set at the 23
greater of 100 kW or 100 percent of demand at the service point 24
for commercial, industrial, and irrigation customers. 25
ELLSWORTH, DI 28
Idaho Power Company
Q. In your opinion, could the Company safely 1
interconnect systems larger than that the proposed demand-based 2
cap? 3
A. Yes. However, not without system upgrades – some of 4
which could be significant. While the on-site generation 5
customer would be responsible for the initial cost of that 6
equipment, the ongoing cost, including maintenance, replacement, 7
property taxes, and other ancillary costs will become the 8
responsibility of the Company. These costs are collectively paid 9
for by all customers. The Company does not routinely install 10
facilities in excess of customer demand in any other instance 11
and it would be inappropriate to do so here. Ultimately, the 12
benefit of tying a system size to customer demand is to ensure 13
Idaho Power does not have oversized distribution equipment on 14
its system necessary to serve those customers. 15
Q. Mr. Anderson’s testimony states that the existing 16
project eligibility cap for Schedule 6 and 8 is larger than 17
their average demand. Does this result in a similar magnitude of 18
concern to the Company? 19
A. No. Most transformers on Idaho Power’s system to 20
serve residential and small general customers are 25 kW or 21
larger, so it is not as common for residential customers to have 22
to complete system upgrades when installing on-site generation. 23
This is not to say it does not happen, as often there may be 24
ELLSWORTH, DI 29
Idaho Power Company
multiple customers on a shared transformer, but the occurrence 1
is less common. 2
Interconnection Process Overview 3
Q. What does the Company generally require of 4
customers installing generation for exporting systems? 5
A. The customer is required to complete the 6
application process as outlined in Schedule 68, Interconnections 7
to Customer Distributed Energy Resources (“Schedule 68”). This 8
process includes requirements for how upgrades to the Company’s 9
system may be treated and what types of equipment is required on 10
the customer’s side of the meter. 11
Q. Please describe the Company’s process to determine 12
whether and to what extent upgrades or modifications may be 13
required on the Company’s side of the point of delivery. 14
A. Because exporting systems are operating in parallel 15
- meaning they are connected to and receiving voltage from Idaho 16
Power’s system - it is critical to implement requirements that 17
will provide for the safety of Company employees and members of 18
the public, as well as integrity of the system through system 19
protection equipment, as necessary. The Company performs a 20
Feasibility Review to evaluate the feeder capacity, phase 21
compatibility, and transformer size. If any of these fail the 22
Feasibility Review, the customer is required to fund upgrades 23
before interconnecting their generation facilities. 24
ELLSWORTH, DI 30
Idaho Power Company
Q. What type of equipment or requirements are imposed 1
on the customer and/or equipment installed on the Company’s side 2
of the point of delivery? 3
A. Safety is critical with any interconnection to the 4
Company’s system. As outlined in Section 1: General 5
Interconnection Requirements of Schedule 68, Idaho Power 6
requires inverters meet Institute of Electrical and Electronics 7
Engineers (“IEEE”) standards; there must be an operable 8
disconnect switch present, proper signage, and the disconnect 9
switch must be readily accessible by the Company at all times. 10
Q. How does the Company validate these customer 11
requirements have been met? 12
A. The Company has developed an initial on-site 13
inspection, that is updated from time to time, to verify that 14
the customer equipment installed matches the information 15
provided on the system verification form and that the 16
interconnection generally complies with the IEEE standards. 17
Additional Interconnection Requirements for Larger Systems 18
Q. Will the Company need to modify any of its existing 19
interconnection requirements for exporting customers if the 20
Commission approves the Company’s proposal to increase the 21
project eligibility cap for Schedule 84? 22
A. Yes. There are additional requirements necessary to 23
interconnect exporting systems larger than 100 kW safely and 24
reliably. Table 2 provides a summary of the additional 25
ELLSWORTH, DI 31
Idaho Power Company
interconnection requirements the Company proposes to include in 1
Schedule 68. 2
Table 2 3
Exporting System Interconnection Requirements 100 kW and Greater 4
5
As illustrated in Table 2, the Company proposes to revise 6
Schedule 68 to require the following: (1) inverter-based 7
generation 100 kW and greater will provide documentation 8
validating inverter settings; (2) for systems 500 kW and 9
greater, a power plant controller (or in the alternative, a 10
properly configured inverter) will be installed on the 11
customer’s side of the point of delivery; (3) for systems 3 MW 12
and greater, the existing uniform interconnection agreement and 13
requirements applicable to non-exporting systems larger than 3 14
MW will apply. 15
Q. Please explain why the Company will require the 16
customer using inverter-based generation to provide 17
documentation validating their inverter settings. 18
A. The larger inverter-based generation systems have a 19
greater potential to negatively impact the Company’s system if 20
not properly configured because of their relative size to the 21
local area load. In order to ensure safe and reliable operation 22
of the Company’s equipment and the service to our customers, it 23
Total Nameplate
Capacity
Inverter Settings
Documentation
Install Plant
Controller
Interconnection
Agreement
100kW - 500 kW
500kW - 3 MW
3 MW+
ELLSWORTH, DI 32
Idaho Power Company
is essential to verify that the larger inverter-based generation 1
systems are properly configured. An improperly configured 2
generation system could lead to power quality issues or damage 3
to equipment for other customers and on the Company’s system. 4
For these reasons, inverter-based generation systems larger than 5
100 kW will need to provide documentation of their inverter 6
settings. 7
Q. What is the rationale for the interconnection 8
requirement of a plant controller for systems 500 kW and 9
greater? 10
A. Pursuant to IEEE 1547-2018, Section 4.2 Reference 11
points of applicability, customer-generators operating 12
Distributed Energy Resources (“DERs”) in aggregate of 500 13
kilovolt-ampere (“kVA”) or greater, are responsible for 14
installing equipment required to monitor voltage, current, and 15
frequency on the customer’s side of the point of delivery. This 16
equipment measures data to calculate and to communicate required 17
operating settings to individual inverters and other devices to 18
control the generation facility output. In order to meet the 19
IEEE standard, the interconnection requirements dictate customer 20
generation facilities 500 kVA and larger be designed with a 21
power plant controller. If all power flows through a single 22
inverter, the inverter may be operated such that it is 23
equivalent to a power plant controller. 24
ELLSWORTH, DI 33
Idaho Power Company
Q. Please describe the interconnection requirement for 1
exporting systems 3 MW and greater. 2
A. Idaho Power proposes to require that exporting 3
systems 3 MW and greater include the same study and 4
communication requirements that are currently applicable to non-5
exporting systems larger than 3 MW under Schedule 68. These 6
requirements align with the interconnection requirements for 7
similar sized generator interconnections. 8
III. CONCLUSION 9
Q. Does the Company’s proposal for methods to value 10
the ECR and modify the Schedule 84 project eligibility cap meet 11
the Company’s primary objectives in this case? 12
A. Yes. I believe the methods described in my 13
testimony to value the ECR result in a fair and accurate 14
valuation of customers’ exported energy and provide for a 15
repeatable method for updating the ECR that will ensure timely 16
recognition of changing conditions on Idaho Power’s system and 17
the broader power markets. I also believe that the Company’s 18
proposed method reasonably balances accuracy with customer 19
understandability. Additionally, the proposed modification to 20
the Schedule 84 project eligibility cap concurrent with approved 21
changes to the compensation structure provides additional 22
flexibly and opportunities for customers to install on-site 23
generation. 24
// 25
ELLSWORTH, DI 34
Idaho Power Company
Q. Does this conclude your testimony? 1
A. Yes. 2
//3
ELLSWORTH, DI 35
Idaho Power Company
DECLARATION OF Jared L. Ellsworth 1
I, Jared L. Ellsworth, declare under penalty of perjury 2
under the laws of the state of Idaho: 3
1. My name is Jared L Ellsworth. I am employed by 4
Idaho Power Company as Transmission, Distribution & Resource 5
Planning Director in the Planning, Engineering & Construction 6
Department. 7
2. On behalf of Idaho Power, I present this pre-8
filed direct testimony and Exhibit Nos. 1-5 in this matter. 9
3. To the best of my knowledge, my pre-filed direct 10
testimony and exhibits are true and accurate. 11
I hereby declare that the above statement is true to the 12
best of my knowledge and belief, and that I understand it is 13
made for use as evidence before the Idaho Public Utilities 14
Commission and is subject to penalty for perjury. 15
16
SIGNED this 1st day of May 2023, at Boise, Idaho. 17
18
Signed: _______________________ 19
20
BEFORE THE
IDAHO PUBLIC UTILITIES COMMISSION
CASE NO. IPC-E-23-14
IDAHO POWER COMPANY
ELLSWORTH, DI
TESTIMONY
EXHIBIT NO. 1
ECR SUMMARY ECR Annual Update
Season ECR
Export Profile
Volume (kWh per kW)Annual 1,465
Capacity Contribution (%)Annual 8.76%
Export Credit Rate by Component (cents/kWh)
Energy On-Peak 8.59 ¢
Including integration and losses Off-Peak 4.91 ¢
Annual* 5.16 ¢
Generation Capacity On-Peak 11.59 ¢
Off-Peak 0.00 ¢
Annual* 0.79 ¢
Transmission & Distribution Capacity On-Peak 0.25 ¢
Off-Peak 0.00 ¢
Annual* 0.02 ¢
Total On-Peak 20.42 ¢
Off-Peak 4.91 ¢
Annual* 5.96 ¢
*Annual values provided for informational purposes only and reflect seasonal weighting for 12
months ending December 2022.
Note: On-Peak defined as June 15 - September 15, Monday - Saturday (exluding holidays), 3pm -
11pm. All other hours defined as Off-Peak.
Exhibit No. 1 Case No. IPC-E-23-14 J. Ellsworth, IPC
Page 1 of 4
Avoided Energy ECR Annual Update
On-Peak Off-Peak
Avoided Energy Calculation Update Update Units Description
ELAP - Weighted Average 84.60$ 49.84$ $/MWh
Plus: Line Loss Gross-up 4.23$ 2.19$ $ Exhibit No. 3 - Analysis of System Losses (March 2023)
Less: Integration Costs (2.93)$ (2.93)$ $/MWh Exhibit No. 4 - Idaho Power 2020 VER Integration Study
Avoided Energy Value 85.90$ 49.10$ $/MWh
Annual Energy Value 51.60$ 51.60$
Monthly Seasonal Energy Calculation
On/Off-Peak Month Value Energy $/MWh
Off-Peak 1 102,879$ 3,144 32.72$
Off-Peak 2 167,545$ 6,362 26.33$
Off-Peak 3 233,461$ 8,973 26.02$
Off-Peak 4 436,204$ 9,977 43.72$
Off-Peak 5 445,602$ 11,077 40.23$
Off-Peak 6 263,414$ 9,105 28.93$
On-Peak 6 57,053$ 1,624 35.14$
Off-Peak 7 385,929$ 6,750 57.17$
On-Peak 7 188,394$ 2,100 89.72$
Off-Peak 8 402,482$ 6,195 64.97$
On-Peak 8 165,264$ 1,767 93.52$
Off-Peak 9 474,169$ 7,779 60.96$
On-Peak 9 118,488$ 764 155.00$
Off-Peak 10 516,061$ 9,157 56.36$
Off-Peak 11 332,075$ 4,809 69.06$
Off-Peak 12 517,249$ 2,494 207.40$
Annual 4,806,268$ 92,076 52.20$
On-Peak 529,199$ 6,255 84.60$
Off-Peak 4,277,069$ 85,821 49.84$
Exhibit No. 1 Case No. IPC-E-23-14 J. Ellsworth, IPC
Page 2 of 4
Avoided Generation Capacity ECR Annual Update
Avoided Generation Capacity Calculation Update Units Description
Effective Load Carrying Capability 8.760% % 3-year rolling average ELCC (CY2020-2022)
(x) Nameplate Capacity 62.86 MW
Total Capacity Contribution 5.51 MW
(x) Levelized Fixed Cost of Avoided Resource 131.60$ $/kW-year 2021 Integrated Resource Plan - Appendix C, page 38
(x) kW to MW conversion 1,000 kW
(/) On-Peak Exports 6,255 MWh CY2022 real-time customer generation exports
On-Peak Avoided Generation Capacity Value 115.86$ $/MWh
Annual Generation Capacity Value 7.87$ $/MWh
Customer Generation Exports - ELCC & Maximum Output | Current Reliability & Capacity Assessment Tool (Historical Data)
Year - 2020
ELCC (MW)2
Maximum Output (MW)27
ELCC (%) 7.50%
Year - 2021
ELCC (MW)5
Maximum Output (MW)40
ELCC (%) 12.42%
Year - 2022
ELCC (MW)4
Maximum Output (MW)63
ELCC (%) 6.36%
3-Year Average 8.76%3-year rolling average ELCC (CY2020-2022)
Exhibit No. 1 Case No. IPC-E-23-14 J. Ellsworth, IPC
Page 3 of 4
Avoided Transmission & Distribution Capacity ECR Annual Update
Avoided T&D Capacity Calculation Update Units Description
Distribution Capacity Savings 307,263$ $ Exhibit No. 2 - Transmission and Distribution Avoided Capacity
Plus: Transmission Capacity Savings - $ Exhibit No. 2 - Transmission and Distribution Avoided Capacity
Total T&D Capacity Savings 307,263$ $
(/) Project Years 20 years Exhibit No. 2 - Transmission and Distribution Avoided Capacity
Annual T&D Capacity Savings 15,363$ $/year
(/) On-Peak Exports 6,255 CY2022 real-time customer generation exports
On-Peak T&D Capacity Value 2.46$ $/MWh
Annual Generation Capacity Value 0.17$ $/MWh
Exhibit No. 1 Case No. IPC-E-23-14 J. Ellsworth, IPC
Page 4 of 4
BEFORE THE
IDAHO PUBLIC UTILITIES COMMISSION
CASE NO. IPC-E-23-14
IDAHO POWER COMPANY
ELLSWORTH, DI
TESTIMONY
EXHIBIT NO. 2
SEE ATTACHED SPREADSHEET
BEFORE THE
IDAHO PUBLIC UTILITIES COMMISSION
CASE NO. IPC-E-23-14
IDAHO POWER COMPANY
ELLSWORTH, DI
TESTIMONY
EXHIBIT NO. 3
SEE ATTACHED SPREADSHEET
BEFORE THE
IDAHO PUBLIC UTILITIES COMMISSION
CASE NO. IPC-E-23-14
IDAHO POWER COMPANY
ELLSWORTH, DI
TESTIMONY
EXHIBIT NO. 4
Analysis of System Losses Idaho Power Company
ANALYSIS OF SYSTEM LOSSES
In
Idaho Power Company
Prepared by:
Jackson Daly
Andrés Valdepeña Delgado
System Planning Department
March 2023
Exhibit No. 4 Case No. IPC-E-23-14 J. Ellsworth, IPC
Page 1 of 17
Analysis of System Losses Idaho Power Company
Contents
Executive Summary ....................................................................................................................................... 3
Introduction .................................................................................................................................................. 4
System Level Description .............................................................................................................................. 4
Transmission System ................................................................................................................................. 4
Distribution System ................................................................................................................................... 4
Stations Level ............................................................................................................................................ 5
Primary Level............................................................................................................................................. 5
Secondary Level ........................................................................................................................................ 5
Energy Loss Coefficient Calculations ............................................................................................................. 6
Transmission Level Energy Losses ............................................................................................................. 6
Distribution Substation Level Energy Losses ............................................................................................ 7
Distribution Level Energy Losses ............................................................................................................... 8
Distribution Line Transformer Losses ....................................................................................................... 8
Primary-Secondary Distribution Losses Split ............................................................................................ 8
Losses Comparison with FERC Form 1 ...................................................................................................... 9
Peak Loss Coefficients ............................................................................................................................. 11
Avoidable Losses by On-Site Customer Generation ................................................................................... 12
Appendix A: 2012 Energy Losses Data Sources ................................................................................ 14
Appendix B: 2012 Peak Losses Data Sources.................................................................................... 16
Appendix D: Reconciliation with FERC Form 1 ................................................................................ 17
Exhibit No. 4 Case No. IPC-E-23-14 J. Ellsworth, IPC
Page 2 of 17
Analysis of System Losses Idaho Power Company
Executive Summary
This study presents the peak and energy loss coefficients for the Idaho Power delivery system. The
analysis was conducted using 2022 data. The delivery system was broken down into four different
system levels, including:
• Transmission: Includes voltage levels between 46 kV and 500 kV
• Distribution Stations: Includes distribution station transformers
• Distribution Primary: Includes distribution lines and facilities between 12.47 kV and 34.5 kV
• Distribution Secondary: Includes distribution service lines and distribution line transformers
The losses documented in this study represent the physical losses that occurred on the Idaho Power
delivery system facilities. Application of the calculated loss coefficients is limited to loads served from
Idaho Power Company facilities. The peak loss coefficients were calculated based on data from the
system peak hour in 2022, which occurred on July 14th, 2022, at 7:00 PM.
The study incorporated various methods to calculate the losses at different voltage levels. For the 161
kV and above transmission system, current readings and resistance from the lines were used to
determine the losses. For the 138 kV transmission system, the losses were determined by calculating the
total inputs into the 138 kV system and subtracting the outputs, leaving the difference as the losses in
the 138 kV system. For the sub-transmission system, electric current or power and resistance readings
were used to determine losses. The total transformer losses were determined by adding the winding
and core losses. The distribution system losses were determined as the difference between the input to
the distribution system and the output, where the output of the distribution system is the end-use
customer usage obtained from the Advance Metering Infrastructure (“AMI”) and the industrial and
commercial usage, MV90 database.
The individual system loss coefficients are determined as the system level inputs, divided by the system
level outputs. The loss coefficients used at each delivery point in the system are calculated as the
product of the individual level loss coefficients. The resulting coefficients for the 2022 study are
summarized in Table 1.
System Level Energy Loss Coefficient Peak Loss Coefficient
Transmission 1.029 1.037
Distribution Station 1.036 1.042
Distribution Primary 1.051 1.056
Distribution Secondary 1.076 1.076
Table 1: Delivery Point Loss Coefficients
Exhibit No. 4 Case No. IPC-E-23-14 J. Ellsworth, IPC
Page 3 of 17
Analysis of System Losses Idaho Power Company
Introduction
Loss coefficients are the ratio of the system input required to provide a given output at a particular
system level. Individual loss coefficient for each system level relates the input and the output by (1):
𝐿𝑜𝑠𝑠 𝐶𝑜𝑒𝑓𝑓𝑖𝑐𝑖𝑒𝑛𝑡=𝐿𝑒𝑣𝑒𝑙 𝐼𝑛𝑝𝑢𝑡
𝐿𝑒𝑣𝑒𝑙 𝑂𝑢𝑡𝑝𝑢𝑡=1 +𝐿𝑒𝑣𝑒𝑙 𝐿𝑜𝑠𝑠𝑒𝑠
𝐿𝑒𝑣𝑒𝑙 𝑂𝑢𝑡𝑝𝑢𝑡
The system loss coefficient is obtained by multiplying all the upstream system level coefficients
together.
System Level Description
The Idaho Power delivery system was split into four categories: transmission, distribution stations,
distribution primary, and distribution secondary. The system inputs and outputs for each level are
described below.
Transmission System
The transmission level includes losses for all facilities and lines from 46 kV up through 500 kV. Losses
from the Generation Step-Up (“GSU”) transformers and transmission tie-bank transformers are included
in the transmission level. Customer owned facilities at the transmission level are not included.
Transmission level inputs consist of the following:
+ Idaho Power Generation
+ Power Purchases/Exchanges
+ Customer Owned Generation Connecting to Transmission Lines
+ Wheeling Transactions
Transmission level outputs consist of the following:
- High Voltage Sales
- Power Exchanges*
- Wheeling Transactions
- Output to Distribution Stations
The exchanges outputs are adjusted to remove the scheduled losses for the Idaho Power share of losses
in the jointly owned Bridger-Idaho and Valmy-Midpoint transmission systems. FERC From 1 includes the
Bridger and Valmy scheduled losses as exchanged out. The calculated losses in this study include the
Idaho Power share of losses on the Bridger and Valmy systems as transmission level losses.
Distribution System
The distribution system consists of all equipment operating at 35 kV and below. This accounts for all
substation transformers, distribution lines, and distribution transformers. The distribution system can be
split into 3 different levels: stations, primary and secondary. These different levels are chosen to account
for the losses most accurately at the different points of delivery.
Exhibit No. 4 Case No. IPC-E-23-14 J. Ellsworth, IPC
Page 4 of 17
Analysis of System Losses Idaho Power Company
Stations Level
Stations level consists only of the substations servicing the distribution system (transformers with a low
voltage side of 7 – 35 kV).
Station level inputs consist of the following:
+ Transmission System Outputs
Station level outputs consist of the following:
- Direct Sales
- Wheeling Transactions
Although this level has no additional inputs, it is chosen as there are several customers who are served
directly from the substation.
Primary Level
The primary level consists of all the primary distribution power lines. Primary lines being lines operated
between 7 - 35 kV.
Primary level inputs consist of the following:
+ Distribution Stations Outputs
+ PURPA/Customer Generation
Primary level outputs consist of the following:
- Customer Sales
- Wheeling Transactions
The primary distribution level contains a large amount of generation under the Public Utility Regulatory
Policies Act (“PURPA”) and customers with on-site generation and customers who connect directly to
the distribution primary level.
Secondary Level
The secondary level consists of all equipment operating at a service voltage. This includes distribution
transformers and distribution lines operating at a service voltage.
Secondary level inputs consist of the following:
+ Primary Level Outputs
+ Net Metering/Customer Generation
Secondary level outputs consist of the following:
- Customer Sales
- Idaho Power Internal use
- Street Lighting/ Unbilled
- Wheeling Transactions
Exhibit No. 4 Case No. IPC-E-23-14 J. Ellsworth, IPC
Page 5 of 17
Analysis of System Losses Idaho Power Company
Customer with on-site generation are inputs to the secondary level and come from both rooftop solar
and small hydro generation.
Energy Loss Coefficient Calculations
Table 8 shows the total system flow diagram for the 2022 energy losses. The table outlines each system
level’s input and output as well as the total energy losses (MWh) and loss coefficient. The transmission
level output (MWh) to the distribution station level is calculated by subtracting the remaining output
and calculated losses from the transmission level inputs
Transmission Level Energy Losses
For the 500 – 161 kV, 69 kV, and 46 kV voltage levels, the transmission losses were calculated using
Ohm’s Law where current readings were available (2).
𝑃𝐿𝑜𝑠𝑠=𝐼2 ⋅𝑅
Where 𝐼 is the current flowing in a particular transmission line in Amperes and 𝑅 is the resistance of the
transmission line in Ohms.
For the lines where current readings were unavailable, the apparent power (S) in MVA and voltage (V)
readings were used to calculate the current using the equation below (3).
𝐼=𝑉
𝑆
Due to the complexity of the 138-kV system, the losses were calculated by obtaining all the energy into
the 138-kV system and subtracting all the energy leaving the 138-kV system.
The summary of losses for the different voltage levels in the transmission system are shown in Table 2:
Loss Type
Voltage Level
500kV 345kV 230kV 161kV 138kV
(Stations)
138kV 69kV 46kV
Lines 23,400 214,741 224,711 3,210 128,558 - 48,061 23,037
Core 7,148 9,909 39,915 990 9,088 36,450 9,210 5,827
Winding 6,005 3,504 18,393 6,222 4,931 35,175 7,065 3,961
Total Losses 36,553 228,154 283,019 10,422 142,577 71,625 64,336 32,825
Table 2: Type of Losses (MWh) by Voltage Level
The losses in the transmission transformers, generator step-up transformers and tie-banks, were
calculated by adding the two components of the losses in a transformer, the winding losses, and the
core losses.
Exhibit No. 4 Case No. IPC-E-23-14 J. Ellsworth, IPC
Page 6 of 17
Analysis of System Losses Idaho Power Company
The winding losses, also called copper losses, were calculated using (4):
𝐿𝑜𝑠𝑠𝑒𝑠 (𝑀𝑊ℎ)= ∑(𝐻𝑜𝑢𝑟𝑙𝑦 𝑈𝑠𝑎𝑔𝑒)2 ⋅𝑅𝑝𝑢
100
𝑁
𝑛=1
Where 𝑅𝑝𝑢 is the total per-unit resistance on a 100 MVA base and 𝐻𝑜𝑢𝑟𝑙𝑦 𝑈𝑠𝑎𝑔𝑒 is the average hourly
usage on the transformer in MWh.
The core losses were obtained using records from the Idaho Power Apparatus department “no-load
losses” records. It was assumed that the transformers were energized the entire year. The total core
losses for each transformer were calculated using (5):
𝐶𝑜𝑟𝑒 𝐿𝑜𝑠𝑠𝑒𝑠 (𝑀𝑊ℎ)=𝑁𝐿𝐿⋅8760
1000
Where 𝑁𝐿𝐿 are the no-load losses in kWh for each transformer, and 8760 is the hours in the year 2022.
The total losses for the transmission level were found by adding the losses for the transmission lines and
the losses for the transmission transformers. The total losses for the transmission system are shown
below, broken down by voltage level and component type Table 3.
Transmission Losses
By Voltage
Transmission Losses
By Component
500kV 36,553 Lines 665,718
345kV 228,154 Core 67,050
230kV 283,019 Winding 39,055
161kV 10,422 Total 771,823
138kV 142,577
69kV 48,061
46kV 23,037
Total 771,823
Table 3: Transmission Losses (MWh) Breakdown
Distribution Substation Level Energy Losses
The distribution station losses were found by calculating the losses in the substation distribution
transformers for the calendar year 2022. Distribution transformers are classified, in this study, as any
transformer with a secondary voltage of 35-kV, 25-kV, or 12.5kV. The losses in other station apparatus
equipment and bus are assumed to be negligible.
The losses in the station transformer were calculated using the same method used to calculate the
losses in the transmission transformers using (3) and (4). For the few transformers that had no metering
data available in Idaho Power’s PI data custodian, the MV90 data was used. The total losses in the
distribution stations are broken down by both voltage level and component type are shown in Table 4.
Exhibit No. 4 Case No. IPC-E-23-14 J. Ellsworth, IPC
Page 7 of 17
Analysis of System Losses Idaho Power Company
Stations Losses
By Voltage
Stations Losses
By Component
500kV - Lines -
345kV - Core 51,487
230kV - Winding 46,201
161kV - Total 97,688
138kV 71,625
69kV 16,275
46kV 9,788
Total 97,688
Table 4: Station Losses (MWh) Breakdown
Distribution Level Energy Losses
The losses in the distribution level were determined by comparing the input to the system (feeder meter
data) to the output (customer billing data). Losses were inputs (feeder meter data) minus outputs
(customer billing data).
Distribution Line Transformer Losses
The distribution system losses can be separated into primary distribution and secondary distribution
losses. The distribution losses can be split between line and transformer losses. The split was done by
taking the average losses of the 138-k, 69-kV, and 46-kV systems as a proxy and determining what
proportion of those losses were line losses and which were transformer losses. These proportions were
then applied to the adjusted distribution losses to determine the distribution line losses and distribution
transformer losses. The results of this calculation can be seen in Table 5 below.
Line vs Transformer losses 2022 System Losses
Line Losses 316,822 Avg Line Loss 64%
Transformer losses 178,213 Avg Transformer Loss 36%
Total Distribution Losses 495,035
Table 5: Line vs Transformer Losses (MWh)
Primary-Secondary Distribution Losses Split
The split between the distribution primary and secondary lines losses was determined using the wire
milage for the distribution primary and secondary systems. The line mileage was obtained from the form
TAX650; the total distribution wire milage was found by adding up the total wire milage for the 12.5-kV,
25-kV, and 34.5-kV systems. From the TAX671 form, the primary line milage can be found broken down
by number of phases; the mile milage was converted to wire mileage by multiplying it by the number of
phases. The result is the total primary wire mileage which we can subtract from the total distribution
wire mileage to find the secondary wire mileage.
Exhibit No. 4 Case No. IPC-E-23-14 J. Ellsworth, IPC
Page 8 of 17
Analysis of System Losses Idaho Power Company
Using the final wire mileage, it was determined that the primary lines make up 68% of the total wire
mileage and the secondary lines make up the other 32%. These percentages can then be applied to the
total distribution line losses to determine the primary and secondary specific line losses. These
calculations can be seen in Table 6 below.
Primary vs Secondary Losses Distribution Wire Mileage
Primary Line Losses 215,080 12.5kV 50,974.12
Secondary Line Losses 101,743 25kV 1,377.87
Total Line Losses 316,822 34.5kV 16,797.35
Primary Losses 215,080 Total Line Mileage 69,149.34
Secondary Losses 279,955 Primary Line Mileage
Total Distribution Losses 495,035 1 – Phase 13,250.97
2 – Phase 928.81
3 – Phase 10,611.49
Primary Wire Mileage 46,943.06
Secondary Wire Mileage 22,206.28
Total Wire Mileage 69,149.34
Table 6: Distribution Losses (MWh) Breakdown
The primary distribution losses consist only of the primary line losses, the total losses for the primary
level is 214,985 MWh. The secondary distribution losses can be found by adding the distribution
transformer losses from Table 5 and the secondary line losses calculated above in Table 6, resulting in
279,955 MWh of losses for the secondary distribution level.
Losses Comparison with FERC Form 1
The losses obtained in the distribution system were added to the losses calculated from the levels above
and compared to the FERC Forum 1 losses. Idaho Power collects hourly data via SCADA for all generation
above 3 MW, for generation under the 3 MW limit there is no SCADA data being collected creating a
mismatch on the total losses calculated via FERC Form 1 and the losses calculated in this study. To adjust
for the generation without SCADA, the losses were adjusted in the distribution system to match the
total losses reported in FERC Form 1. This calculation can be seen in Table 7 below.
Calculated Distribution Losses FERC Forum 1 Comparison
Distribution Input 15,619,939 FERC Total Energy 18,376,323
Distribution Output 15,120,270 FERC Forum 1 Losses 1,238,735
Distribution Losses 499,669 Bridger/Valmy Losses 125,811
Missing Losses (4,634) Total FERC Losses 1,364,546
Corrected Losses 495,035 Calculated Losses 1,369,180
Adjusted Losses (4,634)
Table 7: Calculated Losses (MWh) Correction
Exhibit No. 4 Case No. IPC-E-23-14 J. Ellsworth, IPC
Page 9 of 17
Analysis of System Losses Idaho Power Company
Loss Coefficients Tables
Tables 8 and 9 contain the MWh losses in each of the level as well as the inputs and output to each
level. Table 8 shows the energy coefficients over the entire calendar year 2022 whereas Table 9 shows
the peak coefficients during the peak day in 2022.
2022 Energy Loss Coefficients Table - Wheeling Included (Values in MWh)
Transmission Inputs Loss Coefficients Losses Transmission Outputs
Power Supply 11,325,243 Transmission 1.029 771,823 Retail Sales 151,444
Utility purchases 4,394,440 High Volt 1,318,132
PURPA/Cust Gen 1,950,434 Wheeling 9,114,526
Exchange IN 27,768 Exchange OUT 0
Wheeling IN 9,325,825
Total 27,023,710 Delivery Point Coefficient 1.029 771,823 Total 10,584,102
Stations Inputs Distribution Stations 1.006 97,688 Stations Outputs
From Transmission 15,667,785 Direct Sales 946,593
Wheeling 91,552
Total 15,667,785 Delivery Point Coefficient 1.036 869,511 Total 1,038,145
Primary Inputs Distribution Primary 1.014 215,080 Primary Outputs
From Stations 14,531,952 Sales 3,067,827
PURPA/Cust Gen 805,834 Wheeling 656
Total 15,337,786 Delivery Point Coefficient 1.051 1,084,591 Total 3,068,483
Secondary Inputs Distribution Secondary 1.024 279,955 Secondary Outputs
From Primary 12,054,223 Sales 11,704,706
NET Metering 92,076 Wheeling 117,676
Street lighting 43,961
Total 12,146,929 Total 1.076 1,364,546 Total 11,866,343
Table 8: 2022 Energy Loss (MWh) Coefficients Table
Exhibit No. 4 Case No. IPC-E-23-14 J. Ellsworth, IPC
Page 10 of 17
Analysis of System Losses Idaho Power Company
Peak Loss Coefficients
An identical method to the annual losses coefficients was used in calculating the peak hour loss
coefficients. For the calculated losses, the same equations were used but only for the data from July 14th
at 7:00 PM. The inputs to the system were determined with the use of historical PI data from the same
hour, along with MV90 hourly data. Some aspects were determined to be 0 or small enough to not
influence the end results and were excluded to simplify the calculation. The results of this peak hour
analysis are shown in Table 9 below.
2022 Peak Loss Coefficients Table - Wheeling Included (Values in MWh)
Transmission Inputs Loss Coefficients Losses Transmission Outputs
Power Supply 1,869 Transmission 1.037 181 Retail Sales 19
Utility purchases 1,500 High Volt 0
PURPA/Cust Gen 853 Wheeling 752
Wheeling IN 804
Total 5,026 Delivery Point Coefficient 1.037 181 Total 771
Stations Inputs Distribution Stations 1.005 20 Stations Outputs
From Transmission 4,074 Direct Sales 108
Wheeling 15
Total 4,074 Delivery Point Coefficient 1.042 201 Total 123
Primary Inputs Distribution Primary 1.013 55 Primary Outputs
From Stations 3,931 Sales 404
PURPA/Cust Gen 365 Wheeling 0
Total 4,296 Delivery Point Coefficient 1.056 256 Total 404
Secondary Inputs Distribution Secondary 1.019 72 Secondary Outputs
From Primary 3,837 Sales 3,765
Total 3,837 Total 1.076 328 Total 3,765
Table 9: 2022 Peak Loss (MWh) Coefficients Table
Exhibit No. 4 Case No. IPC-E-23-14 J. Ellsworth, IPC
Page 11 of 17
Analysis of System Losses Idaho Power Company
Avoidable Losses by On-Site Customer Generation
Customers with on-site generation could avoid some of the losses previously discussed in this report.
However, there are losses, such as transformer core losses, that are not a function of load and will not
be able to be avoided by customers with on-site generation
To determine the avoidable losses from customers with on-site generation, the losses due to
transformer core-losses and distribution secondary were removed from the calculation and new
coefficients were calculated. The avoidable losses were separated into two different periods, an on-peak
period that covers June 15th to September 15th from 3:00pm to 11:00pm excluding Sundays and holidays
and an off-peak period that cover the rest of the hours in the year.
Previously, the loss coefficients were determined for the entire year and for the peak hour. In order to
determine the coefficients for the on-peak season, the hourly data from 138-kV system was used as
proxy to modify the peak and energy calculations. The 138-kV system was chosen due to having all
hourly data available and being a better representation on the Company loading at any given time.
The peak losses were modified to capture the load variability (and losses) that occurred from June 15th
to September 15th. Table 10 shows the adjustments to the peak coefficients to determine the on-peak
avoidable losses.
2022 On-Peak Loss Coefficients Table - Adjusted VODER (Values in MWh)
Transmission Inputs Loss Coefficients Losses Transmission Outputs
Power Supply 1,869 Transmission 1.034 164 Retail Sales 19
Utility purchases 1,500 High Volt 0
PURPA/Cust Gen 853 Wheeling 752
Exchange IN 0 Exchange 0
Wheeling IN 804
Total 5,026 Delivery Point Coefficient 1.034 164 Total 771
Stations Inputs Distribution Stations 1.003 14 Stations Outputs
From Transmission 4,091 Direct Sales 108
Wheeling 15
Total 4,091 Delivery Point Coefficient 1.037 178 Total 123
Primary Inputs Distribution Primary 1.012 52 Primary Outputs
From Stations 3,954 Sales 404
PURPA/Cust Gen 365 Wheeling 0
Total 4,319 Delivery Point Coefficient 1.050 230 Total 404
Secondary Inputs Distribution Secondary 1.000 Secondary Outputs
From Primary 3,863 Sales 3,863
Total 3,863 Total 1.050 230 Total 3,863
Table 10: Adjusted VODER Energy Losses (MWh) Coefficients Table
Exhibit No. 4 Case No. IPC-E-23-14 J. Ellsworth, IPC
Page 12 of 17
Analysis of System Losses Idaho Power Company
Similarly, the off-peak coefficients were modified to remove the on-peak data and obtained an off-peak
coefficient. Table 11 shows the modifications to the off-peak coefficients.
2022 Off-Peak Loss Coefficients Table - Adjusted VODER (Values in MWh)
Transmission Inputs Loss Coefficients Losses Transmission Outputs
Power Supply 11,325,243 Transmission 1.026 697,937 Retail Sales 150,532
Utility purchases 4,394,440 High Volt 1,318,132
PURPA/Cust Gen 1,945,752 Wheeling 9,114,526
Exchange IN 53,368 Exchange 25,600
Wheeling IN 9,325,825
Total 27,044,628 Delivery Point Coefficient 1.026 697,937 Total 10,608,790
Stations Inputs Distribution Stations 1.003 45,753 Stations Outputs
From Transmission 15,737,901 Direct Sales 946,593
Wheeling 91,552
Total 15,737,901 Delivery Point Coefficient 1.029 743,690 Total 1,038,145
Primary Inputs Distribution Primary 1.014 212,900 Primary Outputs
From Stations 14,654,003 Sales 3,042,892
PURPA/Cust Gen 805,968 Wheeling 656
Total 15,459,971 Delivery Point Coefficient 1.044 956,589 Total 3,043,548
Secondary Inputs Distribution Secondary 1.000 Secondary Outputs
From Primary 12,203,524 Sales 12,203,524
Total 12,203,524 Total 1.044 956,589 Total 12,203,524
Table 11: Adjusted VODER Peak Losses (MWh) Coefficients Table
The avoidable losses coefficients are shown in Table 12 below.
VODER
System Level
Off-Peak Loss
Coefficient
On- Peak Loss
Coefficient
Transmission 1.026 1.034
Distribution Station 1.029 1.037
Distribution Primary 1.044 1.050
Distribution Secondary 1.044 1.050
Table 12: Adjusted VODER Delivery Point Loss Coefficients
Exhibit No. 4 Case No. IPC-E-23-14 J. Ellsworth, IPC
Page 13 of 17
Analysis of System Losses Idaho Power Company
Appendix A: 2012 Energy Losses Data Sources
Transmission
Inputs
Value
(MWh) Data Source Notes
Power Supply
Generation 11,325,243
FERC Form 1 p 401a line
9
Utility
Purchases 4,394,440
FERC Form 1 p 326.8 -
327.12 col g (Subset of
Utility Purchases FERC
Form 1 p 401a line 10)
OATT Power purchases from
utilities/entities not directly connected to
IPC system
PURPA/Cust
Gen 1,950,434
FERC Form 1 pp 326-
327.7 col g (Subset of
Utility Purchases FERC
Form 1 p 401a line 10)
Power purchased from non-IPC owned
generation connected to IPC transmission
system
Exchange In 27,768
FERC Form 1 p 401a line
12
Details on FORM 1 p 326.12-327.13
See "FF1 326-327.xlsx"
Wheeling In 9,325,825
FERC Form 1 p 401a line
16 File: “Wheeling Form 1 Detail.xlsx”
Transmission
Outputs
High Voltage
Sales 1,318,132
FERC Form 1 p 401a line
24 Details on Form 1 p 311
Exchange Out 25,600
FERC Form 1 p 401a line
12
Details on FORM 1 p 326.12-327.13
See "FF1 326-327.xlsx"
Wheeling Out 9,114,526
FERC Form 1 p 401a line
17 File: “Wheeling Form 1 Detail.xlsx”
Retail
Transmission
Sales 151,444 FERC Forum 1 – p 304
FERC Forum 1 – p 304
Rate 9T, 19T, and Unbilled Rev. Large
Distribution
Station
Outputs
Direct Station
Sales 946,593 FERC Forum 1 – p 304
FERC Forum 1 – p 304
Special Contracts
Wheeling Out 91,552 Operation Data File: “Wheeling Form 1 Detail.xlsx”
Distribution
Primary Inputs
PURPA 805,834
PURPA gen connected to
IPC Primary distribution
system from FERC Form
1 p 326-327.7 col g
Subset of Utility Purchases
FERC Form 1 p 401a line 10
Total from p 401a line 10 is split by system
level on spreadsheet:
“FF1 326-327.xlsx”
Exhibit No. 4 Case No. IPC-E-23-14 J. Ellsworth, IPC
Page 14 of 17
Analysis of System Losses Idaho Power Company
Distribution
Primary
Outputs
Direct Primary
Sales 3,067,827 FERC Forum 1 – p 304
FERC Forum 1 – p 304
Rate 09P, 19P, 08, and Unbilled Rev. Small
Wheeling Out 656 Operations Data File: “Wheeling Form 1 Detail.xlsx”
Distribution
Secondary
Inputs
Net Met/Ore
Solar 92,076 Operations Data “IPC_Exports_by_Class.xlsx”
Distribution
Secondary
Outputs
Distribution
Sales 11,704,706 FERC Forum 1 – p 304
FERC Forum 1 – p 304
07, 09S, 19S, 24S, Total Billed Residential
Sales – Rate 15., and Unbilled Rev.
Street Lighting 43,961 FERC Forum 1 – p 304
FERC Forum 1 – p 304
Rate 15, 40, and TOTAL Billed Public Street
and Highway Lighting
Wheeling Out 117,676 Operations Data File: “Wheeling Form 1 Detail.xlsx”
Exhibit No. 4 Case No. IPC-E-23-14 J. Ellsworth, IPC
Page 15 of 17
Analysis of System Losses Idaho Power Company
Appendix B: 2012 Peak Losses Data Sources
Transmission
Inputs
Value
(MW) Data Source Notes
Power Supply
Generation 1,869 Pi
Utility Purchases 1,500 Pi see file "Peak_day_data.xlsx”
PURPA/Cust Gen 853 Pi
Wheeling In 804 Operations data on peak hour File: “Wheeling Forum 1 Detail.xlsx”
Transmission
Outputs
Retail Sales 19
Transmission customer sales
from MV90 data: filename
"MV90 2022 8760.xlsx”
Wheeling Out 752 Pi File: “Wheeling Forum 1 Detail.xlsx”
Distribution
Station Outputs
Direct Station
Sales 108
Sales from MV90 data:
filename "MV90 2022
8760.xlsx”
Wheeling Out 15 Pi File: “Wheeling Forum 1 Detail.xlsx”
Distribution
Primary Inputs
PURPA 365 Pi
Distribution
Primary Outputs
Direct Primary
Sales 404
Sales from MV90 data:
filename "MV90 2022
8760.xlsx”
Distribution
Secondary
Outputs
Wheeling Out 36.9 Pi File: “Wheeling Forum 1 Detail.xlsx”
Exhibit No. 4 Case No. IPC-E-23-14 J. Ellsworth, IPC
Page 16 of 17
Analysis of System Losses Idaho Power Company
Appendix D: Reconciliation with FERC Form 1
The data used in the development of the energy loss coefficients in this report is consistent with that
reported in the 2022 FERC Form 1, page 401a. Values used in Figure 1 are reconciled with values in 2022
FERC Form 1 below.
System Losses
Item Figure 1
MWh
2012 FERC
Form 1 MWh
Comment
Total System Losses 1,364,546 1,238,725 Form 1, pg 401a, line 27
Adjustment for Bridger Loss
Transactions
124,135 Bridger Loss transactions counted as
system outputs in Form 1 (part of
total in Form 1, pg 401a, line 13)
Adjustment for Valmy Loss
Transactions
1,676 Valmy Loss transactions counted as
system outputs in Form 1 (part of
total in Form 1, pg 401a, line 13)
Adjusted Total 1,364,546 1,364,180
The ratio of Adjusted FERC Form 1 losses to Figure 1 losses is 99.66%. Reasons for the small discrepancy
may include non-uniformity between the calculation method used to determine transmission losses on
the Bridger and Valmy subsystems in this study versus the calculation method used to determine the
actual loss transactions and estimation methods used where small amounts of data were missing in the
tabulation of individual level losses.
Exhibit No. 4 Case No. IPC-E-23-14 J. Ellsworth, IPC
Page 17 of 17
BEFORE THE
IDAHO PUBLIC UTILITIES COMMISSION
CASE NO. IPC-E-23-14
IDAHO POWER COMPANY
ELLSWORTH, DI
TESTIMONY
EXHIBIT NO. 5
Variable Energy Resource
Idaho Power Company
December, 2020
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 1 of 88
Appendix 4.17 Idaho Power 2020 VER Integration Study
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 1 of 88
ii | P a g e
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 2 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 2 of 88
© 2020 Copyright. All Rights Reserved.
Energy and Environmental Economics, Inc.
44 Montgomery Street, Suite 1500
San Francisco, CA 94104
415.391.5100
www.ethree.com
Variable Energy Resource
Idaho Power Company
December 2020
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 3 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 3 of 88
iv | P a g e
Executive Summary
Energy and Environmental Economics, Inc. (E3) was retained by Idaho Power to
investigate the integration cost of variable energy resources in Idaho Power’s
service territory. These costs are incurred due to increased dispatchable unit
cycling (from increased unit stops and starts; increased load following ramping)
and imperfect unit commitment and dispatch (resulting in higher average thermal
unit heat rates and/or lower net market earnings); and, in cases in which
economic VER curtailment is allowed, increased curtailment costs. E3’s analysis
calculates both average and incremental integration costs on a $/MWh basis,
using the proposed unit additions and retirements to Idaho Power’s 2023 system
as a base case.
The study examines eleven cases of potential future VER builds in Idaho Power
territory. These cases are illustrated below in Table ES1. These include high wind
and high solar builds; cases in which low, average and high annual hydro energy
budgets are simulated; cases in which there is solar plus investment tax credit
(ITC)-enabled storage; cases in which solar can be economically curtailed; and a
case in which a planned unit retirement at the Bridger coal plant is not in effect
in 2023. As can be seen in Table ES1, the overall incremental integration costs
were found to range from $0.64/MWh-$4.65/MWh. Generally, these costs are
lower than those in the 2018 Idaho Power VER Integration Analysis, although it is
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 4 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 4 of 88
notable that the method of deriving integration costs was substantially different
in the last study.1
Table ES1: Case Description and Results Summary
E3 believes that the integration costs in this study are lower than previous studies
primarily due to four factors: 1) Reduced need for modeled ancillary services, 2)
The fact that the remaining 2023 coal fleet is modeled as must-run (i.e. its
commitment decisions are not affected by VER penetration), 3) Access to the
Energy Imbalance Market (EIM) makes it easier to use market transactions to
1https://docs.idahopower.com/pdfs/AboutUs/PlanningForFuture/wind/VariableEnergyResourceIntegrationAn
alysis.pdf
Case Description
First
Bridger
Unit
Existing
2023
Wind
Capacity
(MW)
Hydro
Year
New
2023
Solar
Build
(MW)
New
2023
Wind
Build
(MW)
Total
Integration
Cost
1 Base 2023 Case Retired 561 728 Normal 0 0 No 0 $ 2.93
2
Base Case + First
Bridger Unit Online Online 561 728 Normal 0 0 No 0 $ 3.61
3 High Solar Retired 561 728 Normal 794 0 No 0 $ 3.86
4
High Solar, Low
Hydro Retired 561 728 Low 794 0 No 0 $ 4.55
5 High Wind Retired 561 728 Normal 0 669 No 0 $ 0.77
6
High Solar, High
Wind Retired 561 728 Normal 794 669 No 0 $ 2.46
7
Existing Solar Base
Case Retired 310 728 Normal 0 0 No 0 n/a
8 Hydro Retired 561 728 High 794 0 No 0 $ 4.65
9
High Solar + 200
MW Storage Retired 561 728 Normal 794 0 No 200 $ 0.64
10
High Solar + 400
MW Storage Retired 561 728 Normal 794 0 No 400 $ 0.93
11 Curtailable Solar Retired 561 728 Normal 794 0 Yes 0 $ 3.13
Proposed
Existing
2023
Solar
Capacity
(MW)
Amount of New
VER Added to
Existing 2023
Builds
New
Solar-
Coupled
4-hr Li-
Ion
Battery
Build
(MW)
Can New
Solar be
Curtailed?
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 5 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 5 of 88
vi | P a g e
integrate VERs (the EIM was not included in the previous study) and 4) Allowing
additional system flexibility, in some cases (e.g. from batteries).
The integration costs calculated in this current effort specifically do not consider
fuel savings or capacity contributions from, nor do they consider the capital costs
of new VERs. Therefore, this VER integration cost study serves as a valid basis for
calculating integration costs but may not include all economic and operational
factors required to integrate VERs on the Idaho Power system.
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 6 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 6 of 88
Table of Contents
Executive Summary ...................................................................................... iv
1 Introduction ............................................................................................ 1
1.1 Motivation and Background ................................................................ 1
2 Methodology ........................................................................................... 2
2.1 Calculating VER Integration Costs .................................................... 2
2.2 Production Cost Modeling ................................................................... 6
2.3 Reserve Modeling ................................................................................. 8
3 Data Collection, Processing and Methods ....................................... 10
3.1 PLEXOS Modeling .............................................................................. 10
3.1.1 Load Profiles, VER Profiles and Dispatchable
Generation Fleet ................................................................. 10
3.1.2 External Market Representation ...................................... 13
3.2 RESERVE Modeling........................................................................... 16
3.2.1 Derivation of 2023 VER Profiles ...................................... 16
3.2.2 Deriving Reserves Components ...................................... 18
3.3 Case Matrix .......................................................................................... 19
4 Results ................................................................................................... 22
4.1 RESERVE Outputs ............................................................................. 22
4.1.1 Annual Average results ..................................................... 22
4.1.2 Detailed Reserve Results ................................................. 26
4.2 2019 PLEXOS to Historical Case Benchmarking ......................... 31
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 7 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 7 of 88
viii | P a g e
4.3 2023 Case Result Summary ............................................................ 33
4.4 System Dispatch Results .................................................................. 35
4.4.1 Existing Solar, 2023 Base Case and Jim Bridger First
Unit Online Cases .............................................................. 35
4.4.2 High Solar, High Wind, and High Solar + Wind Cases 39
4.4.3 High Solar with Low, Average and High Hydro Budgets
............................................................................................... 42
4.4.4 High Solar With and Without Storage ............................ 45
4.4.5 High Must Take Solar and Curtailable Solar Cases .... 49
5 Discussion ............................................................................................ 53
5.1 Discussion of Current Study Results .............................................. 53
5.1.1 Effects of Binding Pmin Constraints on VER Integration
Costs ..................................................................................... 53
5.1.2 High Solar With Storage Cases ....................................... 56
5.2 Comparison to Data in Literature and 2018 Idaho Power VER
Study ..................................................................................................... 59
5.3 Methodological Differences between 2020 and 2018 Idaho Power
Company Variable Energy Resource Analysis ............................. 60
5.3.1 Overview .............................................................................. 60
5.3.2 Reserves .............................................................................. 61
5.3.3 Treatment of External Markets ........................................ 65
5.3.4 Multistage vs. Single Stage Model .................................. 65
6 Conclusions .......................................................................................... 67
6.1 Integration Costs ................................................................................. 67
7 Appendix 1: Process Document ........................................................ 68
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 8 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 8 of 88
7.1 Introduction ........................................................................................... 68
7.2 Results Processing ............................................................................. 74
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 9 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 9 of 88
x | P a g e
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 10 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 10 of 88
P a g e | 1 |
© 2010 Energy and Environmental Economics, Inc.
1 Introduction
1.1 Motivation and Background
In 2019, Idaho Power committed to using 100 percent clean energy by 2045.
While more than 50 percent of Idaho Power’s annual load was served by clean
resources in 2018 (primarily from hydroelectricity, with some additional wind and
solar resources), Idaho Power may potentially add significant amounts of variable
energy resources (VERs), such as wind and solar power, to achieve this 2045 goal.
Energy and Environmental Economics (E3) was retained by Idaho Power to
perform a study of the cost of integrating new VERs into Idaho Power’s system.
Idaho Power has periodically performed these studies and analyses to inform
regulatory proceedings, and to determine integration charges included in Public
Utility Regulatory Policies Act (PURPA) contracts. Idaho Power hired E3 to update
integration costs. E3 conducted this analysis by designing a suite of scenarios that
are relevant to the 2023 timeframe.
The following report details the modeling methodology, data collection and
assumptions, and results from E3’s 2020 investigation of VER integration costs for
Idaho Power.
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 11 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 11 of 88
P a g e | 2 |
2 Methodology
2.1 Calculating VER Integration Costs
E3 used five metrics to estimate the total cost of VER integration to Idaho Power’s
system. These were:
Start/Stop Costs: The costs resulting from changes in unit start and stop
counts due to differing patterns of net load fluctuations caused by higher
VER penetration
Ramping Costs: The costs resulting from changes in unit ramping due to
differing patterns of net load fluctuations caused by higher VER
penetration
Imperfect Unit Commitment and Dispatch Costs (Fuel Costs): The costs
resulting from holding a greater amount of committed dispatchable
resources operating at part load and lower efficiency. These resources
operate at part load to provide reserves necessary to manage increased
VER-induced forecast error and subhourly net load variability
Imperfect Unit Commitment and Dispatch Costs (Net Import Costs): The
costs resulting from suboptimal market transactions due to holding more
headroom and footroom on generators to account for increased VER-
induced forecast error and subhourly net load variability
Curtailment Costs: In all cases, VERs are assumed to be contracted on a
take-or-pay basis (i.e. all VER energy is paid for whether it is consumed or
not). However, in the case in which solar can be economically curtailed,
Idaho Power would incur a cost from no longer generating a renewable
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 12 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 12 of 88
P a g e | 3 |
© 2010 Energy and Environmental Economics, Inc.
energy credit (REC). This REC cost is factored into the integration cost for
that case.
The total VER integration cost for each case is calculated by summing each of
these costs.
To calculate these costs, E3 performed three model runs for each of the eleven
analyzed cases. In the first model run, E3 ran a 2023 “base case” model that
served as the reference point for each of the subsequent investigated cases. The
base case included potential unit additions and retirements, the relevant hydro
budget, as well as projected load growth from 2019 through 2023. Next, E3 ran
an intermediate “perfect foresight” case in which any new VER additions beyond
the 2023 base case have perfect foresight (i.e. no new forecast error reserves are
held vs. the base case), and the new VER profiles are “smoothed” on a subhourly
timescale (i.e. no new regulation reserves are held vs. the base case). This perfect
foresight case is designed specifically to look at the effect of forecast error and
subhourly variability from VERs on integration costs, not factoring in savings from
extra energy provided by new VER additions. Finally, E3 ran a case with higher
VER-induced regulation reserves and higher net load forecast error reserves. The
formulae for calculating integration costs from these three cases are provided
below. In the formulae, “Case j” refers to an individual case for which E3
calculated the VER integration costs. The “base case” is the 2023 base case
common to all but two of the evaluated cases. The remaining two cases are the
2023 base case and the base case with Bridger Unit 1 cases. These use the existing
solar case instead of the 2023 base case due to the need for an incremental VER
build to assess the integration costs in the equations provided below. The
resulting Total Integration Costs pursuant to this study are calculated in units of
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 13 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 13 of 88
P a g e | 4 |
$/MWh. The graphical depiction of this three-part process is also shown below in
Figure 1.
𝑰𝒏𝒄𝒓𝒆𝒎𝒆𝒏𝒕𝒂𝒍 𝑺𝒕𝒂𝒓𝒕 𝑪𝒐𝒔𝒕𝒔 𝒇𝒐𝒓 𝑪𝒂𝒔𝒆 𝒋
= 𝛴 𝑆𝑡𝑎𝑟𝑡 𝐶𝑜𝑠𝑡 ∗ (𝐴𝑛𝑛𝑢𝑎𝑙 𝑆𝑡𝑎𝑟𝑡 𝐶𝑜𝑢𝑛𝑡 ,
− 𝐴𝑛𝑛𝑢𝑎𝑙 𝑆𝑡𝑎𝑟𝑡 𝐶𝑜𝑢𝑛𝑡 , )
𝑰𝒏𝒄𝒓𝒆𝒎𝒆𝒏𝒕𝒂𝒍 𝑹𝒂𝒎𝒑𝒊𝒏𝒈 𝑪𝒐𝒔𝒕𝒔 𝒇𝒐𝒓 𝑪𝒂𝒔𝒆 𝒋
= 𝛴 𝑅𝑎𝑚𝑝𝑖𝑛𝑔 𝐶𝑜𝑠𝑡
∗ (𝐶𝑢𝑚𝑢𝑙𝑎𝑡𝑖𝑣𝑒 𝑅𝑇5 𝑀𝑊 𝑅𝑎𝑚𝑝𝑖𝑛𝑔 ,
− 𝐶𝑢𝑚𝑢𝑙𝑎𝑡𝑖𝑣𝑒 𝑅𝑇5 𝑀𝑊 𝑅𝑎𝑚𝑝𝑖𝑛𝑔 , )
𝑰𝒏𝒄𝒓𝒆𝒎𝒆𝒏𝒕𝒂𝒍 𝑰𝒎𝒑𝒆𝒓𝒇𝒆𝒄𝒕 𝑼𝒏𝒊𝒕 𝑪𝒐𝒎𝒎𝒊𝒕𝒎𝒆𝒏𝒕 & 𝑫𝒊𝒔𝒑𝒂𝒕𝒄𝒉 𝑪𝒐𝒔𝒕 𝒇𝒐𝒓 𝑪𝒂𝒔𝒆 𝒋
= 𝛴 𝐹𝑢𝑒𝑙 𝐶𝑜𝑠𝑡 ∗ (𝐹𝑢𝑒𝑙 𝑈𝑠𝑒 ,
− 𝐹𝑢𝑒𝑙 𝑈𝑠𝑒 ,"Perfect Foresight" ) + (𝑁𝑒𝑡 𝐼𝑚𝑝𝑜𝑟𝑡 𝐶𝑜𝑠𝑡
− 𝑁𝑒𝑡 𝐼𝑚𝑝𝑜𝑟𝑡 𝐶𝑜𝑠𝑡" " )
𝑰𝒏𝒄𝒓𝒆𝒎𝒆𝒏𝒕𝒂𝒍 𝑪𝒖𝒓𝒕𝒂𝒊𝒍𝒎𝒆𝒏𝒕 𝑪𝒐𝒔𝒕𝒔 𝒇𝒐𝒓 𝑪𝒂𝒔𝒆 𝒋
= 𝛴 𝐶𝑢𝑟𝑡𝑎𝑖𝑙𝑚𝑒𝑛𝑡 𝐶𝑜𝑠𝑡
∗ (𝐶𝑢𝑚𝑢𝑙𝑎𝑡𝑖𝑣𝑒 𝑅𝑇5 𝑀𝑊 𝐶𝑢𝑟𝑡𝑎𝑖𝑙𝑚𝑒𝑛𝑡 ,
− 𝐶𝑢𝑚𝑢𝑙𝑎𝑡𝑖𝑣𝑒 𝑅𝑇5 𝑀𝑊 𝐶𝑢𝑟𝑡𝑎𝑖𝑙𝑚𝑒𝑛𝑡 ," " )
𝑻𝒐𝒕𝒂𝒍 𝑰𝒏𝒕𝒆𝒈𝒓𝒂𝒕𝒊𝒐𝒏 𝑪𝒐𝒔𝒕𝑰𝒏𝒄.,𝑪𝒂𝒔𝒆 𝒋
=(𝐼𝑛𝑐.𝑆𝑡𝑎𝑟𝑡 𝐶𝑜𝑠𝑡, + 𝐼𝑛𝑐.𝑅𝑎𝑚𝑝𝑖𝑛𝑔 𝐶𝑜𝑠𝑡, +
𝐼𝑛𝑐𝑟𝑒𝑚𝑒𝑡𝑛𝑎𝑙 𝐼𝑚𝑝𝑒𝑟𝑓𝑒𝑐𝑡 𝑈𝑛𝑖𝑡 𝐶𝑜𝑚𝑚𝑖𝑡𝑚𝑒𝑛𝑡 𝑎𝑛𝑑 𝐷𝑖𝑠𝑝𝑎𝑡𝑐ℎ 𝐶𝑜𝑠𝑡, + 𝐼𝑛𝑐.𝐶𝑢𝑟𝑡.𝐶𝑜𝑠𝑡, )
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 14 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 14 of 88
P a g e | 5 |
© 2010 Energy and Environmental Economics, Inc.
𝑻𝒐𝒕.𝑰𝒏𝒕𝒆𝒈𝒓𝒂𝒕𝒊𝒐𝒏 𝑪𝒐𝒔𝒕𝑰𝒏𝒄.,𝑪𝒂𝒔𝒆 𝒋=
(𝐼𝑛𝑐.𝑆𝑡𝑎𝑟𝑡 𝐶𝑜𝑠𝑡, + 𝐼𝑛𝑐.𝑅𝑎𝑚𝑝𝑖𝑛𝑔 𝐶𝑜𝑠𝑡, +
𝐼𝑛𝑐.𝐼𝑚𝑝𝑒𝑟𝑓𝑒𝑐𝑡 𝑈𝑛𝑖𝑡 𝐶𝑜𝑚𝑚.𝑎𝑛𝑑 𝐷𝑖𝑠𝑝.𝐶𝑜𝑠𝑡,
𝑉𝐸𝑅 𝐸𝑛𝑒𝑟𝑔𝑦 𝑃𝑜𝑡𝑒𝑛𝑡𝑖𝑎𝑙, − 𝑉𝐸𝑅 𝐸𝑛𝑒𝑟𝑔𝑦 𝑃𝑜𝑡𝑒𝑛𝑡𝑖𝑎𝑙,
Figure 1: VER Integration Cost Calculation Methodology
This methodology for deriving VER integration costs does not calculate various
costs and benefits from the VER additions. Additionally, this method does not
calculate fuel cost savings due to VER deployment, nor the capacity value of new
VERs in offsetting the need for firm generation unit additions. This method also
does not calculate capital or PPA costs associated with contracting new VERs.
Therefore, the future use of these VER integration costs must be done with
knowledge and awareness of what costs and benefits they omit.
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 15 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 15 of 88
P a g e | 6 |
2.2 Production Cost Modeling
E3 used Energy Exemplar’s PLEXOS 7.2 Software2 to calculate the total production
costs associated with each evaluated case. The model uses load, VER, generator,
fuel cost and external market data provided by Idaho Power and other data
sources to calculate annual production costs for Idaho Power under varying
scenarios, which are then used to calculate VER integration costs. This is shown
schematically below in Figure 2.
In order to perform this modeling, E3 used a four-stage PLEXOS model. For each
day, the model sequentially solves the day-ahead (DA), hour-ahead (HA), 15-
minute (RT15) and 5-minute (RT5) markets. In each stage, the model is solved
completely (i.e. all units and transmission committed and dispatched). Then, any
unit commitment or other model decisions with a lead time longer than the next
phase’s lead time to the real time are passed down to the next stage. In this
manner, the model approximates the actual unit commitment and dispatch
constraints associated with the longer commitment times of thermal and
transmission markets. This captures the effects of greater average forecast error
and higher average reserves in model stages that are farther from the real time
on the ability of Idaho Power to efficiently commit long start units. This daily
sequential model execution process is depicted in Figure 3.
2 https://energyexemplar.com/solutions/plexos/
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 16 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 16 of 88
P a g e | 7 |
© 2010 Energy and Environmental Economics, Inc.
Figure 2: Using PLEXOS to Calculate VER Integration Costs
Figure 3: PLEXOS Multistage Modeling
The change in start/stop cost, and the imperfect unit commitment costs are
calculated endogenously in PLEXOS. However, E3 used data from the 2013
National Renewable Energy Laboratory’s (NREL) Western Wind and Solar
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 17 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 17 of 88
P a g e | 8 |
Integration Study: Phase 23 to estimate $/MW ramping costs for Idaho Power’s
thermal units. The annual total ramping costs were calculated as a post-
processing step by calculating the total annual MW of ramping in the RT5 stage
for each thermal unit, and multiplying that by the per MW ramping cost from
NREL. The $/MW values that E3 used are shown in Table 2 below.
Table 2: Ramping Costs Used in Study (Sourced from NREL4)
Value Coal Gas GT Gas CCGT
Median Ramping Cost ($/MW) $3 $2 $1
2.3 Reserve Modeling
E3 used its RESERVE tool5 to model 2019 and 2023 levels of hourly reserves that
Idaho Power needs to hold in each PLEXOS interval. This is done to account for
the fact that Idaho Power needs to hold reserves to manage net load forecast
error and subhourly net load variations in its daily operations.
Idaho Power’s participation in the California Independent System Operator’s
(CAISO’s) Energy Imbalance Market (EIM) means that Idaho Power holds reserves
of CAISO’s Flexible Ramping Product6 (FRP). It must do this so that it can trade in
the RT15 and RT5 EIM markets. Additionally, Idaho Power holds amounts of
regulation reserves and contingency reserves dictated by the North American
3 https://www.nrel.gov/docs/fy13osti/55588.pdf
4 https://www.nrel.gov/docs/fy13osti/55588.pdf
5 https://www.ethree.com/tools/reserve-model/
6http://www.caiso.com/informed/Pages/StakeholderProcesses/CompletedClosedStakeholderInitiatives/Flexib
leRampingProduct.aspx
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 18 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 18 of 88
P a g e | 9 |
© 2010 Energy and Environmental Economics, Inc.
Electric Reliability Corporation (NERC) and the Western Electricity Coordinating
Council (WECC).
While the derivation of contingency reserves is standardized (calculated as 3
percent of load and 3 percent of generation total, with at least half held as for
spinning reserves and the rest as non-spinning reserves), Idaho Power’s future
CAISO FRP and regulation reserve needs are unknown. This is because future VER
additions and load growth will increase the level of net load forecast uncertainty
on Idaho Power’s system relative to current conditions. Therefore, E3 used its
RESERVE tool along with Idaho Power’s 2019 forecast and actual load and VER
data to simulate reserves that approximate the CAISO FRP and regulation needs.
E3 also used RESERVE to calculate CAISO FRP and regulation reserves in 2019 to
enable a consistent baseline for model comparisons.
These contingency, CAISO FRP and regulation reserves were input to the PLEXOS
model such that the reserves are held in all time intervals. Further information on
the derivation of the 2023 load and VER profiles for each analyzed case are
provided in subsequent sections of this report.
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 19 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 19 of 88
P a g e | 10 |
3 Data Collection, Processing
and Methods
3.1 PLEXOS Modeling
3.1.1 LOAD PROFILES, VER PROFILES AND DISPATCHABLE GENERATION
FLEET
E3 collected forecast and actual gross load, wind and solar profiles for 2019 from
Idaho Power for the DA, HA, RT15 and RT5 phases. The VER data was on a plant-
level basis and covered most of Idaho Power’s existing PURPA and Idaho Power-
owned facilities, with only a few small wind and solar plants omitted from the
data collection process due to their small effect on net load forecast error. Idaho
Power also provided the total 2019 wind and solar nameplate build in Idaho
Power territory so that E3 could use a correct baseline VER build in its analysis.
Idaho Power’s 2019 average load was 1,980 aMW. To estimate 2023 loads, E3
used load growth projections from Idaho Power to uniformly increase 2019 loads
by approximately 5 percent total to 2,081 aMW. The method for deriving new
2023 VER profiles is detailed below, but the 2019 historical VER profiles were used
in all cases to derive the 2023 VER profiles.
In all cases, E3 modeled existing and proposed solar, solar + storage and wind
plants as qualifying facilities (QF) operating under PURPA. This means that, under
all circumstances except for one case, these resources are treated as must take
facilities.
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 20 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 20 of 88
P a g e | 11 |
© 2010 Energy and Environmental Economics, Inc.
E3 chose to use 2019 load and VER data to derive 2023 load and VER profiles in
order to capture the spatial and temporal correlations between load, wind and
solar production and forecast error, as well as the typical hourly and seasonal
distributions of load, and VER production. Most of Idaho Power’s existing solar
capacity is modern, single-axis tracking utility solar, meaning that future
installations were likely to have similar annual capacity factors as existing arrays.
Idaho Power’s solar and wind is mostly distributed across the Snake River Plain
and Eastern Oregon, as shown below in Figure 4, because this is where the
majority of existing Idaho Power transmission and load is, and it is also the best
solar resource in Idaho Power’s service territory. Idaho Power stated that they
are likely to continue to add new VER resources within the Snake River Plain.
Therefore, E3’s use of 2019 VER profiles to represent future profiles is reasonable.
Idaho Power proposed that, for the 2023 base case, it was reasonable to assume
that 251 MW of new solar was online in their service territory (131 MW of
unspecified PURPA contracts and 120 MW from the planned Jackpot Solar
facility). Idaho Power also proposed that the 2023 wind capacity remain the same
as that from 2019.
Idaho Power provided detailed information on each of its thermal (coal, natural
gas combustion turbine, natural gas combined cycle and diesel) plants, as well as
its hydroelectric fleet. Unit outages, heat rates, fuel prices and other relevant
data were collected. Coal plants are modeled as must-run units with seasonal
outages for Idaho Power’s North Valmy Generating Station. Combined Cycle
plants (Langley Gulch) are committed in the hour-ahead timeframe and the gas
combustion turbine fleet has subhourly commitment intervals.
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 21 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 21 of 88
P a g e | 12 |
Figure 4: Existing Idaho Power VER Units for which E3 was Provided 2019 DA,
HA, RT15 and RT5 Profiles
Given the large share of hydroelectricity on Idaho Power’s system, E3 focused on
ensuring proper representation of the hydro fleet’s capacity, ramping capability,
daily energy budgets, hourly maximum and minimum power ratings and other
such data. Additionally, E3 considered three hydro years, comprising
representative “low,” “average,” and “high,” hydro years. These profiles were
determined by Idaho Power by choosing from historical data. The average daily
energy profiles for these low, average and high hydro years are shown in Figure
5.
Planned future coal unit retirements through 2023 were modeled per Idaho
Power input. The overall planned change in fleet composition from 2019 to 2023,
as well as the total unit capacities by generation type are provided in Table 3.
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 22 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 22 of 88
P a g e | 13 |
© 2010 Energy and Environmental Economics, Inc.
Idaho Power’s projected base case load and resource balance is shown below in
Figure 6.
Table 3: 2019 and 2023 Base Case Unit Capacities by Generator and Resource
Type
Figure 5: Daily High, Average and Low Hydro Energy Budget Profiles for Idaho
Power
3.1.2 EXTERNAL MARKET REPRESENTATION
Idaho Power was modeled as being able to trade with external electricity markets
at the Palo Verde and Mid C hubs. In the DA and HA stages of the model, Idaho
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 23 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 23 of 88
P a g e | 14 |
Power can make bilateral trades with its neighbors, while incurring a hurdle rate
to do so.
Figure 6: Base Case Load and Resource Balance in Idaho Power through 2030
E3 determined historical 2019 bilateral energy prices, hurdle rates, and transfer
limits through discussions with Idaho Power. In the RT15 and RT5 stages, Idaho
Power can trade with its neighbors in a manner consistent with Idaho Power’s
participation in the CAISO EIM, i.e. there are no hurdle rates, but there are
transfer limits. In the RT15 and RT5, Idaho Power trades electricity at the RTPD
(RT15) and RTD (RT5) 2019 EIM prices for the DGAP_IPCO_APND node, which is
an aggregated node that averages Idaho Power prices. E3 benchmarked the 2019
DGACP_IPCO_APND node prices versus 2019 nodal prices for the Elkhorn, High
Mesa and Rockland plants and found that the aggregated node provided a
satisfactory representation of these various wind plants.
In Q1 of 2019, there was a natural gas pipeline outage in the Alberta Electricity
System Operator (AESO) service territory, which inflated market prices in the
Pacific Northwest. Accordingly, E3 replaced the Q1 2019 market prices with Q1
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 24 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 24 of 88
P a g e | 15 |
© 2010 Energy and Environmental Economics, Inc.
2020 market prices for the DA, HA, RT15 and RT5 phases. Additionally, given the
2023 timeframe of the model, E3 used its AURORA Market Price forecasts to
create a month-hourly average basis differential between 2023 and 2019. This
was added to the historical market prices in order to reflect the effect of
anticipated growth of VERs and storage across the Western Interconnection from
2019 through 2023, among other changes.
E3 benchmarked the historical interaction of the Elkhorn, High Mesa and
Rockland wind plants with the EIM. E3 found its representation of Idaho Power’s
interactions with the EIM to be reasonable.
Finally, E3 combined Idaho Power’s multiple hydroelectric projects into two units
for modeling simplicity. One unit consisted of aggregated run-of-river (RoR)
plants, whose output is largely inflexible and in flat hourly blocks, and the other
consisted of the combined Hells Canyon Complex (HCC) units (consisting of the
Oxbow, Brownlee and Hell’s Canyon dams), whose output can be varied within
certain time windows. This division of Idaho Power’s hydroelectric assets into two
aggregated units was done to reflect the variation in flexibility, water storage and
dispatchability across Idaho Power’s hydro fleet.
Planned future coal unit retirements through 2023 were modeled per Idaho
Power input. The overall planned change in fleet composition from 2019 to 2023,
as well as the total unit capacities by generation type are provided in Table 3.
Idaho Power’s projected base case load and resource balance is shown in Figure
6.
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 25 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 25 of 88
P a g e | 16 |
3.2 RESERVE Modeling
3.2.1 DERIVATION OF 2023 VER PROFILES
As new VER resources are added, the average forecast error and subhourly
variability of the aggregated fleet will decline on a per MW of installed resources
basis. This is due to well-known diversity effects (i.e. as solar and wind plants are
installed at different locations, the average forecast error and subhourly variation
across all units will tend to decline on a per MW basis). Additionally, based on
experience in other jurisdictions, E3 assumed that there will be improvements to
VER forecast error in the future.
In order to capture these effects while using the 2019 VER data, E3 assessed the
reduction in forecast error and subhourly variability that Idaho Power has
observed to date. A similar approach was taken in Idaho Power’s 2018 Variable
Energy Resource Analysis. E3 performed this through the following steps
Randomly order the forecast and actual profiles for existing wind and
solar that Idaho Power provided to E3
Sequentially add solar profiles or wind profiles, each time calculating the
average forecast error and regulation reserves of the aggregated solar or
wind profiles using RESERVE
Fit a polynomial trend to the average reserves versus the total MW of
online VERs for the solar and wind profiles
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 26 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 26 of 88
P a g e | 17 |
© 2010 Energy and Environmental Economics, Inc.
From prior work in the CAISO Extended Day Ahead Market project7, E3
assumed a 2 percent per annum improvement in VER forecasting
(average mean average percentage error reduction)
For each future VER build, linearly scale up the 2019 VER forecast and
actual profiles by the ratios of future VER build total online MW to 2019
online MW
Reduce the forecast error equally in all intervals using the polynomial
trend fit to forecast error data and using the estimated 2 percent per
annum improvement in forecast error from 2019 to 2023
Reduce the subhourly interval-to-interval variation by the amount
derived from the polynomial trend fit to the regulation error data
Run RESERVE for this new set of profiles; and
Input these new set of profiles to PLEXOS
Using this process, the average standalone (i.e. not net-load-based) HA forecast
error reserves and regulation reserves for wind and solar would decline as shown
below in Table 4. These data show the reduction in average forecast error and
regulation needs across all hours of the year, relative to a case with no diversity
benefits or forecast error improvements and the same VER unit additions.
As can be seen in Table 4, E3 projects that regulation reserves will drop more on
a percentage basis than CAISO FRP reserves needs will in the high solar and high
wind cases. This is consistent with the larger percentage increase in solar build
than wind build in the high solar versus high wind cases, respectively.
7 https://stakeholdercenter.caiso.com/StakeholderInitiatives/Extended-day-ahead-market
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 27 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 27 of 88
P a g e | 18 |
Table 4: Average Projected Improvement in Forecast Error and Regulation
Reserves from Diversity and Forecasting Improvements
Case Average CAISO FRP
Reserve Improvement
Average Regulation
Reserve
Improvement
Base 2023 Case Solar (251 MW
new solar added to 2019 build)
11.7 % 14.2 %
Base 2023 Wind Case (0 MW new
wind added to 2019 build)
7.8 % 0.0 %
2023 Hi Solar Case (794 MW new
solar added to 2019 build)
17.2 % 31.6 %
2023 Hi Wind Case (669 MW new
wind added to 2019 build)
13.2 % 19.1 %
3.2.2 DERIVING RESERVES COMPONENTS
The CAISO FRP’s reserves for each interval consist of an uncertainty component,
plus a net load change from the previous interval, minus a credit component
based on the lesser of either the EIM-wide average footprint diversity or the
Balancing Authority’s (BA) trading position-derived credit. E3 used the
information provided by Idaho Power on forecast and actual load, wind and solar
to derive uncertainty requirements for the CAISO FRP. Given E3’s simplified
representation of Idaho Power’s external market transactions, E3 assumed that
the credit component of the reserve created a 40 percent reduction versus the
uncertainty component alone. This 40 percent value is an approximate value, and
was calculated using average historically-observed EIM footprint diversity in
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 28 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 28 of 88
P a g e | 19 |
© 2010 Energy and Environmental Economics, Inc.
2019.8 This derivation, and its differences from the 2018 Idaho Variable Energy
Resource Integration Study is further discussed in Section 5.3.2.
3.3 Case Matrix
E3 and Idaho Power worked together to derive a total of eleven 2023 cases to
examine, in addition to a 2019 base case, which are described below. Table 5
details the specifics of each case.
Case 1 is the 2023 base case for Cases 3-6 and Cases 8-11, which has
proposed unit additions and retirements and also includes the known
2019 through 2023 load growth
Case 2 explores the effect of not retiring one of the Bridger coal plant’s
units, but is otherwise identical to Case 1
Case 3 builds on Case 1 by exploring the effect of adding enough new
solar (794 MW of new solar) such that 10 percent of the 2023 Idaho
Power average gross load is provided by this new solar build
Case 4 extends the Case 3 analysis to a low, rather than average hydro
year
Case 5 builds on Case 1 and explores the integration costs of a high wind
build. Case 5 assumes a new wind build that can supply 10 percent of the
annual 2023 Idaho Power gross load (669 MW of new wind)
Case 6 builds on Case 3 and Case 5, including both high solar and high
wind builds (794 MW of new solar and 669 MW of new wind)
8 https://www.westerneim.com/Pages/About/QuarterlyBenefits.aspx
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 29 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 29 of 88
P a g e | 20 |
Case 7 is identical to Case 1, except that none of proposed solar additions
come online from 2019 to 2023, resulting in 251 MW fewer of solar than
Case 1 and lower reserves needs
Cases 8 extends the Case 3 analysis to a high, rather than average hydro
year
Case 9 extends the Case 3 analysis to have 200 MW of 4-hour, Federal
Investment Tax Credit (ITC)-enabled Li-Ion battery storage
Case 10 extends the Case 3 analysis to have 400 MW of 4-hour, ITC-
enabled Li-Ion battery storage
Case 11 extends the Case 3 analysis to allow economic curtailment of the
794 MW of new solar resource, while the 561 MW of existing and
proposed solar remain must-take resources
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 30 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 30 of 88
P a g e | 21 |
© 2010 Energy and Environmental Economics, Inc.
Table 5: Case Matrix for 2023 Cases
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 31 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 31 of 88
P a g e | 22 |
4 Results
The following section provides detailed results from this work. A discussion of the
implications of these detailed results on VER integration in Idaho Power’s system
is provided in Section 5.
4.1 RESERVE Outputs
4.1.1 ANNUAL AVERAGE RESULTS
The average annual reserves for each of the cases is shown below in Table 6. It
should be noted that actual reserves vary on an hourly or subhourly basis in all
stages. However, E3 provided these average annual reserves as a general
indicator of how reserves needs change from case to case. These same data are
displayed below for the hour-ahead forecast’s CAISO FRP, regulation and
contingency reserves on a percentage of average monthly load basis for each
unique combination of solar and wind in Table 7, Table 8, Table 9, Table 10 and
Table 11. As observed in Table 6, wind reserves have more forecast error (CAISO
FRP reserves), whereas solar reserves have more subhourly variability. This trend,
observed here, is also expressed elsewhere in the literature.
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 32 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 32 of 88
P a g e | 23 |
© 2010 Energy and Environmental Economics, Inc.
Table 6: Average 2023 Case Reserves Needs
Case Total
MW
Wind
(MW)
Total
MW
Solar
(MW)
Avg.
RT15
FRP
Up
(MW)
Avg.
RT15
FRP
Down
(MW)
Avg.
Reg.
Up
(MW)
Avg.
Reg.
Down
(MW)
Avg.
Conting.
Res.
(MW)
Avg.
Total
Res. Up
(Percent
of Avg.
Load)
Avg. Total
Reserves
Down
(Percent of
Avg. Load)
1. 2023
Base
Case 728 561 100 97 40 41 104 13 % 7 %
2. Jim
Bridger
Online 728 561 100 97 40 41 104 13 % 7 %
3. Hi
Solar 728 1,354 147 142 71 72 104 17 % 11 %
4. Hi
Solar,
Low
Hydro
728 1,354 147 142 71 72 104 17 % 11 %
5. Hi
Wind 1,396 561 152 147 50 52 104 16 % 10 %
6. Hi
Solar, Hi
Wind 1,396 1,354 193 186 79 81 104 19 % 13 %
7.
Existing
Solar
Case
728 561 87 86 32 33 104 11% 6%
8. Hi
Solar, Hi
Hydro 728 1,354 147 142 71 72 104 17 % 11 %
9. Hi
Solar,
200 MW
Battery 728 1,354 147 142 71 72 104 17 % 11 %
10. Hi
Solar,
400 MW
Battery 728 1,354 147 142 71 72 104 17 % 11 %
11.
Curtail.
Solar 728 1,354 147 142 71 72 104 17 % 11 %
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 33 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 33 of 88
P a g e | 24 |
Table 7: 2023 Monthly Average, Load Normalized CAISO FRP, Regulation and
Contingency Reserves, Base 2023 Cases (Case 1 and Case 2)
Table 8: 2023 Monthly Average, Load Normalized CAISO FRP, Regulation and
Contingency Reserves, Existing Solar 2023 Case (Case 7)
Month
Hour Ahead
FRP + Reg. +
Contingency
Headroom,
Total
(% of Load)
Hour
Ahead FRP
+ Reg.
Solar
(% of Load)
Hour
Ahead FRP
+ Reg.
Headroom,
Wind
(% of Load)
Hour Ahead
FRP + Reg. +
Contin.
Headroom,
Load
(% of Load)
Hour
Ahead FRP
+ Reg.
Footroom,
Total
(% of Load)
Hour
Ahead FRP
+ Reg.
Footroom,
Solar
(% of Load)
Hour
Ahead FRP
+ Reg.
Footroom,
Wind
(% of Load)
Hour
Ahead FRP
+ Reg.
Footroom,
Load
(% of Load)
1 11.6% 0.5% 2.9% 8.2% 5.1% 0.5% 2.8% 1.7%
2 11.2% 0.5% 2.5% 8.3% 5.8% 0.6% 3.5% 1.6%
3 12.8% 1.5% 3.0% 8.2% 6.2% 1.5% 3.0% 1.7%
4 13.3% 1.6% 3.5% 8.2% 8.0% 1.8% 4.6% 1.6%
5 12.4% 1.6% 2.7% 8.2% 7.4% 2.0% 3.8% 1.6%
6 12.1% 1.4% 2.6% 8.1% 4.8% 1.0% 2.2% 1.6%
7 10.6% 1.0% 1.4% 8.2% 3.9% 0.8% 1.7% 1.4%
8 10.7% 1.0% 1.5% 8.2% 4.1% 0.8% 1.8% 1.5%
9 12.3% 1.1% 2.7% 8.5% 5.5% 1.0% 2.7% 1.8%
10 12.2% 1.2% 2.8% 8.3% 7.2% 1.3% 4.3% 1.6%
11 12.1% 1.2% 2.5% 8.4% 6.7% 1.1% 3.8% 1.8%
12 10.9% 0.5% 2.3% 8.1% 6.3% 0.6% 4.1% 1.6%
Avg. 11.86% 1.1% 2.5% 8.2% 5.9% 1.1% 3.2% 1.6%
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 34 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 34 of 88
P a g e | 25 |
© 2010 Energy and Environmental Economics, Inc.
Table 9: 2023 Monthly Average, Load Normalized Regulation Reserves, High
Solar Cases (Cases 3, 4, 8-11)
Table 10: 2023 Monthly Average, Load Normalized Regulation Reserves, High
Wind Case (Case 3)
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 35 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 35 of 88
P a g e | 26 |
Table 11: 2023 Monthly Average, Load Normalized Regulation Reserves, High
Solar and High Wind Case (Case 6)
4.1.2 DETAILED RESERVE RESULTS
While additions of new solar and wind both cause a similar increase in average
reserves needs, the hours in which they increase reserves are very different. The
following discussion illustrates these differences.
As observed in Table 6, wind reserves have more forecast error (CAISO FRP
reserves), whereas solar reserves have more subhourly variability. This trend,
observed here, is also expressed elsewhere in the literature.9
Conversely, the incremental FRP needs from adding solar shown in Figure 11
indicate that CAISO FRP reserves increase primarily during solar hours. FRP
reserves do increase at night because caps on the level of uncertainty imposed
9 https://www.nrel.gov/docs/fy13osti/55588.pdf
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 36 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 36 of 88
P a g e | 27 |
© 2010 Energy and Environmental Economics, Inc.
by the CAISO FRP derivation10 (see further discussion in Section 5.3.2) also
increase. Similarly, solar regulation needs increase only during solar hours.
Because reserves can only be provided with dispatchable resources in the PLEXOS
model, it is important to compare the need for reserves with the availability of
dispatchable resources. Figure 13 and Figure 14 show month-hourly average
residual net load, calculated as load minus wind, solar, and RoR hydro for the High
Solar and High Wind cases. This residual net load is the average load that must be
met by dispatchable resources and imports. If the need for reserves is greater
than the residual net load, then the model must export power to the market to
be able to serve Idaho Power’s reserves needs while not violating minimum
generation setpoints for online units. As discussed below, this can result in
exports to the market at a loss.
As can be seen from Figure 13, in the High Solar case, in March, April, May and
October, the residual net load is very low during the midday hours in which there
is high demand on reserves. Alternatively, as can be seen in the high wind case
for Figure 10, the residual net load is significantly higher during those midday
hours, and as shown earlier, average reserves needs are not especially high
midday.
10 See https://bpmcm.caiso.com/Pages/BPMDetails.aspx?BPM=Market percent20Operations for a discussion of
these caps; E3 derives its own caps from P98 and P2 values of the seasonal forecast error.
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 37 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 37 of 88
P a g e | 28 |
Figure 7: Average Month-Hourly CAISO FRR Headroom Needs for Base 2023
Case
Figure 8: Average Month-Hourly Regulation Reserves Headroom Needs for 2023
Base Case
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
1 54 72 54 103 82 83 99 84 73 95 98 95 103 103 103 103 103 103 100 103 54 69 75 93 88
2 71 52 35 63 75 57 43 50 67 95 103 103 103 103 103 103 103 103 80 84 103 103 103 103 84
3 68 81 80 92 65 80 84 80 102 110 127 127 71 124 127 127 127 127 127 73 64 78 75 71 95
4 56 47 58 74 80 97 83 67 84 127 127 127 105 105 127 127 127 127 127 113 67 90 93 67 96
5 67 90 84 63 68 71 67 86 112 127 127 127 89 127 127 127 127 127 127 121 104 71 72 78 99
6 71 78 130 151 151 151 151 151 151 151 151 151 151 151 151 128 101 151 151 151 116 99 101 80 132
7 57 63 53 50 41 44 59 67 151 142 151 128 115 151 144 122 147 151 151 151 120 103 87 53 104
8 25 50 63 59 61 57 54 53 113 151 142 151 151 117 134 144 151 151 151 151 98 85 103 61 103
9 71 72 76 66 80 72 92 108 129 129 129 129 129 129 129 129 129 129 129 122 122 83 69 68 105
10 76 69 73 56 53 63 53 54 61 113 129 129 129 129 125 129 129 129 74 50 89 74 54 78 88
11 56 54 65 58 57 75 80 78 93 122 118 129 118 109 129 129 129 95 110 87 70 59 59 56 89
12 71 55 65 63 66 80 70 67 51 78 91 103 103 103 94 71 71 66 47 50 61 68 87 87 74
62 65 70 75 73 77 78 79 99 120 124 125 114 121 124 120 120 122 115 105 89 82 81 74
Average Modeled CAISO FRR Headroom (MW)
Hour of Day
Month
Hour Average
Month
Average
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
1 24 24 24 24 24 25 25 27 27 32 39 40 41 48 47 53 51 41 26 26 25 24 24 24 32
2 25 25 25 25 25 25 26 28 27 33 39 41 41 48 46 55 51 42 27 26 26 25 25 25 33
3 21 21 21 22 23 23 23 33 49 53 50 74 74 85 99 109 99 90 67 34 22 22 21 21 48
4 21 21 20 21 22 23 23 34 48 49 46 58 65 71 82 85 91 83 66 35 24 24 22 22 44
5 20 21 21 21 21 22 22 38 43 42 42 53 58 62 67 77 86 74 63 34 24 23 22 21 41
6 25 24 23 23 23 23 28 45 60 60 40 37 41 44 44 62 68 71 59 54 35 34 29 27 41
7 29 27 25 24 24 25 29 57 74 73 49 42 45 45 44 68 76 81 59 52 35 37 38 34 46
8 26 24 24 23 23 24 27 56 72 61 48 44 46 47 49 61 66 77 83 53 35 36 33 29 45
9 23 23 23 22 23 23 24 24 35 54 51 47 49 53 61 78 77 57 36 26 27 26 25 24 38
10 22 21 21 21 23 23 24 24 35 58 51 56 51 61 69 84 83 57 36 23 23 24 23 22 39
11 23 22 22 22 22 22 23 23 34 57 54 57 62 68 78 87 82 57 36 23 22 22 23 23 40
12 23 23 23 23 23 23 24 24 25 30 37 40 40 50 50 57 53 43 25 24 24 24 23 23 31
23 23 23 23 23 23 25 34 44 50 46 49 51 57 62 73 74 64 49 34 27 27 26 24
Average Regulation Headroom - RMS Combined Load + Wind + Solar (MW)
Hour Average
Month
Hour of Day Month
Average
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 38 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 38 of 88
P a g e | 29 |
© 2010 Energy and Environmental Economics, Inc.
Figure 9: High Wind Minus Base Case CAISO FRR Headroom
Figure 10: High Wind Minus Base Case Regulation Headroom
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
1 16 17 17 16 16 17 17 16 15 14 11 11 11 9 10 9 9 11 15 15 15 16 16 16 14
2 17 17 17 17 17 17 17 16 16 14 13 12 12 11 11 10 10 12 16 16 16 16 16 17 15
3 13 13 13 13 13 13 13 10 8 7 7 5 5 4 4 3 4 4 5 9 12 12 12 13 9
4 16 17 17 16 15 15 14 11 8 8 8 7 7 7 6 6 6 6 7 11 15 15 16 17 11
5 14 13 13 14 13 13 11 8 7 7 7 6 6 6 6 6 5 6 7 11 14 14 14 13 10
6 15 14 14 14 14 13 11 8 6 7 8 8 8 8 8 8 7 7 8 9 13 13 14 14 10
7 14 14 15 14 14 13 12 7 6 6 6 7 7 7 8 6 6 7 8 9 13 13 12 13 10
8 13 13 13 13 13 12 11 6 5 6 7 6 6 6 6 6 6 5 5 8 12 11 12 13 9
9 15 15 15 14 14 14 13 13 10 7 7 8 7 8 7 6 5 7 10 13 14 14 14 15 11
10 15 15 15 14 14 14 15 14 11 8 8 8 9 7 6 5 5 8 12 15 15 15 15 15 12
11 13 12 13 13 13 12 13 12 9 6 6 6 5 5 4 4 4 6 10 13 13 13 13 13 10
12 15 14 14 14 14 14 13 13 12 11 9 8 8 7 7 6 7 8 12 13 13 14 14 15 11
15 15 15 14 14 14 13 11 9 8 8 8 7 7 7 6 6 7 10 12 14 14 14 14 11
Month
Hour Average
Difference, Hi Wind to Base Case, Average Regulation Headroom (MW)
Hour of Day Month
Average
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 39 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 39 of 88
P a g e | 30 |
Figure 11: High Solar Minus Base Case CAISO FRR Headroom
Figure 12: High Solar Minus Base Case Regulation Headroom
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
1 0 0 0 3 0 0 0 0 -1 19 47 61 60 60 60 60 60 58 2 6 0 0 0 0 21
2 0 0 0 0 0 0 0 1 7 40 60 60 60 60 60 60 60 60 7 0 42 21 60 44 29
3 0 0 0 0 0 0 0 12 28 75 114 114 63 103 114 114 114 114 104 23 0 0 0 0 45
4 0 0 0 0 0 1 10 36 52 114 114 114 124 116 114 114 114 114 114 56 4 0 0 0 55
5 0 0 0 0 0 2 24 41 129 114 114 114 119 114 114 114 114 114 114 88 12 0 0 0 60
6 0 0 0 9 45 99 89 116 124 124 124 124 124 124 124 147 67 124 124 124 43 1 0 0 77
7 0 0 0 0 0 11 31 72 124 133 124 113 106 124 116 136 126 124 124 124 27 1 0 0 67
8 0 0 0 0 0 2 9 70 135 124 132 124 101 69 72 104 124 124 124 104 4 0 0 0 59
9 0 0 0 0 0 0 3 20 85 111 111 111 111 111 111 111 111 111 111 23 0 0 0 0 52
10 0 0 0 0 0 0 4 19 40 127 111 111 111 109 95 111 111 111 30 0 0 0 0 0 45
11 0 0 0 0 0 0 0 7 15 28 58 82 81 71 111 111 111 37 0 0 0 0 0 0 30
12 0 0 0 0 0 0 0 0 3 33 62 60 60 60 68 49 35 6 0 0 0 0 0 0 18
0 0 0 1 4 10 14 33 62 87 98 99 93 93 97 103 96 91 71 46 11 2 5 4 47Hour Average
Difference, Hi Solar to Base Case, Average CAISO FRR Headroom (MW)
Hour of Day Month
Average
Month
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
1 0 0 0 0 0 0 0 0 0 19 37 39 42 55 53 65 60 41 0 0 0 0 0 0 17
2 0 0 0 0 0 0 0 0 0 20 35 40 41 53 49 60 59 42 0 0 0 0 0 0 17
3 0 0 0 0 0 0 0 27 58 64 60 97 97 106 131 144 134 122 87 33 0 0 0 0 48
4 0 0 0 0 0 0 0 30 56 57 54 71 82 91 108 112 121 106 85 32 0 0 0 0 42
5 0 0 0 0 0 0 1 38 48 46 46 63 72 77 85 99 107 91 78 28 0 0 0 0 37
6 0 0 0 0 0 0 12 45 71 63 26 16 27 32 33 55 66 73 70 51 3 0 0 0 27
7 0 0 0 0 0 0 5 62 92 74 30 21 32 36 37 77 93 100 85 52 3 0 0 0 33
8 0 0 0 0 0 0 5 65 87 66 34 19 34 38 44 60 77 88 105 52 3 0 0 0 32
9 0 0 0 0 0 0 0 1 31 59 56 46 56 59 70 99 99 67 30 0 0 0 0 0 28
10 0 0 0 0 0 0 0 0 30 68 60 65 61 73 86 117 110 71 35 0 0 0 0 0 32
11 0 0 0 0 0 0 0 0 30 72 65 71 80 89 105 118 111 71 35 0 0 0 0 0 35
12 0 0 0 0 0 0 0 0 0 19 36 42 42 59 60 71 64 45 0 0 0 0 0 0 18
0 0 0 0 0 0 2 22 42 52 45 49 55 64 72 90 92 76 51 21 1 0 0 0 31
Month
Hour Average
Difference, Hi Solar to Base Case, Average Regulation Headroom (MW)
Hour of Day Month
Average
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 40 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 40 of 88
P a g e | 31 |
© 2010 Energy and Environmental Economics, Inc.
Figure 13: Residual Net Load, High Solar Case 3
Figure 14: Residual Net Load, High Wind Case 5
4.2 2019 PLEXOS to Historical Case Benchmarking
E3 and Idaho Power performed rigorous benchmarking to ensure that the PLEXOS
model was able to reasonably replicate actual 2019 historical behavior. E3 and
Idaho Power verified that the following were in line with historical 2019 behavior:
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 41 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 41 of 88
P a g e | 32 |
Hydro and thermal unit flexibility (ramping rate) and dispatch
(distribution of ramps);
Total generation by unit and technology class;
Market transaction behavior and external market prices;
Average Idaho Power nodal energy prices;
Unit capacities;
Unit outages;
Number of unit starts; and
Average unit marginal operational cost
Particular attention was paid to the HCC to ensure its operation was reasonable.
This was critical because of the large amount of Idaho Power’s energy from
hydroelectricity in a typical year, as well as the crucial role that this unit has in
providing flexibility. Figure 15 below shows a sample of the verification of the
model wherein actual dispatch of the PLEXOS HCC is shown to be within the daily
maximum and minimum power output ranges, and the dispatch of the HCC
adheres to the input daily hydro budget.
Additionally, after initial results were analyzed, the Idaho Power team thought
that EIM transactions were unrealistically high in the PLEXOS model, given that
the model operates a price taker for market transactions. In reality, if Idaho
Power made particularly large sales or purchases in the EIM, prices would be
affected. Therefore, E3 and Idaho Power worked together to limit total net sales
and purchases in the EIM to +/- 300 MW in price taker mode. In instances in which
the model traded between +/- 300 MW up to the line limits in the real time, the
model paid a hurdle rate of $150/MW, which was implemented to approximate
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 42 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 42 of 88
P a g e | 33 |
© 2010 Energy and Environmental Economics, Inc.
“price setting” behavior. Overall, there were few hours in which the model
accessed this additional EIM flexibility.
Figure 15: PLEXOS HCC Dispatch vs. Historical Power and Hydro Budget Bounds
4.3 2023 Case Result Summary
The Incremental specific integration costs for each of the cases is provided below
in Table 12. These results are discussed in greater detail below in Chapter 5.
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 43 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 43 of 88
P a g e | 34 |
Table 12: Summary of Incremental VER Integration Costs
Case Inc. Start
Costs
($Million/
yr)
Inc.
Ramping
Costs
($ Million/ yr)
Total
Inc.
Imperf.
Unit Commit. & Dispatch
Costs
($Mill./yr)
Total
Curtail.
Costs
($Million/yr)
Total
Inc.
Integrat.
Costs
($Million/ yr)
Total
Product.
Cost
($Million/ yr)
Total
Inc.
VER
Gen.
(GWh /yr)
Total
Inc.
Specific
Integrat. Costs
($/MWh)
1. 2023
Base
Case -$0.15 $0.22 $1.62 $0.00 1.69 $181 577 $2.93
2. Jim
Bridger
Online
-$0.17 $0.37 $1.88 $0.00 $2.08 $180 577 $3.61
3. Hi
Solar $0.80 $0.45 $5.78 $0.00 $7.04 $146 1,824 $3.86
4. Hi
Solar,
Low
Hydro
$0.60 $0.53 $7.16 $0.00 $8.29 $172 1,824 $4.55
5. Hi
Wind $0.35 -$0.07 $1.12 $0.00 $1.41 $143 1,823 $0.77
6. Hi
Solar +
Hi Wind
$1.63 $0.33 $7.01 $0.00 $8.96 $109 3,647 $2.46
7.
Existing
Solar
Base
n/a n/a n/a n/a n/a $193 0 n/a
8. Hi
Solar, Hi
Hydro
$2.41 $0.19 $5.87 $0.00 $8.47 $75 1,823 $4.65
9. Hi
Solar,
200 MW
Battery
$0.58 $0.02 $0.56 $0.00 $1.16 $144 1,823 $0.64
10. Hi
Solar,
400 MW
Battery
$0.58 -$0.34 $1.46 $0.00 $1.69 $142 1,823 $0.93
11. Hi
Curtail.
Solar
$0.72 $0.39 $4.31 $0.29 $5.71 $147 1,823 $3.13
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 44 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 44 of 88
P a g e | 35 |
© 2010 Energy and Environmental Economics, Inc.
4.4 System Dispatch Results
In the following subsections, detailed day plots and other modeling results will be
used to illustrate how the Idaho Power system responds to adding different
capacities and kinds of VERs, and increasing or decreasing system flexibility. To
facilitate this, this study will examine the following case groupings:
Existing Solar (Case 7), Base Case (Case 1) and Jim Bridger First Unit Online
(Case 2)
High Solar (Case 3), High Wind (Case 5) and High Solar + Wind (Case 6)
High Solar with Low (Case 4), Average (Case 3) and High (Case 8) Hydro
Budgets
High Solar with (Cases 9 and 10) and without (Case 3) battery storage
Hi Solar with (Case 11) and without the ability to economically curtail
solar (Case 3)
4.4.1 EXISTING SOLAR, 2023 BASE CASE AND JIM BRIDGER FIRST UNIT
ONLINE CASES
This case comparison illustrates the effect of adding successively more VERs, as
well as increasing the aggregate system thermal minimum power level (Pmin).
The salient differences between cases are outlined as follows
Total online solar
o Existing Solar (Case 7): 310 MW
o 2023 Base Case (Case 1): 561 MW
o Jim Bridger Online Case (Case 2): 561 MW
Jim Bridger Coal Plant Pmin/Pmax
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 45 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 45 of 88
P a g e | 36 |
o Existing Solar (Case 7): 89 MW / 533 MW
o 2023 Base Case (Case 1): 89 MW / 533 MW
o Jim Bridger Online Case (Case 2): 118 MW / 707 MW
In the modeled year of 2023, there will be periods during the daytime in the
spring and fall in which external electricity prices are low or negatively priced.
This is due to the growing penetration of solar across the WECC footprint and the
low net loads during these periods. Figure 16 illustrates the Idaho Power system
operation operating during a day (April 23, 2023) that exhibits these conditions.
Beginning with the “Existing Solar Case,” which models the Idaho Power system
with the 2019 levels of wind and solar, the model will choose to purchase power
from the market rather than generate its own power during these periods. This is
shown by the purchase of electricity 4 am through 8 pm MST in Figure 16.
In the 2023 base case, 561 MW of solar is assumed to be online, which increases
Idaho Power’s total VER Pmin during midday periods. This decreases Idaho
Power’s ability to purchase negatively priced electricity from the market. This is
shown in Figure 16, wherein purchases are now only made in the morning and
evening periods.
Per discussions with Idaho Power, the Jim Bridger coal plant is modeled as a must-
run unit. As such, in the first Jim Bridger unit online case, the aggregate thermal
Pmin increases during all hours by 29 MW. Having both more solar and Jim
Bridger’s first unit online further increases Idaho Power’s aggregate Pmin. In
Figure 16, this results the model no longer purchasing negatively priced electricity
in the afternoon.
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 46 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 46 of 88
P a g e | 37 |
© 2010 Energy and Environmental Economics, Inc.
Though not depicted here, during periods of high net load (e.g. during summer
peaking events), the addition of extra solar and the ability to dispatch more power
from Jim Bridger can prove beneficial in reducing system costs by displacing
expensive market purchases and/or natural gas combustion turbine (CT) and/or
combined cycle (CCGT) generation. Per Table 13, as more solar is added, and if a
Jim Bridger unit is not retired, total incremental specific VER integration costs rise
but total production costs fall.
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 47 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 47 of 88
P a g e | 38 |
Figure 16: Existing Solar vs. 2023 Base Case vs. First Bridger Unit Online Daily
Dispatch Plots
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 48 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 48 of 88
P a g e | 39 |
© 2010 Energy and Environmental Economics, Inc.
Table 13: Summary of Results for Existing Solar, Base Case Solar and Jim Bridger
Cases
Case Inc. Start
Costs
($Million/
yr)
Inc.
Ramping
Costs
($ Million/ yr)
Total
Inc.
Imperf.
Unit Commit. & Dispatch
Costs
($Million
/ yr)
Total
Curtail.
Costs
($Million/yr)
Total
Inc.
Integrat.
Costs
($Million/ yr)
Total
Product.
Cost
($Million/ yr)
Total
Inc.
VER
Gen.
(GWh /yr)
Total
Inc.
Specific
Integrat. Costs
($/MWh)
1. 2023
Base
Case
-$0.15 $0.22 $1.62 $0.00 1.69 $181 577 $2.93
2. Jim
Bridger
Online
-$0.17 $0.37 $1.88 $0.00 $2.08 $180 577 $3.61
7.
Existing
Solar
Base
n/a n/a n/a n/a n/a $193 0 n/a
4.4.2 HIGH SOLAR, HIGH WIND, AND HIGH SOLAR + WIND CASES
This set of cases illustrates the difference in the ease of integrating equivalent
amounts of new VER energy from solar and wind. Additionally, the effects of
combining these solar and wind additions is shown.
The salient differences in VER capacities between these cases are as follows:
Total Online Solar
o High Solar Case (Case 3): 1,355 MW
o High Wind Case (Case 5): 561 MW
o High Solar + High Wind Case (Case 6): 1,355 MW
Total Online Wind
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 49 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 49 of 88
P a g e | 40 |
o High Solar Case (Case 3): 728 MW
o High Wind Case (Case 5): 1,397 MW
o High Solar + High Wind Case (Case 6): 1,397 MW
This case builds on the phenomena observed in Section 4.4.1, wherein adding
more VERs reduces the model’s ability to optimally perform market transactions
during low net load, springtime conditions. Figure 17 below depicts the high wind,
high solar, and high solar + high wind cases on the same low net load spring day
(April 27, 2023).
Starting with the high wind case, one observes that during periods of low net load,
the system is fairly balanced in terms of imports and exports, only exporting to
the low to negatively priced EIM market in the afternoon when wind generation
begins to climb. Additionally, the system is able to provide the required reserves
for carrying wind with only the coal and HCC units. This is due to the relatively
low level of reserves required to integrate wind, as shown in Figure 9 and Figure
10.
In the high solar case, the increased midday reserves needs shown in Figure 11
and Figure 12 coincide with high solar production. The increase in reserves needs
causes the model to start a CCGT unit, as the reserve can no longer just be
provided with hydro and coal. Bringing the CCGT unit online when there is high
solar production causes the model to make significant exports to the EIM market
during low and negatively priced hours. This, along with the start costs of the
CCGT, increases the costs of integrating solar relative to the costs of integrating
wind.
Finally, adding both high solar and high wind further exacerbates the issues that
arise during the high solar case. Due to the increase in production of wind during
the afternoon, the model must make further exports to a low and negatively
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 50 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 50 of 88
P a g e | 41 |
© 2010 Energy and Environmental Economics, Inc.
priced market. Additionally, the model turns on a CT instead of a CCGT to provide
the additional reserves required due to wind and solar.
Figure 17 presents daily operations from the imperfect foresight cases. However,
as described in Section 2.1, the difference in total market transactions and
generator costs for each case are calculated using the difference between each
case’s perfect and imperfect foresight cases. Though not shown here, on the day
shown in Figure 17, the model chooses to not start CCGTs or CTs in the respective
high solar and high wind + high solar cases in the perfect foresight cases. This is
due to the lower reserve need of the perfect foresight case.
Table 14: Summary of Results for High Solar, High Wind and High Solar + High
Wind Cases
Case Inc. Start
Costs
($Million/
yr)
Inc.
Ramping
Costs
($
Million/
yr)
Total
Inc.
Imperf.
Unit
Commit.
& Dispatch Costs
($Million/
yr)
Total
Curtail.
Costs
($Million/
yr)
Total
Inc.
Integrat.
Costs
($Million/
yr)
Total
Product.
Cost
($Million/
yr)
Total
Inc.
VER
Gen.
(GWh
/yr)
Total
Inc.
Specific
Integrat.
Costs
($/MWh)
3. Hi
Solar $0.80 $0.45 $5.78 $0.00 $7.04 $146 1,824 $3.86
5. Hi
Wind $0.35 -$0.07 $1.12 $0.00 $1.41 $143 1,823 $0.77
6. Hi
Solar +
Hi Wind
$1.63 $0.33 $7.01 $0.00 $8.96 $109 3,647 $2.46
As shown in Table 14, total incremental VER integration costs are highest in the
high solar + high wind case, followed by the high solar case and the high wind
case. However, the total specific incremental VER integration cost is lower for the
high wind + high solar than the high solar case because, while the total integration
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 51 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 51 of 88
P a g e | 42 |
cost rises with more VERs, there is also more total incremental VER generation in
the high wind + high solar case versus the high solar case.
Figure 17: High Wind vs. High Solar vs. High Solar + Hi Wind
4.4.3 HIGH SOLAR WITH LOW, AVERAGE AND HIGH HYDRO BUDGETS
This set of cases compares the effects of varying hydro budgets under high solar
conditions. On a typical year, Idaho Power derives the majority of their power
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 52 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 52 of 88
P a g e | 43 |
© 2010 Energy and Environmental Economics, Inc.
from their hydro fleet, but the total annual energy derived from hydro varies
considerably year-to-year. The simulated conditions considered in this set of
cases is depicted below in Figure 18.
Figure 18: Hydro Conditions in Low, Average and High Hydro Cases
In the model, RoR hydro is treated as an inflexible, must take resource, whereas
HCC is dispatchable. The high hydro budget case capacity factor shown in Figure
18 indicates that both HCC and RoR hydro must operate near their Pmax
throughout the year in order to not violate daily hydro energy budgets, which
greatly reduces hydro system flexibility. As shown in Figure 15, hydro conditions
are generally highest in the spring due to runoff from snow melt. Figure 19 below
compares a spring day (April 20, 2023) in which the combination of low electricity
market prices, hydro availability and VERs interact with one another.
Starting with the high hydro case, the model must sell HCC and RoR output to the
market all day, due to the high hydro budget. This includes sales during periods
of negative external market prices. Additionally, the model must start a CT to
provide solar reserves during midday. Conversely, during average hydro
conditions, this need to sell to the market at a loss is reduced, and the model
shifts HCC production to avoid selling hydro at a loss during the morning. The
model switches from using a CT to a CCGT to provide solar reserves. Finally, during
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 53 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 53 of 88
P a g e | 44 |
low hydro conditions, Idaho Power’s system can buy from the market during
negatively priced hours, but the model must run the CCGT more due to lower
hydro budgets.
Table 15: Summary of Results for High Solar with Low, Average and High Hydro
Budgets Cases
Case Inc. Start
Costs
($Million/
yr)
Inc.
Ramp
Costs
($
Million/
yr)
Total
Inc.
Imperf.
Unit
Commit.
&
Dispatch
Costs
($Million
/ yr)
Total
Curtail.
Costs
($Million
/yr)
Total
Inc.
Integrat.
Costs
($Million
/ yr)
Total
Product.
Cost
($Million
/ yr)
Total
Inc.
VER
Gen.
(GWh
/yr)
Total
Inc.
Specific
Integrat.
Costs
($/MWh)
3. Hi
Solar $0.80 $0.45 $5.78 $0.00 $7.04 $146 1,824 $3.86
4. Hi
Solar,
Low
Hydro
$0.60 $0.53 $7.16 $0.00 $8.29 $172 1,824 $4.55
8. Hi
Solar,
Hi
Hydro
$2.41 $0.19 $5.87 $0.00 $8.47 $75 1,823 $4.65
As shown in Table 15, total incremental specific VER integration costs are higher
in both the low and high hydro year cases. Moving from low to high hydro
conditions, market purchases and thermal generation decreases. This causes
production costs to decrease.
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 54 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 54 of 88
P a g e | 45 |
© 2010 Energy and Environmental Economics, Inc.
Figure 19: Low, Average and High Hydro Case Comparison
4.4.4 HIGH SOLAR WITH AND WITHOUT STORAGE
This set of cases compares the cost of integrating solar with and without battery
storage. Because Idaho Power is a vertically integrated utility, there is no ancillary
services market for these PURPA facilities. Therefore, batteries do not provide
reserves to the Idaho Power system in these cases. Additionally, the model treats
solar + storage systems having ITC-eligible battery storage. Per ITC regulations,
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 55 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 55 of 88
P a g e | 46 |
this requires storage to charge solely using solar power production. At the time
of this study’s completion, compensation rate methodologies had not been
finalized for PURPA solar + battery storage facilities pursuing contracts with Idaho
Power. Thus, the model used a simplified approach of allowing the battery to only
discharge between 4 pm and 10 pm daily. However, the model allowed the
battery dispatch to minimize total Idaho Power production costs when during the
permitted charging and discharging periods. Finally, as shown in Table 6, the
reserves needs are modeled as identical in each of these cases.
In all of these cases, the model uses a high solar build (1,355 MW of total solar),
but only the 794 MW of the solar (i.e. the incremental solar built vs. the 2023
Base Case) is coupled with an ITC-eligible battery. The differences in these cases
are as follows:
Total Battery Capacity
o High Solar Case: 0 MW
o High Solar + 200 MW Battery Case: 200 MW, 4-hour (800 MWh)
Li-Ion Battery
o High Solar + 400 MW Battery Case: 400 MW, 4-hour (1,600 MWh)
Li-Ion Battery
As can be seen in Figure 20 and Figure 21, on a typical medium-load spring day
(5/10/2023), the battery is used to move solar energy from morning and evening
solar production hours to increase net sales to the market and reduce Idaho
Power coal generation.
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 56 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 56 of 88
P a g e | 47 |
© 2010 Energy and Environmental Economics, Inc.
Figure 20: High Solar vs. High Solar + 200 MW Battery, Medium Load Spring Day
Figure 21: High Solar vs. High Solar + 400 MW Battery, Medium Load Spring Day
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 57 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 57 of 88
P a g e | 48 |
The average month-hourly dispatch of charging and discharging for the ITC-
eligible storage is depicted in Figure 22. As can be seen in each of these figures,
having greater battery capacity does not fundamentally alter when charging and
discharging occur on a given day, or across the year.
Figure 22: Month-Hourly Average Battery Charge and Discharge Power for 200
MW and 400 MW ITC-Eligible Batteries
Table 16 shows the summary of results for these cases. The total production costs
are lowest for the 400 MW battery, increasing in the 200 MW battery case and
further increasing in the no battery cas+es. However, the total specific
integration costs are lowest for the 200 MW battery size. Both storage cases
exhibit dramatically lower VER integration costs than the high solar without
storage case. This is discussed in greater detail in Section 5 of this report.
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 58 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 58 of 88
P a g e | 49 |
© 2010 Energy and Environmental Economics, Inc.
Table 16: Summary of Results for High Solar with and without Storage
Case Inc. Start
Costs
($Million/
yr)
Inc.
Ramping
Costs
($
Million/
yr)
Total
Inc.
Imperf.
Unit
Commit.
&
Dispatch
Costs
($Million/
yr)
Total
Curtail.
Costs
($Million/
yr)
Total
Inc.
Integrat.
Costs
($Million/
yr)
Total
Product.
Cost
($Million/
yr)
Total
Inc.
VER
Gen.
(GWh
/yr)
Total
Inc.
Specific
Integrat.
Costs
($/MWh)
3. Hi
Solar $0.80 $0.45 $5.78 $0.00 $7.04 $146 1,824 $3.86
9. Hi
Solar,
200
MW
Battery
$0.58 $0.02 $0.56 $0.00 $1.16 $144 1,823 $0.64
10. Hi
Solar,
400
MW
Battery
$0.58 -$0.34 $1.46 $0.00 $1.69 $142 1,823 $0.93
4.4.5 HIGH MUST TAKE SOLAR AND CURTAILABLE SOLAR CASES
Idaho Power is not able to perform economic solar curtailment of PURPA
facilities. The high must take solar and high curtailable solar cases were therefore
implemented to show how being able to economically curtail PURPA solar would
change the cost of integrating VERs.
In the high solar case, the model can only perform reliability-based curtailment,
i.e. the model will curtail VERs only when the alternative is to have unserved
energy or face some other infeasibility. In the curtailable case, the model may
economically curtail power for the incremental 794 MW of solar installed vs. the
2023 base case. This allows the model to curtail power to reduce Idaho Power’s
total production costs. There would be no difference in short-run marginal energy
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 59 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 59 of 88
P a g e | 50 |
costs from economically curtailing PURPA solar, however Idaho Power may have
to pay for the lost renewable energy credit (REC) due to curtailing solar.
Therefore, the model assumes a $20/MWh curtailment penalty, which is a typical
REC price in WECC. Similarly to the solar with storage cases, the VER reserves
needs are modeled as identical between the must take and curtailable cases.
Figure 23 and Figure 24 respectively show the difference between the must take
and curtailable cases on a low net load spring day (4/21/2023) and a high net load
summer day (7/21/2023). In Figure 23, the model chooses to curtail power both
when the external market price is below the curtailment penalty (i.e. below
-$20/MWh), as well as during the middle of the day. The model chooses to curtail
power midday because, while the market price is not below -$20/MWh, the
model performs reliability curtailment of solar in the must take case as well. In
other words, this low net load day requires VER curtailment of some sort. Total
annual curtailment in the curtailable solar case is 3.8% of potential generation for
the 794 MW of new solar. This curtailment is largely confined to spring hours,
when the net load is very low.
Alternatively, Figure 24 shows that the model does not curtail solar when solar
helps reduce total production costs. This is because solar increases net sales to a
high-priced market.
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 60 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 60 of 88
P a g e | 51 |
© 2010 Energy and Environmental Economics, Inc.
Figure 23: High Must Take Solar and High Curtailable Solar, Low Load Day
Figure 24: High Must Take Solar vs. High Curtailable Solar, High Load Day
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 61 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 61 of 88
P a g e | 52 |
Table 17 shows that while the total incremental specific integration cost is lower
in the curtailable solar case than the must take solar case, the total production
costs are essentially identical between the two cases.
Table 17: Summary of Results for High Must Take and Curtailable Solar
Case Inc. Start
Costs
($Million/ yr)
Inc.
Ramping
Costs
($ Million/ yr)
Total
Inc.
Imperf. Unit Commit. & Dispatch Costs
($Million/
yr)
Total
Curtail.
Costs
($Million/yr)
Total
Inc.
Integrat. Costs
($Million/ yr)
Total
Product.
Cost ($Million/ yr)
Total
Inc.
VER Gen.
(GWh /yr)
Total
Inc.
Specific Integrat. Costs
($/MWh)
3. Hi
Solar $0.80 $0.45 $5.78 $0.00 $7.04 $146 1,824 $3.86
11. Hi
Curtail.
Solar
$0.72 $0.39 $4.31 $0.29 $5.71 $147 1,823 $3.13
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 62 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 62 of 88
P a g e | 53 |
© 2010 Energy and Environmental Economics, Inc.
5 Discussion
5.1 Discussion of Current Study Results
E3’s results provide several high-level insights about integrating VERs:
Integration costs are driven by the need for procuring system flexibility
on dispatchable generators during periods of low net load
Integrating solar is more expensive that integrating new wind resources
VER integration costs can be lowered by adding flexibility to the Idaho
Power system, such as battery storage, allowing economic curtailment
and reducing the must-run thermal Pmin of the system
VER integration costs increase during abnormal hydro conditions (low or
high annual budgets)
The integration costs found in this 2020 Idaho Power VER integration
study are lower than the 2018 Idaho Power Variable Energy Resource
Analysis
These results are discussed in more detail below.
5.1.1 EFFECTS OF BINDING PMIN CONSTRAINTS ON VER INTEGRATION
COSTS
As discussed in Section 3.2, as more VERs are added to Idaho Power’s system, the
aggregate reserve and flexibility needs tend to increase. Only HCC, coal, CTs and
CCGTs are modeled as eligible to provide reserves. Because all these generators
have a non-zero Pmin, the aggregate thermal Pmin grows when more generators
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 63 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 63 of 88
P a g e | 54 |
are brought online to provide reserves. Idaho Power has a large penetration of
PURPA VERs, which are treated as must take units by Idaho Power. When these
must take resources produce large amounts of power, the net load on Idaho
Power’s system can fall to very low values. In order to maintain supply-demand
equilibria on Idaho Power’s system, Idaho Power must export power to the
market when the aggregate system Pmin, plus the required system footroom, is
greater than the system net load. This is depicted schematically below in Figure
25.
Figure 25: Effects of Additional Solar on Unit Commitment and Market
Transactions
During these “binding Pmin” events, exporting power to the market does not by
itself cause VER integration costs to rise. However, due to the growing
penetration of solar across the EIM footprint, 2023 EIM market prices are
projected to be, on average, below typical marginal thermal unit generation costs
during daytime hours in the spring and fall, as shown in Figure 26. These periods
of low EIM prices are also when Idaho Power’s solar generators will be producing
enough power to significantly lower Idaho Power’s net load to binding Pmin
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 64 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 64 of 88
P a g e | 55 |
© 2010 Energy and Environmental Economics, Inc.
levels. Therefore, under high solar builds, Idaho Power is often exporting power
at a financial loss to a low- or negative-priced EIM market. At other times, Idaho
Power may have to shift its hydro production to non-optimal hours (e.g. away
from times when hydro could earn the greatest amount of export revenues) in
order to provide sufficient flexibility on HCC while adhering to the HCC daily
energy budget.
Figure 26: Month-Hourly Average 2023 EIM Market Prices
As shown in Section 3.2, in contrast to the High Solar case, in the High Wind case,
the reserves profile is more uniform across time. Additionally, the period of
highest reserves needs do not necessarily coincide with low net loads resulting
from high Idaho Power wind production because Idaho Power wind production
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 65 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 65 of 88
P a g e | 56 |
tends to be highest during wintertime evenings. This results in fewer binding Pmin
intervals in the High Wind case that force suboptimal market transactions.
Not retiring a Bridger unit and high hydro conditions increases the cost of
integrating new solar. In these cases, having higher levels of must run coal or must
take hydro has the effect of decreasing the solar production level at which these
binding Pmin events take place.
As shown in
Table 12, the VER integration costs are typically dominated by the costs of
imperfect unit commitment and dispatch costs. Therefore, the reader can largely
focus on periods in which these binding Pmin events occur when seeking to
understand what drives integration costs for the different cases.
5.1.2 HIGH SOLAR WITH STORAGE CASES
A paradoxical finding of this analysis is that the total specific integration cost of solar
is lower for the High Solar + 200 MW Battery case than the High Solar + 400 MW
Battery case.
The reason for this is due to the way in which this study calculates VER integration
costs. As discussed in Section 2.1, the VER integration costs are calculated as the
sum of the ramping and start costs, plus the total imperfect unit commitment and
dispatch costs. The total imperfect unit commitment and dispatch cost is calculated
for each case as the difference of production costs for the imperfect foresight and
perfect foresight cases. The only difference between these cases is how much VER
forecast error, subhourly VER variability and reserves are carried for the
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 66 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 66 of 88
P a g e | 57 |
© 2010 Energy and Environmental Economics, Inc.
incremental VER build. Due to its greater capacity, the larger 400 MW battery
allows for a greater production cost savings than the 200 MW battery when moving
from the imperfect foresight to the perfect foresight case. This larger savings is
added into the integration cost. Therefore, the apparent integration cost is higher
for the 400 MW battery than the 200 MW battery. However, there are limitations
to how this study was able to model a PURPA solar + ITC-enabled solar fleet in
PLEXOS. These limitations are discussed below.
The PLEXOS model used to calculate Idaho Power’s VER integration costs has
multiple stages that reflect different levels of uncertainty the DA, HA, RT15, and
RT5 time intervals. Storage dispatch can change between the stages due to
different grid conditions and solar forecasts. If storage provides more flexibility
ahead of real time, it can leave real-time dispatch with lower levels of flexibility, or
vice versa. The difference between storage dispatch in perfect and imperfect
foresight cases, propagated through multiple modeling time horizons, results in the
potential for small, unexpected swings in VER integration costs. Considerations
with respect to storage scheduling include:
Storage scheduling between different commitment timeframes will
evolve as more storage is deployed. Currently, there is not a standard
practice for battery storage scheduling
The scheduling of PURPA-contracted storage over multiple timeframes is
especially uncertain given the lack of experience with this type of
resource
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 67 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 67 of 88
P a g e | 58 |
The scheduling of PURPA-contracted storage in a perfect foresight
counterfactual will never be known with any precision because grids are
not operated with perfect foresight.
The impact of storage sizing on unit commitment may be non-linear – a bigger
battery may cause a large Idaho Power unit to alter its commitment schedule
whereas a small battery would not be able to cause as big of an impact.
Additionally, the interaction between storage dispatch and Idaho Power market
revenues can create significant swings in the VER integration cost. The extent to
which Idaho Power has control over PURPA-contracted battery operations can
impact market revenues, especially during periods of extreme EIM prices.
The considerations above imply that there is uncertainty around future PURPA-
contracted storage dispatch and VER integration costs. E3 has included many of
the relevant dynamics in the PLEXOS model, and believes that the two integration
cost calculations for storage are within reasonable bounds of error given what is
known currently about PURPA-contracted storage. However, E3 believes it is
appropriate to use the results from these two cases to derive an average solar +
storage VER integration cost, rather than assign discrete values to different storage
capacities.
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 68 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 68 of 88
P a g e | 59 |
© 2010 Energy and Environmental Economics, Inc.
5.2 Comparison to Data in Literature and 2018 Idaho
Power VER Study
In its Western Wind and Solar Integration Study: Phase 211, NREL calculated
integration costs for up to 33 percent penetration of wind and solar in the
Western Interconnection. The summary integration costs by scenario from the
NREL study, the 2018 Idaho Power VER integration study and this study are shown
below in Table 18, in 2020 dollars. Generally, it can be seen that the values from
this study vary considerably more than the values from the NREL study. The NREL
study integrated wind and solar across the Western Interconnection versus a
small individual balancing area, and did not use the same reserves derivation
process as this study. Modeling the entire Western Interconnection meant that
NREL did not assess the effects of suboptimal market trades on integration costs
at the interconnection footprint level. Additionally, the greater resource diversity
across the entire Western Interconnection likely reduces specific VER integration
costs. However, the general takeaway from this modeling is that VER integration
costs in the 2018 and 2020 Idaho Power VER integration studies are generally
higher than those from prior NREL work.
11 https://www.nrel.gov/docs/fy13osti/55588.pdf
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 69 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 69 of 88
P a g e | 60 |
Table 18: Comparison of 2020 Idaho Power VER Study Results to Other VER
Integration Cost Results
Case Total percent of
Annual Load
Supplied by VERs
(Total VER
Generation/Gross
Load)
Specific
Integration
Cost, Low
Bound
(2020$/MWh
VER)
NREL High Wind 33 % $0.25-0.75
NREL High Solar 33 % $0.22-0.56
NREL Mixed Resources 33 % $0.16-0.43
2020 Idaho Power VER Study High Solar Cases
(no storage or curtailment allowed) 28 % $3.86-4.65
2020 Idaho Power VER Study High Wind Case 28 % $0.77
2020 Idaho Power VER Study High Wind and
Solar Case 38 % $2.46
2018 Idaho Power VER Study 1,000 MW of
Wind Case 14 % $6.17
5.3 Methodological Differences between 2020 and
2018 Idaho Power Company Variable Energy
Resource Analysis
5.3.1 OVERVIEW
The incremental integration costs shown in this study are lower than those from
the 2018 Idaho Variable Energy Resource Analysis. While it was not in scope for
E3 to perform a detailed analysis of the 2018 study and how its methodology
differed from that of this analysis, several things stand out as important
differences between the two studies.
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 70 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 70 of 88
P a g e | 61 |
© 2010 Energy and Environmental Economics, Inc.
5.3.2 RESERVES
The 2018 study calculates reserves in a very different manner than in the 2020
study. The resulting average reserves levels are higher in the 2018 study than
those investigated in the 2020 study. The 2020 study includes CAISO FRP reserves,
regulation reserves and contingency reserves. The 2018 study included regulation
reserves and contingency reserves, but the regulation reserves were calculated
differently.
In the 2020 study, to derive the CAISO FRP reserves, E3 used a method that
approximates the method used to derive the CAISO FRP within reasonable
bounds.12 The CAISO FRP has RT15 and RT5 stages. For the RT15 stage, E3
calculated the uncertainty component of the FRP using the difference between
2019 HA forecast net load and RT5 actual net load. Similarly to CAISO’s derivation
methodology, E3 then binned this net load forecast error by month-hour and
used a 95 percent confidence interval (as does CAISO) to determine headroom
and footroom components of the uncertainty reserves. After capping these net
load-based reserves using P98 and P2 values for footroom and headroom,
respectively, E3 assumes a 40 percent diversity credit to reduce the uncertainty
component by the same percentage in all hours, based on historical levels of EIM
footprint diversity. This 40 percent value approximates the caps and “credit”
system that the CAISO FRP uses.13 Finally, E3 calculates the RT5 CAISO FRP using
12 See, e.g. http://www.caiso.com/InitiativeDocuments/DMMResourceSufficiencyEvaluationPresentation-
EnergyImbalanceMarketofferRulesTechnicalWorkshop.pdf for a description of CAISO FRR components.
13 See, e.g. http://www.caiso.com/InitiativeDocuments/DMMResourceSufficiencyEvaluationPresentation-
EnergyImbalanceMarketofferRulesTechnicalWorkshop.pdf for a description of CAISO FRR components.
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 71 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 71 of 88
P a g e | 62 |
historical data derived from the ratio of 2019 CAISO RT5 FRP uncertainty reserves
to the 2019 CAISO RT15 FRP uncertainty reserves.14
E3 calculates regulation reserves for the individual load, wind and solar profiles
using a persistence forecast of the 5-minute data. Solar data are then binned by
season, hour and percent output, whereas load and wind are binned by percent
of maximum observed load and output, respectively. A 95 percent confidence
interval is then used to derive headroom and footroom needs for these reserves,
and they are then combined using a root mean square, assuming that the load,
wind and solar regulation components have no covariance on this short
timescale. Finally, spinning contingency reserves are calculated at 3 percent of
load. This results in the average reserves shown below in Table 19.
Table 19: Reserves Summary for Different 2020 Idaho Power VER Integration
Cost Cases
Case Total
MW Online Wind
(MW)
Total
MW Online Solar
(MW)
Avg.
RT15 FRP Up
(MW)
Avg.
RT15 FRP Down
(MW)
Avg.
Reg. Up
(MW)
Avg.
Reg. Down
(MW)
Avg.
Conting.
Res.
(MW)
Avg.
Total Res. Up
(Percent
of Avg. Load)
Avg.
Total Reserves Down
(Percent of Avg. Load)
1. 2023
Base
Case
728 561 100 97 40 41 104 13 % 7 %
2. Jim
Bridger
Online
728 561 100 97 40 41 104 13 % 7 %
3. Hi
Solar 728 1,354 147 142 71 72 104 17 % 11 %
14 http://oasis.caiso.com/mrioasis/logon.do
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 72 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 72 of 88
P a g e | 63 |
© 2010 Energy and Environmental Economics, Inc.
4. Hi
Solar,
Low
Hydro
728 1,354 147 142 71 72 104 17 % 11 %
5. Hi
Wind 1,396 561 152 147 50 52 104 16 % 10 %
6. Hi
Solar,
Hi Wind
1,396 1,354 193 186 79 81 104 19 % 13 %
7.
Existing
Solar
Base
Case
728 561 87 86 32 33 104 11% 6%
8. Hi
Solar,
Hi
Hydro
728 1,354 147 142 71 72 104 17 % 11 %
9. Hi
Solar,
200
MW
Battery
728 1,354 147 142 71 72 104 17 % 11 %
10. Hi
Solar,
400
MW
Battery
728 1,354 147 142 71 72 104 17 % 11 %
11. Hi
Curtail.
Solar
728 1,354 147 142 71 72 104 17 % 11 %
In the 2018 study, Idaho Power calculated the regulation reserves using 2HA
forecasted wind and load, and 1-minute actual wind and load data. These data
were then binned by percentage of wind output or maximum load. It is not clear
from the study if confidence intervals are subsequently applied to this data, but
the resulting reserves, as a percentage of normalized load, are shown below as
Table 20 and Table 21. Spinning reserves are calculated as 3 % of the hourly load,
which is identical to the method E3 used.
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 73 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 73 of 88
P a g e | 64 |
Table 20: 2018 Idaho Power VER Integration Study Wind Reserves
Winter Spring Summer Fall
Wind Quartile of Forec. Output
Reg Up % of Avg Wind Forec.
Reg Down % of Avg (Namplate – Forec.)
Reg Up % of Avg Wind Forec.
Reg Down % of Avg (Namplate – Forec.)
Reg Up % of Avg Wind Forec.
Reg Down % of Avg (Namplate – Forec.)
Reg Up % of Avg Wind Forec.
Reg Down % of Avg (Namplate – Forec.)
1. 100% 28 % 100% 62 % 100 % 48 % 100 % 66 %
2. 86 % 51 % 94 % 79 % 93 % 75 % 80 % 65 %
3. 55 % 65 % 71 % 81 % 68 % 85 % 76 % 75 %
4. 49 % 34 % 43 % 69 % 59 % 82 % 39 % 43 %
As shown in Table 20 and Table 21, the 2018 study had much higher reserves than
the 2020 study, particularly for VERs. This likely results in higher costs for
integrating VERs in the 2018 study, due to the high reserves levels causing more
binding Pmin constraints for a given VER penetration level.
Table 21: 2018 Idaho Power VER Integration Study Load Reserves
Winter Spring Summer Fall
Load Quartile of Forecast Maximum
Reg Up % of Avg Load
Reg Down % of Avg
Load
Reg Up % of Avg Load
Reg Down % of Avg Load
Reg Up % of Avg Load
Reg Down % of Avg
Load
Reg Up % of Avg Load
Reg Down % of Avg Load
1. 4.9 % 9.1 % 8.1 % 10.5 % 7.9 % 11.5 % 8.0 % 10.6 %
2. 9.3 % 6.8 % 6.8 % 11.3 % 8.1 % 6.0 % 7.5 % 8.9 %
3. 9.5 % 5.8 % 9.9 % 6.7 % 9.7 % 9.8 % 9.9 % 8.5 %
4. 7.9 % 6.9 % 8.3 % 7.0 % 6.2 % 13.3 % 7.3 % 7.1 %
E3 believes that its 2020 reserve derivation methodology is closer to standard
practice than the method used in the 2018 study. There was negligible observed
unserved energy in E3’s models. Similar normalized levels of reserves (MW per
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 74 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 74 of 88
P a g e | 65 |
© 2010 Energy and Environmental Economics, Inc.
MW of installed VERs) and confidence intervals of historical forecast error have
been used elsewhere.15 16 17
In both the 2018 study and the 2020 study, there were a significant number of
hours in which the AURORA and PLEXOS models were unable to hold sufficient
reserves to meet the requirements outlined above. In the PLEXOS model, the
reserve violation penalties were set up such that regulation reserves were
typically not met whereas CAISO FRP reserves and contingency reserves were
nearly always met.
5.3.3 TREATMENT OF EXTERNAL MARKETS
The 2020 study is modeled with an EIM market, whereas the 2018 study is not.
Because Idaho Power joined the EIM in Q2 2018, this omission was reasonable in
the 2018 study. In the 2020 study, the presence of the EIM market allows the
model to balance forecast error from the DA and HA intervals to the real time.
The 2018 model had less flexibility in its ability to trade, which likely reduces the
ability of Idaho Power’s system to buy and sell from the market to enable
procuring reserves relative to a scenario with the EIM.
5.3.4 MULTISTAGE VS. SINGLE STAGE MODEL
The 2020 study used a multistage PLEXOS model, which contains information
about typical net load forecast error and subhourly net load variability, whereas
15 Z. Zhou, T. Levin, G. Conzelmann, “Survey of U.S. Ancillary Services Markets.”
https://publications.anl.gov/anlpubs/2016/01/124217.pdf
16http://www.ercot.com/content/wcm/key_documents_lists/137978/9_2019_Methodology_for_Determining_Mini
mum_Ancillary_Service_Requirements.pdf
17 http://www.caiso.com/Documents/Addendum-DraftFinalTechnicalAppendix-FlexibleRampingProduct.pdf
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 75 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 75 of 88
P a g e | 66 |
the 2018 study used a single hourly stage AURORA model that did not reflect
forecast error. In executing its multistage PLEXOS model, E3 did not observe
significant levels of unserved energy. Therefore E3 believes its reserves derivation
method provides reasonable reserve levels.
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 76 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 76 of 88
P a g e | 67 |
© 2010 Energy and Environmental Economics, Inc.
6 Conclusions
6.1 Integration Costs
Overall, it was found that integration costs for new VERs on Idaho Power’s system
vary from $0.64/MWh up to $4.65/MWh. Generally, solar integration costs are
significantly higher than those for new wind. Adding more must-run resources,
such as hydro operating at very high capacity factors, or keeping must run thermal
units online, increases VER integration costs. Increasing system flexibility, such as
by pairing solar with dispatchable storage, or by allowing solar to be economically
curtailed, reduces VER integration costs.
Additionally, the VER integration costs found herein are significantly lower than
those from the 2018 Idaho Power VER integration study. This is due to multiple
factors, but likely the single greatest cause is the reduction in growth in reserves
per unit of online wind and solar capacity in the 2020 study versus the 2018 study.
Finally, the results from this study are contingent upon VERs being must take; coal
units being committed as baseload, must run units; maintaining strategies for
deploying Idaho Power’s HCC hydroelectric resources; storage paired with solar
not being able to provide reserves; and other assumptions about current
practices that may change in the future.
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 77 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 77 of 88
P a g e | 68 |
7 Appendix 1: Process
Document
7.1 Introduction
This Appendix is provided as a guide to further understand how E3 developed its
PLEXOS model for this study.
The production cost simulation software, PLEXOS, was used to calculate VER
integration costs in this study. This was done by using PLEXOS to generate the
outputs necessary to derive the VER integration costs: start/stop costs, ramping
cost, imperfect unit commitment and dispatch fuel costs, imperfect unit
commitment and dispatch net import costs and curtailment costs.
To yield results, PLEXOS requires various inputs into E3’s four stage model. The
inputs to the PLEXOS model were developed by E3, Idaho Power, and in some
instances in collaboration between Idaho Power and E3. These include:
Load Profiles: The 2019 profiles were developed by Idaho Power and E3
and consist of 4 comma separated value (CSV) files to represent load
forecasts at the DA, HA, and RT15 stages with the RT5 profile seen as the
actual load profile, and these were scaled to 2023 load profiles by E3.
Renewable Profiles: Solar and wind profiles were developed by E3 using
Idaho Power’s data and consist of 4 CSV files to represent generation
forecasts at the DA, HA, and RT15 stages with the RT5 profile seen as the
actual output.
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 78 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 78 of 88
P a g e | 69 |
© 2010 Energy and Environmental Economics, Inc.
Hydro Profiles: Daily hydro budgets were created by E3 using Idaho
Power’s historical hydro data, and Pmax/Pmin levels were derived using
Idaho Power input. These are fed into the model using separate CSVs for
daily HCC maximum power, daily HCC minimum power, daily HCC energy
budget and daily RoR power outputs
Generator Characteristics: Generator characteristics were provided by
Idaho Power as E3’s part of the data collection process and include
properties such as maximum and minimum capacity, ramp rates, start-up
costs, VO&M, as well as any must-run flags or particular generating
patterns. These are input for each generator using the PLEXOS UI.
Reserve Policies and Profiles: Reserve profiles for the “perfect foresight”
and “imperfect foresight” cases were developed using E3’s RESERVE tool,
along with the renewable and load profiles provided by E3. Each case has
its own set of reserve profiles, which are in the form of CSVs read in for
the flexible ramping requirement and the regulation needs. Contingency
reserves are enforced within the PLEXOS UI.
Topology and Transmission: The transmission and zonal topology of the
model was created by E3 with input from Idaho Power towards
transmission capacity to the Mid C and PV market nodes. These limits and
the topology were input to the PLEXOS UI.
Markets: Market transaction limits were provided by Idaho Power for the
two markets nodes, Mid C and PV, represented within this model.
Forward Q2-Q4 2019 and Q1 2020 market prices were provided to E3 by
Idaho Power, and E3 downloaded historical Q2-Q4 2019 and Q1 2020 EIM
market prices. These prices are then modified using E3’s in-house
AURORA price forecasts to adjust them to 2023 expected market prices.
These adjusted prices are fed into the model using CSVs for each market
and model stage.
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 79 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 79 of 88
P a g e | 70 |
Fuel Prices: Fuel prices were provided to E3 for each of the generators,
and are enforced within the PLEXOS UI.
When running a case within PLEXOS, it is important to ensure that the appropriate
renewable profiles are added as data files in the model. These are found in the
‘Wind Profiles’ and ‘Solar Profiles’ subfolders within the ‘Data’ directory and ‘Data
Files’ folder illustrated in Figure 27. In addition, if need be, updated reserve
profiles must also be added to the PLEXOS model. These data files are named to
correspond with the relevant case they will be used for and can be found under
the ‘Reserves Idaho Power’ subfolder in the ‘Data’ directory and within the ‘Data
Files’ folder. Daily hydro budget profiles can be added or adjusted within the
‘Hydro Budgets’ subfolder within the ‘Data Files’ folder.
Figure 27: PLEXOS Data Directory
Creating a new case or editing an existing case’s properties can be done within
the PLEXOS UI’s ‘Scenarios’ folder seen in Figure 28 under ‘Idaho Power Core
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 80 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 80 of 88
P a g e | 71 |
© 2010 Energy and Environmental Economics, Inc.
Cases’. Each Scenario represents an individual case. The properties that are
tagged with this case ‘Scenario’ will only be used if this case is being run.
Figure 28: PLEXOS Scenario Directory
A specific case is only run if the ‘Scenario’ associated with it is included in the
‘Membership’ of each monthly stage model (DA, HA, RT15, RT5) and can be
identified as shown in Figure 29. Only one ‘Idaho Power Core Cases’ ‘Scenario’
can be linked to the models at any one time. If multiple case ‘Scenarios’ are
included in the model ‘Memberships’, errors may occur while attempting to
execute the full model or may yield incorrect results.
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 81 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 81 of 88
P a g e | 72 |
Figure 29: PLEXOS Membership view
To derive VER integration costs, the overall PLEXOS model is run twice for each
case, once using the perfect foresight profiles for the relevant VER resources and
reserves, and then once using the imperfect foresight reserve and VER profiles.
The individual cases are expressed as individual PLEXOS models with custom
modifications and, in some instances, CSV files. The primary differences between
the cases are described below.
Case 1 is the 2023 base case for Cases 3-6 and Cases 8-11, which has all
known unit additions and retirements and also includes the known 2019
through 2023 load growth. The Solar and Wind objects are scaled to the
appropriate size for Case 1
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 82 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 82 of 88
P a g e | 73 |
© 2010 Energy and Environmental Economics, Inc.
Case 2 explores the effect of not retiring one of the Bridger coal plant’s
units, but is otherwise identical to Case 1. The Bridger coal plant Pmin
and Pmax are increased to reflect this change
Case 3 builds on Case 1 by exploring the effect of adding enough new
solar (794 MW of new solar) such that 10 percent of the 2023 Idaho
Power average gross load is provided by this new solar build. This is done
using the existing aggregated solar plant from Case 1
Case 4 extends the Case 3 analysis to a low, rather than average hydro
year. The hydro budgets and daily Pmin/Pmax levels are updated using
the CSVs fed into the model
Case 5 builds on Case 1 and explores the integration costs of a high wind
build. Case 5 assumes a new wind build that can supply 10 percent of the
annual 2023 Idaho Power gross load (669 MW of new wind). This is
performed using the existing wind object from Case 1
Case 6 builds on Case 3 and Case 5, including both high solar and high
wind builds (794 MW of new solar and 669 MW of new wind). This is done
using the existing solar and wind objects from Case 1
Case 7 is identical to Case 1, except that none of proposed solar additions
come online from 2019 to 2023, resulting in 251 MW fewer of solar than
Case 1 and lower reserves needs. This is done using the existing solar
object from Case 1
Cases 8 extends the Case 3 analysis to a high, rather than average hydro
year, and as in Case 4, this is accomplished by feeding in different CSVs
to adjust the energy budgets and Pmax/Pmin levels
Case 9 builds on Case 3 by adding a 200 MW 4-hour Battery object with
a roundtrip efficiency of 85% and can only charge from the additional 794
MW of new solar
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 83 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 83 of 88
P a g e | 74 |
Case 10 adds a 400 MW 4-hour Battery object with an 85% roundtrip
efficiency and is only able to charge from the additional 794 MW of new
solar
Case 11 splits the solar object in Case 3 into two distinct generator
objects: an ‘Idaho Solar’ and ‘Idaho Solar Curtailable’. The ‘Idaho Solar’
resource is modeled as must-take, while the ‘Idaho Solar Curtailable’
object is allowed economically curtail
7.2 Results Processing
The results viewer enables us to display annual PLEXOS ST data in a more user-
friendly format and consists of several different tabs. Below, we explain how to
navigate and manipulate each tab in the order of their use when processing
results:
Cover: this tab provides a high-level overview of the workbook and is not
of any practical use in processing results
Params: The Params tab is used as a library that the embedded excel
macro will read and use to pull outputs from individual properties in the
PLEXOS solutions zip files. The ‘ParentClassName’ column corresponds to
the tabs within the PLEXOS UI either ‘System’ or ‘Simulation’ as seen in
Figure 29. The ‘ParentName’ is the system name within PLEXOS which is
given as ‘IPC’ in this model. ‘ChildClassName’ is the subfolder name
within any of the ‘Production’, ‘Transmission’, ‘Generic’, ‘Data’ folders.
For example, ‘Generators’ or ‘Lines’. The ‘PropertyName’ column is the
name of the property to be output to the results viewer. ‘ChildName’ is
the name of the object that the output property belongs to. If the
generation of a generator called ‘GEN1’ needed to be brought into the
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 84 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 84 of 88
P a g e | 75 |
© 2010 Energy and Environmental Economics, Inc.
results viewer then the ‘PropertyName’ would be ‘Generation’ and the
‘ChildName’ would be ‘GEN1’.
Figure 30: PLEXOS UI
If pulling in individual object properties, the ‘AggregrationEnum_type’
column by default should be input as ‘AggregationEnum_None’ and the
‘agg_category’ column should be left blank; however if it is more
beneficial to load properties from all objects within a subfolder of the
‘ChildClassName’ folders such as ‘IPC Solar’ as seen in Figure 30, then it is
possible to do this by leaving the ‘ChildName’ column blank, changing the
‘AggregrationEnum_type’ column entry to ‘AggregationEnum_Category’,
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 85 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 85 of 88
P a g e | 76 |
and changing the ‘agg_category’ entry to ‘IPC Solar’. Finally, the ‘Units’
column should contain the units of the property that is being selected.
One should ensure that the properties that are being listed in the Params
tab in the results viewer are being output by the PLEXOS model. It is
possible to verify and, if need be, add the property to be output as part
of the PLEXOS solution zip file through the PLEXOS UI. As seen in Figure
31, by clicking on the ‘Simulation’ tab in the PLEXOS UI and double clicking
on the object within the ‘Reports’ subfolder, the ‘Field List’ tab will show
the entire list of possible outputs from the model.
Figure 31: PLEXOS Reports
Ensure that the desired outputs have the ‘Period’, or ‘Flat File’ boxes
checked. PLEXOS Help documentation is extremely thorough in providing
additional detail in understanding the full amount of available output
properties. This must be done before running the models to ensure that
the selected outputs are created in the PLEXOS solution zip files.
Control: Once the desired outputs are set in the ‘Params’ tab, the results
viewer can be run. The ‘Control’ tab contains a few cells that must be filled
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 86 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 86 of 88
P a g e | 77 |
© 2010 Energy and Environmental Economics, Inc.
out before running the Macro. The ‘Start Solution Month’ and ‘End
Solution Month’ allows the flexibility to run the results viewer for one
month or a set of months if need be, though use caution as the results
viewer capacity factor calculations are set up to calculate over the whole
year so may yield incorrect results if not run over the whole year. In
addition, ensure that the ‘Stage Name’ and ‘Model Name Constant’ inputs
are aligned with the model names as seen in Figure 32, where the ‘Stage
Name’ is ‘RealTime5’ and the ‘Model Name Constant’ is ‘IPC’. The rest of
the values within the ‘Control’ tab should not be touched. Ensure
calculations within the workbook are set to manual and then click the ‘Do
all the PLEXOS things NOW!’ button to start the results viewer.
Figure 32 PLEXOS Model Naming Convention
TimeSeries Data: Once the results viewer is finished compiling the
PLEXOS outputs these will all appear in the ‘TimeSeries Data’ tab.
Plot: The ‘Plot’ tab provides dispatch plots, price plots, and market
transaction plots of a user-selected date. The day chosen can be toggled
between any days represented within the output data. The ‘Plot’ tab also
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 87 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 87 of 88
P a g e | 78 |
provides an annual look at capacity factor, cost, generation, number of
starts by generator and provides annual cost and generation figures
associated with market transactions to provide an overall production cost
for the system over the year.
Month-Hour Summary: This tab converts the 5-minute data within the
‘TimeSeries Data’ tab to hourly average values which is then used to
create heat maps.
Month-Hour: This tab is used as a data visualizing tool to display output
data as month-hour average heat maps. The data being shown in the heat
map can be toggled by the user via the dropdown menu.
SummaryAll: The ‘SummaryAll’ tab offers a quick average value of each
of the properties listed in the ‘Params’ tab.
Hydro Budget: This tab provides information on Hells Canyon Complex
hydro budgets.
Conversion: This tab provides conversion figures within the workbook.
Appendix 4.17 - Idaho Power 2020 VER Integration Study Page 88 of 88
Exhibit No. 5 Case No. IPC-E-23-14 J. Ellsworth, IPC Page 88 of 88