20131129DIRECT P.DeVol.pdf

BEFORE THE TDAHO PUBLIC UTILITIES COMMTSSTON
IN THE MATTER OF THE APPLICATION
OF TDAHO POWER COMPANY TO
UPDATE TTS WIND INTEGRATION RATES
AND CHARGES.
CASE NO. TPC-E-I3_22
)
)
)
)
)
IDAHO POTnIER COMPANY
DIRECT TESTIMONY
OF
PHILIP B. DeVOL
1 Q. Please state your name and business address.
2 A. My name is Philip B. DeVol- and my business
3 address is !227 West Idaho Street, Boise, Idaho 83702.
4 Q. By whom are you employed and in what capacity?
5 A. I am employed by Idaho Power Company ("Idaho
6 Power" or "Company") as the Resource Planning Leader.
7 Q. Please describe your educational background
8 and work experience with Idaho Power.
9 A. In May of 1989, I received a Bachelor of
10 Science Degree in Mathematics from Miami University in
11 Oxford, Ohio. I then received a Master of Scj-ence Degree
1,2 in Biostatistics from the University of Mlchigan in May of
13 1991.
74 O. Please describe your work history at Idaho
15 Power.
76 A. I began my employment with Idaho Power in 2001,
77 as an Engineering Specialist in the Water Management
18 Department. In this position, I was responsible for
19 modeling of the Idaho Power hydroelectric system for the
20 Integrated Resource Pl-an ("fRP") and relicensing studies.
27 In 2004, I became a Water Management Operations Analyst
22 where I continued to be responsible for hydroel-ectric
23 system modeling.
24 In 2005, I became a Planning Analyst in the Power
25 Supply Planning Department. fn this position, I was
DeVOL, DI 1
Idaho Power Company
t_
2
3
4
5
6
7
8
9
10
11
L2
13
L4
15
16
t1
18
t9
20
2t
22
23
24
25
responsible for the compilation of the Idaho Power long-
term operating plan prepared on a monthly basis as part of
the Company's plan for managing risk. My duties in this
position also expanded to include the study of wind
integration.
I became the Power Supply Planning Leader in 20L0
and Resource Planning Leader in 2073. My duties i-n these
positions have included project management for the most
recent Idaho Power wlnd integration study.
I have been involved in regional and national
proceedings related to the study of wind integration. I
partici-pated in methodology discussions for the 2007 Wind
Integration Actj-on Plan produced by the Northwest Wind
Integration Forum. I have attended numerous Utility Wind
Integration Group ("UWIG") workshops, and presented at UWIG
workshops in Okl-ahoma City in 2006 and Portland, Oregon, in
20A7. I al-so presented to the Idaho Wind Working Group at
their September 207L meeting. Eina11y, earlier this month,
I presented at a Centre for Energy Advancement through
Technol-ogical Innovatj-on (*CEATI") workshop focused on
forecasting uncertainties for renewable energy supply.
O. What is the purpose of your testimony in this
matter?
A. Idaho Power is
Public Utilities Commission
requesting that the Idaho
("Commission" ) author j-ze the
DeVOL, DI 2
Idaho Power Company
1
2
3
4
5
6
7
I
Y
10
11
t2
13
74
15
76
I7
18
19
20
21
22
23
24
25
Company to update its wind integration rates and charges
consistent with its 20L3 Wind Integration Study Report.
The purpose of my testimony is to provide the Commission
with information regarding the design and execution of the
study and to provide the resul-ts.
I. 2OL3 IIIITD INTEGRATION STUDY
O. Has Idaho Power updated the initial wind
integration study that was filed, along with its addendum,
in 2007?
A.Yes. Idaho Power has conducted an updated
wind integration study (*2013 Study"). Idaho Power filed
this wind integration study with the Commission on Eebruary
14, 2013, with its 20Ll IRP Update informational filing,
Case No. IPC-E-11-11. The 201,3 Study is attached hereto as
Exhibit No. 1.
O.Has the 20L3 Study been updated to incorporate
inputs from the 20L3 IRP?
A.Yes. The 201,3 Study was conducted using
inputs from the 20Ll IRP. Subsequent to the development of
the 20L3 Study, the Company has filed its 2013 IRP. The
Company has updated the 2013 Study based upon inputs from
the 2013 IRP, including the load forecast, Mid-C electric
market prices, natural gas prices forecast, and the coal-
price forecast ("Updated 2013 Study").
DeVOL, Df 3
Idaho Power Company
1 Q. Please provide a high level descripti-on of the
2 Company's 20L3 Study.
3 A. The Company's 2013 Study determined wind
4 integration costs for installed capacities of 800 megawatts
5 ("MW"); 1,000 MW; and 1,200 MW. Synthetic wind generation
5 data and corresponding day-ahead wind generation forecasts
1 at these build-outs were provided by 3TIER and Energy
8 Exemplar (formerly PLEXOS Solutions).
9 The 2073 Study employed the foll-owing two-scenario
10 design:
11 o Base scenario for which the system is not
t2 burdened wlth the incremental bal-ancing reserves necessary
13 for integrating wind; and
14 o Test scenario for which the system is
15 burdened with the incremental balancing reserves necessary
76 for integrating wind.
L7 System simul-ations for the two scenarios were
18 identi-cal-, except that generation scheduling for the test
19 scenario incl-uded the condition that dispatchabl-e thermal
20 and hydro generators must provide the appropriate amount of
27 incremental balancing reserves.
22 System simul-ations were conducted for a 2071 test
23 year. Customer demand for 2077, ds projected for the 20LL
24 IRP, was used in system modeling. To investigate the
25 effect of water conditions on wind integration, the 20L3
DeVOL, DI 4
Idaho Power Company
1 Study also considered Snake River Basin stream flows for
2 three separate historic years representing 1ow (2004),
3 average (2009) , and high (2006) water years. Fina11y, the
4 natural gas price and Mid-C wholesale electric market
5 prices as forecast for 20L1 in the 201-7 IRP were used in
6 the system simulations. The forecast gas and market prlces
7 were converted to year 20L0 base dol1ars.
O. Why was the 20L1 test year selected?
A. The primary reason for sel-ecting the 201-7 test
10 year was the 2077 IRP's projected in-service date of 2016
11 for the Boardman to Hemingway transmission (*B2H") project.
72 By sel-ectj-ng the 2017 test year, it made sense in the study
13 to evaluate integration costs for a system under two
14 scenarios-with B2H and without B2H. The Company now
15 expects B2H will- not be completed prior to 2020.
L6 O. Why are there costs associated wlth
t7 integrating wind generation on an electrical system?
18 A. Due to the variabl-e and intermittent nature of
19 wind generation, an electrical system operator must provide
20 operating reserves from other dispatchable resources that
21 are capable of j-ncreasing or decreasing generatj-on on short
22 notice to offset changes in the non-dispatchable wind
23 generation. The effect of having to hold operating
24 reserves on dispatchable resources is that the operation of
25 those resources is restricted and they cannot be
DeVOL, DI 5
Idaho Power Company
I
2
3
4
5
6
1
't
9
10
11
L2
13
L4
t_5
76
71
18
79
20
21
22
23
24
25
economicalJ-y dispatched
results in higher power
to their fullest capability. This
supply costs that are subsequently
passed on to customers
O.Are hydroelectric generators good resources to
use to integrate wind?
A. Yes. Operationally, the quick response
capabillties of a hydro unit makes it ideal- for responding
to changes in wind generation. However, many people
believe that because operationally hydro resources are good
resources for integrating wind, the cost of using them for
this purpose shoul-d be low; however, this is not the case.
The flexibility and quick response characteristics of hydro
units, especially when coupled with a storage reservoir
that can be used for shaping generation over longer time
periods, provides considerabl-e operational value as well as
economic value when water can be stored or shaped so that
it is used to produce electricity at times when it is the
most val-uabl-e.
The figure below, which depicts model results from
Idaho Power's ]atest wind i-ntegration study, shows this
impact on hydro generation at Idaho Power's Hel-Is Canyon
Complex during a typical week in June. The teal- line
represents how the generators woul-d be operated if
additional operating reserves were not necessary due to
In comparison, the red l-ine
DeVOL, DI 6
Idaho Power Company
wind generation on the system
1
2
3
4
5
6
8
9
10
11
72
13
74
15
t6
l1
18
79
20
27
22
23
24
25
shows how the range of generation j-s limited both upwards
and downwards in order to provide reserves for intermittent
wind resources. The result is less water can be run, and
el-ectricity generated, during heavy load hours when it is
more valuable.
Impact of Wind Generation on Hydroelectric Generators
1,400
1,200
1,000
800
600
400
200
n1v I i ---
Jun-5 .lun-6 Jun-7 Jun-8 Jun-9 Jun-10 Jun-11 Jun-12
O. Are natural gas and coal units used to
integrate wind?
A. Yes, they are. However, they are not able to
respond as quickly as hydro units. Natural gas units can
respond to changes in wind generatj-on, but they have to be
operating to do so. Because natural gas Combined-Cyc1e
Combusti-on Turbine (*CCCT") units are typically on the
margin relative to market prices, there are times when they
do not operate. Simple-Cycle Combustion Turbine ("SCCT")
units are typically operated as "peaker" plants due to
their lower efficiency/higher heat rate, and operate much
DeVOL, DI 7
Idaho Power Company
I less frequently than CCCTs. The cost of using natural gas
2 resources to integrate wind increases substantially when
3 the electrical- system operator has to operate natural gas
4 units for the sole purpose of providing operating reserves,
5 at times when the gas unit would otherwise not be
6 dispatched due to economics.
7 Coal units can also be used to integrate wind;
I however, operationally they are not able to rapidly change
9 generation output. Therefore, generation from coal- units
10 will typically be used l-ast and only if a sizeable
11 adjustment in total generation is needed to account for
72 changes in wind generation.
13 O. Are the costs to i-ntegrate wind affected by
L4 B2H?
15 A. Later in my testimony, I will- provide wind
76 integration costs found from Idaho Power's analysis based
71 on a system without B2H. Idaho Power's analysis indicates
18 that B2H reduces integration costs by 5 to 8 percent. Both
t9 sets of costs are provided in Exhibit No. 1.
20 It is important to point out that the modeling
21, performed for the 201,3 Study indicates that the primary
22 reason for the integration cost reduction is simply that
23 B2H al-lows greater access to wholesale market
24 opportunities, and is not related to operating reserves
25 provided by B2H.
DeVOL, DI I
Idaho Power Company
1
2
3
4
q
6
1
8
9
10
11
72
13
l4
15
1,6
11
18
79
20
2L
22
23
24
25
O. Is Idaho Power's wind integration study design
the same method used by all utilities to calculate the cost
of wind integration?
A.fdaho Power desj-gned its wind integration
study with the objective of isolating in its operations
modeling the effects directly related to integrating wind
generation. While this is a common study design used
towards meeting this objective, it is not a "specific"
design used by al1 utilities.' I do not believe it is
possible to detail a "specific" method because of
differences in electrical systems and the available
analysis tools. However, I think as a general principle,
the concept that has been used by various utilities is the
same-comparing the cost of operatJ-ng the electrical- system
both with and without intermittent wind generation on the
system. In addition, while many utilities have done wind
J-ntegration studies, not all utilities use the same
computer model- when modeling their electrical systems.
Therefore, it woul-d be difficult to define a specific
method due to potential limitations on the capabilities of
each model.
O. Have wind integration study methodologies
I
changed dramatically from study to study, potentially
resulting in large changes in the calculated reserve
requirements and wind integration costs?
DeVOL, DI 9
Idaho Power Company
1
2
3
4
5
6
7
I
9
10
11
t2
13
L4
15
76
71
18
19
20
2L
22
23
24
25
A.No. In fact, the basic framework of the Idaho
Power study has remained the same sj-nce 2008. Idaho
Power's study recognizes that a load-serving entity must
operate its dispatchable resources differently when wind is
part of its fleet. The study isol-ates the effects of wind
on the operation of the dispatchable resources by looking
at two scenarios. Eirst, the study models the operation of
dispatchabl-e resources when they are burdened with
incremental balancing reserves caused by wind generation.
Second, the study runs the same model without the
additional balancing reserves. This study design was the
model for Idaho Power's first wind integration study, and
has remained the mode1 for the second study.
Eor the Company's latest study, Idaho Power did make
one change to allow the model- to consider scenarios where
integration was not possibl-e. The Company made this change
because Idaho Power's dispatchabl-e resources are not always
capable of providing the bal-ancing reserves necessary to
integrate wind given the rapid expansion of instal-l-ed wind
capacity on Idaho Power's system. Even with this change,
however, the basic framework designed to estimate the costs
of modifying the operation of dispatchable resources such
that they are ready to respond to wind is unchanged.
0.In your opinion, how often should wind
DEVOL, Dr 10
Idaho Power Company
integration studies be updated?
1
2
3
4
5
6
7
I
9
10
11
1,2
13
1,4
15
16
t7
1B
19
20
2L
22
23
24
25
A.Generally, I believe that wind integration
studies shoul-d be updated every three years. Three years
is sufficient time to prepare the next study, yet short
enough that results are 1ikely to remain relevant between
studies. That said, it may be possible to update wind
integration costs on a more frequent basis if the update is
l-imited to updating only the load forecast, natural gas
prj-ces, and forward market prices. Under this scenario,
future wind buil-d-outs and wind data would remain unchanged
from the original study.
It may also be necessary to fulJ-y update wind
integration studies more frequently based on changes in the
Company's instal-1ed wind capacity. Erom a long-term
plannlng perspective, it has been challenging to predict
the expansion of installed wind capacity. With the
exception of the Elkhorn Valley wind project, which
resul-ted from the 2004 IRP's identification of a utility-
scale wind project in the preferred resource portfolio, the
wind projects connecting to Idaho Power's system have been
deveJ-oped as Qualifying Eacility ("QF") projects outside of
an IRP process. Wind fleet expansion has been
characterized by fits and starts, with periods where wind
penetration remains fairly stable, followed by periods with
very rapid growth. It is difficult to predict whether wj-nd
integration studies in the coming years will- need to be
DEVOL, Dr 11
Idaho Power Company
1
2
3
4
5
6
7
I
9
10
11
72
13
74
15
16
L1
l_B
19
20
2L
22
23
24
25
updated frequently to keep up with rapid wind development
when or if it occurs.
Other factors may also trigger the need for an
updated study. For example, systemic changes to el-ectric
market practices, the impJ-ementation of new regional
balancing initiatives, significant fuel price changes, or
the addition of new generating or demand-side resources,
particularly flexible resources providing wind-balancing
capability, may aII result in the need for a new
integration study.
O. Would the creation of an Energy Imbalance
Market ("EIM") facilitate wind integratj-on and reduce costs
and impacts?
A. Idaho Power has been participating in a
detailed analysis by the Northwest Power Pool ("NWPP") of
potential EIM designs. This analysis suggests that the
benefits of an EIM market on a NWPP-wide scal-e slightly
outweigh the costs necessary to implement and run such a
market. However, because of a degree of uncertainty with
the costs and benefits, an EIM shoul-d not be deveJ-oped
without caution. Eor an EIM to perform correctly, a
reasonably large footprint involving a large number of NWPP
participants would be necessary; Idaho Power by itself
cannot control- the development of an EIM. From a wind
integratj-on perspectj-ve, dn EIM woul-d facilitate sharing
DEVOL, DI 72
Idaho Power Company
1
2
3
4
5
6
7
I
9
10
1t_
t2
13
14
15
76
77
18
19
20
2t
22
23
24
25
wind diversity across a much greater footprint, so the
capacity necessary to service a wind fl-eet shoul-d be 1ower
However, the effect of an EIM on integration costs depends
on many factors related to the EIM program desJ-gn. Most of
these factors have yet to be finalized; therefore, the
existence of an EIM in the near term was not considered i-n
the 2073 Study analysis.
o.Idaho Power's wind study calculates balancing
reserve requirements based on day-ahead schedule errors as
opposed to hour-ahead schedule errors. Can you explain the
significance of both day-ahead scheduling and hour-ahead
scheduling as they rel-ate to wind integration?
A. Yes. In both cases, the issue is uncertainty.
Devj-ations between forecast and actual wind production must
be covered by other resources in order to maintain the
balance between supply and demand. Not surprisingly,
longer l-ead forecasts are more uncertaj-n than shorter lead
forecasts; therefore, deviations are typically larger for
forecasts of day-ahead wind production versus hour-ahead
wind production. Thus, the balancing reserve requirements
are greater when using day-ahead scheduj-i-ng.
o.Why does the Company use day-ahead scheduling
to determine its wind integration costs?
A. Idaho Power views the simulation of day-ahead
scheduling as appropriate due to system scheduling
DEVOL, DT 13
Idaho Power Company
1 practices. Day-ahead scheduling is reflective of the time
2 frame in which ldaho Power makes dispatching decisions and
3 is the reasonable and prudent tj-me frame in which to do so.
4 The use of day-ahead errors can be explained by considering
5 the implications of the alternative, where the amount of
6 balancing reserve is smaller because it is based on the
7 hour-ahead errors in forecast wind. As stated above, all
8 deviations between forecast and actual wind production need
9 to be covered. Thus, in scheduling the system day-ahead,
10 which is performed for each day, the dispatchabl-e
11 generators wou1d be scheduled to carry a smaller amount of
t2 response allowing them to cover deviations as determlned
13 from analysis of hour-ahead forecast errors. The
L4 dispatchable generators would not be schedul-ed to allow
15 them to respond to day-ahead forecast errors, meaning that
L6 the response to these larger errors is only achieved by
17 some other means, which in Idaho Power's view would too
18 often translate to a risky reliance on the wholesale
19 el-ectric market. Consequently, the prudent simulation of
20 day-ahead system scheduling is to ensure that dispatchable
2I generators are capable of responding in real time to
22 uncertainty in wind production as determined by analysis of
23 day-ahead forecast errors.
24 O. Can you describe the source of the wind
25 generation data used in the wind integration study?
DEVOL, DI t4
Idaho Power Company
1
2
3
4
5
6
7
I
9
10
11
72
13
74
15
16
L1
18
1,9
20
2!
22
23
24
25
A. Yes. As stated earlier, Idaho Power used
synthetic wind generation data and day-ahead wind
generation forecasts provided by 3TIER and Energy Exemplar
(formerly PLEXOS Solutions). The geographic dispersion of
the synthetic wind data used in Idaho Power's study is
representative of the geographic dispersion of wind build-
outs Idaho Power is like1y to integrate. The wind data
used for the study was provided by 3TIER, an industry
l-eader in renewable energy risk analysis. Indeed, 3TIER
developed the data set that was used by the National
Renewable Energy Laboratory for its Western Wlnd and Solar
Integration Study (*WWSIS"), which when completed in 20L01
was one of the largest and most comprehensive studj-es of
wind and solar resources to date. The WWSIS inc]uded data
for more than 32,000 existing or hypothetical- wind project
si-tes.
For Idaho Power's study, 3TIER developed a new time
series directly from the WWSIS data set for 43 locations
requested by Idaho Power. These locations correspond to
project sites that either have a current contract or have
requested a contract with Idaho Power. The 43 locations
are spread across a wide regi-on, with locations j-n five
states-Oregon, Idaho, Utah, Wyoming, and Montana. The
majority of the locations are in or peripheral to the Snake
River plain in southern Idaho.
DEVOL, Dr 15
Idaho Power Company
I believe the methodology used to develop the wind
2 generation data used for the study ensures it accurately
3 represents wind generation that is currently connected to
4 and would likely be connected to ldaho Power's system in
5 the future.
O. Is the cost of integrating wind considered in
7 Idaho Power's IRP when comparlng the costs of utility-owned
8 generation resources?
A. Yes, it is. The cost of integrating wind is
10 incurred regardless of whether the wind resource is
11 utility-owned or contracted through a third party, and
1,2 ultimately j-ncreases power supply costs that are passed on
13 to customers. It would be inappropriate to ignore these
L4 costs when evaluating new resources in the IRP.
t_5 O. Is the cost of integrating wind generation the
L6 same for anyone operating an el-ectrj-cal- system?
L7 A. No, it is not. As I explained previously, the
18 costs associated wlth wind integration are specific and
19 unique for each individual- electrical system based on the
20 amount of wind being integrated and the other types of
2L resources that are used to provide the necessary operating
22 reserves. In general terms, the cost.of integrating wind
23 j-ncreases as the amount of nameplate wind generation on the
24 el-ectrical- system increases.
25
DEVOL, Dr 76
Idaho Power Company
1
2
3
4
q,
6
7
B
9
10
11
l2
13
L4
15
1,6
l1
18
t9
20
21-
22
23
24
25
o.What is unique about the Idaho Power system
that influences integration costs?
A.The operating reserves Idaho Power uses to
integrate wj-nd are overwhelmingly provided by its
hydroelectric system. As I stated earlier, hydroel-ectric
generating faci-1ities, particularly those with large
storage reservoirs, are very effective at quickly
respondi-ng to wind's variability and intermittency.
However, maintaining this capability to respond comes at a
relatively high opportunity cost. If we consider as an
example the need to hol-d un-dispatched generating capacity
in reserve during on-peak hours, where this capacity is
held to respond to wj-nd down ramps, then the cost to hol-d
this capacity on hydroel-ectric generators is essentially
equal to the market cost of power. On the other hand, if
the reserve capacity is carried on thermal generators, then
the cost of holding capacity in reserve is equal to the
market cost of power -l.ess the variable cost to fuel and
operate the generators. In short, operatlon of the
hydroelectric system can be very effectively optimized, and
the de-optimization needed to ready the system to integrate
wind has noticeable and costly impacts.
Idaho Power is also unique in the high level of wind
generation on its system relative to its system loads and
other availabl-e dispatchable generation. fdaho Power has
DEVOL, Dr l1
Idaho Power Company
1 seen very rapid growth of wind generation on its system,
2 especially relative to system Ioad and other generation
3 resources. This has l-ed to the recognition that Idaho
4 Power's finite capability for integrating wind is nearing
5 depletion. Even at the current level of wind penetrati-on,
6 dispatchable thermal and hydro generators are not always
7 capable of providing the balancing reserves necessary to
8 integrate wind. This situation is expected to worsen as
9 wind penetration levels increase, particularly during
10 periods of l-ow customer demand.
11
72
II. 2OL3 WI![D INTEGRATION STT'DY REST'LTS
O. Based on the resul-ts of the 20L3 Study, what
13 is the cost of integrating wind generation on Idaho Power's
74 electri-cal system?
15 A. As previously discussed, the 2073 Study
16 analyzed three different level-s of wind penetratJ-on:
7"7 800 MW; 1,000 MW; and 1,200 MW. The results of the
18 analysis, based upon 2077 IRP inputs, showed integration
79 costs of $8.06/megawatt-hour ('MWh"), $13.06/MWh-, and
20 $19.01/MWh, respectively. These wind integration costs are
2L associated with total wind generation at any given time,
22 not ;ust incremental additions.
23 O. How are the results different j-n the Updated
24 201,3 Study?
25
DEVOL, Dr 18
Idaho Power Company
I
2
3
4
5
6
1
I
9
10
11
72
13
t4
15
L6
l1
18
L9
20
2L
22
23
24
25
A. As previously mentioned, subsequent to the
development of the 2013 Study, the Company filed its 201-3
IRP. Because the 20L3 Study was developed using inputs
from the 20Lt IRP, the Company has updated the
determination of wind integration costs based upon the
inputs in the 2073 IRP.
o.What inputs were used from the 20L3 IRP in the
Updated 20L3 Study?
A.The Company updated inputs to determine the
current wind integration costs using the 20L3 IRP load
forecast, Mid-C market prices, natural gas price forecast,
and the coal- price forecast.
o.What was the resul-t of recafculating the wind
integration costs based upon inputs from the 201,3 IRP?
A.The resul-t of updating the inputs used in the
study to those from the 201-3 IRP was a reduction in the
wind integration costs. As before , for the 2017 test year,
the updated integration costs per MWh associated with total
wind generation at the 800 MW; 1,000 MW; and 1,,200 MW
penetration levels were $6.83, $10.22, and $74.22,
respectively.
o.What would be the revised incremental- costs of
wind integratj-on for the Updated 2073 Study?
A.Maintalning the conservative assumption that
all- 678 MW of current wind generation were assessed the cap
DEVOL, Dr l_9
Idaho Power Company
1
2
3
4
5
6
7
8
9
10
11
\2
13
14
15
76
L7
18
L9
20
21,
22
23
24
25
of $6.50/lrtwh, the respective incremental costs
integration woul-d be $8.67 , $24.00, and $34.70
o.How much wind generation capaclty
of wind
per MWh.
does Idaho
Power currently have on its system?
A.Idaho Power currentl-y has 577 MW of wind
generation capacity from Public Utility Regulatory Policies
Act of 7978 projects and an additional 101 MW of wind
generation capacity from the Elkhorn VaIIey wind project,
for a total of 618 MW of wind generation capacity currently
on-line.
O.Do the wind J-ntegration costs identified for
the three different 1evels of wind penetration represent
the cost per MWh to integrate the ful-l- instal-l-ed wind at
the respective penetratlon level-?
Yes, the integration costs stated above
represent the cost per MWh to integrate the fuII instal-led
wind generation capacity at the respective penetration
Ievels studied. For example, the resul-ts indicate that the
fuI1 fleet of wind generators making up the 800 MW
penetration level brings about costs of $6.83 for each MWh
integrated. However, wind generators comprising the 618 MW
of current i-nstal-led capacity on the Idaho Power system are
assessed an integration cost based upon a percentage of the
avoided cost rate contained in their power purchase
agreement and is capped at only $5.SO/ltWfr.
DEVOL, DI 20
Idaho Power Company
A.
1
2
3
4
5
6
7
I
9
10
11
L2
13
l4
15
1,6
L7
18
19
20
21
22
23
24
25
0.Based upon a conservative assumption that all
of the current 678 MW of wind generation were currently
being assessed the cap of $6.sO/UWfr (which they are not)
and that they woul-d continue to be assessed just $5.50/MWh,
what then would be the j-ncremental cost of wind integration
for new wind generation?
A.In order to fuI1y recover the $6.83/MWh
lntegration costs associated with 800 MW of installed wind
capacity, wind generators in the increment between the
current penetration level (678 MW) and the 800 MW
penetration 1eve1 will need greater assessed integration
costs. Study analysis indicates that if the current 618 MW
of wind generation were to be assessed the ful-l- cap of
$6.50/UWfr, and were to continue to receive this cdp, the
new wind generators will need to recognize integration
costs of $8.67ltqWh to al-l-ow ful1 recovery of integration
costs associated with 800 MW of installed wind capacity.
Similarly, generators between the 800 MW and 1000 MW
penetration levels introduce incremental system operatj-ng
costs requiring the assessment of lntegration costs of
$24.00/MWh, and generators between 1000 MW and L,200 MW
require incremental integratj-on costs of $34.70lMWh.
The 2073 Study results and the Updated 2013 Study
results are summarized in the tables beIow.
I
DEVOL, DI 2t
Idaho Power Company
2013 STUDY (using 2OLt IRP inputs)
Penetration Level 8OO MW 1,000 MW L,200 MW
Allocated EquaIIy to
alt_ Ir,Iind (/MWh)
$8 06 $13.06 $19.01
Incremental Cost
Allocation (/MWh)
$15.70 $33.42 $49 .46
UPDATED 2OL3 STTDY (using 2OL3 IRP inputs)
Penetration Level-8OO MW 1,000 MW L,200 MW
Al-Iocated Equally
al-I Wind (/MWh)
to $6.83 $10 .22 $14.22
Incremental- Cost
All-ocation (/UWn)
$8.67 $24.00 $34.70
4
5
6
1
I
9
10
11
72
13
L4
15
L6
L7
18
79
O. Has Idaho Power proposed a similar integration
charge for solar QFs?
A. Not at this time. Idaho Power's proposal
addresses only wind integration costs. However, upon
completion of a solar-specific integration study, Idaho
Power believes it woul-d be appropriate to assess a similar
integration charge for solar QFs
o.Does this conclude your testimony?
Yes
DEVOL, Dr 22
Idaho Power Company
A
1
2
3
4
5
6
7
8
9
10
11
L2
13
L4
15
76
L1
18
19
20
2t
22
23
24
25
26
21
28
29
30
31
32
STATE OF IDAHO
County of Ada
SS.
ATTESTATIOI{ OF TESTIMOIIY
SWORN to before me this 29th day of
T, Philip B. DeVol, havlng been duly sworn to
testify truthfully, and based upon my personal knowledge,
state the following:
I am employed by Idaho Power Company as the Resource
Plannj-ng Leader j-n the Water and Resource Planning
Department and am competent to be a witness in this
proceeding.
I declare under penalty of perjury of the laws of
the state of Idaho that the foregoing pre-filed testimony
is true and correct to the best of my informatlon and
bel-ief .
DATED this 29th day of November 20L3.
SUBSCRIBED AND
November 20].3.
DEVOL, Dr 23
Idaho Power Company
Philip
ooI rl f
apo't'.C,ubl,'
tary Public
Residing at:
expl-res:My commission
BEFORE THE
IDAHO PUBLIG UTILITIES COMMISSION
GASE NO. IPC-E-13-22
IDAHO POWER COMPANY
DeVOL, DI
TESTIMONY
EXHIBIT NO. 1
^llmloNIPOIIIR-
Wind lntegration
Study Report
February 2013
@ 2012 Idaho Power Exhibit No. 1
Case No. IPC-E-13-22
P. DeVoI, IPC
Page 1 of 48
ldaho Power,1221W ldaho Street, Boise, ldaho 83702
ldaho Power Company Wind lntegration Study Report
TneLe oF CoNTENTS
Executive Summar... ................5
Balancing Reserves.... ..........5
Wind Integration Costs......... ..................7
Incremental Cost of Wind lntegration .......................8
Technical Review Committee. ..............12
Energy Exemplar Contribution................ ................13
Idaho Power System Overview... ................15
Hydroelectric Generating Projects.. ......15
Coal-Fired Generating Projects...... .......16
Natural Gas-Fired Generating Projects...... ..............16
Transmission and Wholesale Market................. ......16
Power Purchase Agreements .................18
System Demand .................18
System Scheduling. ............19
Balancing Reserves Calculations and Operating Reserves............... ...................23
Balancing Reserves for Variability and Uncertainty in System Demand..... ....................25
Contingency Reserve Obligation ..........25
Day-Ahead Scheduling ......27
Demand and Wind Forecasts ................28
Transmission System Modeling... .........28
Overgeneration in System Modeling .......................29
Exhibit No. 't
Case No. IPC-E-'!3-22
P. DeVoI, IPC
Page i
Page 3 of 48
Wind lntegration Study Report ldaho Power Company
Wind Integration Costs ......31
lncremental Cost of Wind Integration................ ........................32
Spilling Water........ ............33
Maximum Idaho Power System Wind Penetration......... ............34
Effect of Wind Integration on Thermal Generation. ...................36
Recommendations and Conclusions.......... ....................37
lssues Not Addressed by the Study... .......................38
Measures Facilitating Wind Integration.............. ........................39
Future Study of Wind Integration .........39
Table I
Table2
Table 3
Table 4
Table 5
Table 6
Table 7
Table 8
Table 9
Table l0
Table I I
Table 12
Table 13
Table 14
Table 15
Table Bl
LIsT oF TABLES
Balancing reserves requirements (MW) .....................6
Wind integration costs ($/MWh) .............7
Wind integration costs with the Boardman to Hemingway transmission
line ($/\4Wh)................ ........7
Incremental wind integration costs (SA{Wh) .............9
Balancing reserve requirements (MW)..... ................25
Modeled transmission constraints (MW)...... ............28
Modeled transmission constraints-simulations with 500-kV Boardman to
Hemingway transmission line (MW)................. .......29
Wind penetration levels and water conditions.......... ...................31
Integration costs ($/IrtWh)............... ......32
Integration costs with the Boardman to Hemingway transmission line(s^4wh) ..........32
Incremental wind integration costs ($A{Wh) ...........33
lncremental Hells Canyon Complex spill (thousands of acre-feet)...................................34
Curtailment of wind generation (annual MWh)........ ...................35
Annual generation for thermal generating resources for the test case(Gwh) ..............36
Integration costs ($/NIWh)................. .......................37
Monthly and annual capacity factors (percent of installed nameplatecapacitv).... ;;.;;*fu:Xil
P. DeVol, IPC
Page 4 of 48Page ii
ldaho Power Company Wind lntegration Study Report
Figure I
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
Figure 8
Figure 9
Figure 10
LIST oF FIGURES
Installed wind capacity connected to the Idaho Power system..... ....................5
Curtailment of wind generation (average annual MWh) ...............8
Integration costs with incremental integration costs ($/\4Wh)............... ..........9
Installed wind capacity connected to the Idaho Power system (MW)...............................11
Idaho Power transmission paths....... ......17
Wind-forecasting and generation data ......................23
Deviations between forecast and actual wind generation with monthly
balancing reserves requirements (M\M) ....................24
Integration costs with incremental integration costs ($/\tIWh)............... ........33
Curtailment of wind generation (average annual MlVh) .............35
Curtailment of wind generation (average annual MWh) .............38
LIST OF APPENDTcES
Appendix A. May 9,2012, Explanation on wind data............ ..........43
Appendix B. Wind data summaries.............. .................45
Exhibit No. 1
Case No. IPC-E-13-22
P. Devol, IPC
Page iii Page 5 of48
Wind lntegration Study Report
This page left blank intentionally.
Exhibit No. 1
Case No. IPC-E-13-22
P. DeVoI, IPC
Page 6 of 48
Page iv
ldaho Power Company Executive Summary
ExecunvE SUMMARY
As a variable and uncertain generating resource, wind generators require ldaho Power to modify power
system operations to successfully integrate such projects without impacting system reliability.
The company must build into its generation scheduling extra operating reserves designed to allow
dispatchable generators to respond to wind's variability and uncertainty.
Idaho Power, similar to much of the Pacific Northwest, has experienced rapid growth in wind generation
over recent years. As of January 2013,ldaho Power has reached on-line wind generation totaling 678
megawatts (MW) of nameplate capacity. The rapid growth in wind generation is illustrated in Figure l.
"-tt "C^od *C "-d "--s ^nnu' ,ond "d "-*-"^n€' ,ono'"d "*d"d "-d
Figure I lnstalled wind capacity connected to the ldaho Power system
This rapid growth has led to the recognition that Idaho Power's finite capability for integrating wind is
nearing depletion. Even at the current level of wind penetration, dispatchable thermal and hydro
generators are not always capable of providing the balancing reserves necessary to integrate wind.
This situation is expected to worsen as wind penetration levels increase.
Balancing Reserves
This investigation quantified wind integration costs for wind installed capacities of 800 MW,
1,000 MW, and 1,200 MW. Synthetic wind generation data and corresponding day-ahead wind
generation forecasts at these build-outs were provided by Energy Exemplar (formerly PLEXOS
Exhibit No. 1
Case No. IPC-E-13-22
';3ili';l'.?
Page 5
Executive Summary ldaho Power Company
Solutions) and 3TIER. Based on analysis of these dat4 the following monthly balancing reserves
requirements were imposed in system modeling.
Table 1 Balancing reserves requirements (MW)
Wind Gen 800 Mw 1,000 Mw 1,200 Mw
Reg Up Reg Down Reg Up Reg Down Reg Up Reg Down
January
February
March
April
May
June
July
August
September
Oclober
November
December
199
252
226
255
258
266
274
172
242
217
226
267
-262
-246
-295
-353
-290
-285
-256
-179
-219
-248
-336
-338
246
319
281
331
328
339
355
21s
309
275
277
326
-325
-297
-368
-450
-366
-363
-322
-224
-280
-308
421
424
295
379
339
395
392
409
423
257
371
329
333
394
-390
-351
-444
-540
439
436
-384
-267
-337
-367
-507
-510
The term Reg Up is used for generating capacity that can be brought online in response to a drop in wind
relative to the forecast. Reg Down is used for on-line generating capacity that can be turned down in
response to a wind up-ramp. The balancing reserves requirements assume a 90 percent confidence level
and thus are designed to cover deviations in wind relative to forecast except for extreme events
comprising 5 percent at each end.
Study Design
The study employed the following two-scenario design:
o Base scenario for which the system was not burdened with the incremental balancing reserves
necessary for integrating wind
o Test scenario for which the system was burdened with the incremental balancing reserves
necessary for integrating wind
System simulations for the two scenarios were identical, except that generation scheduling for the test
scenario included the condition that dispatchable thermal and hydro generators must provide the
appropriate amount of incremental balancing reserves. Having the prescribed balancing reserves
positions these generators such that they can respond to changing wind.
System simulations were conducted for a2017 test year. Customer demand for 2017, as projected for the
201I Integrated Resource Plon (lW), was used in system modeling. To investigate the effect of water
conditions on wind integration, the study also considered Snake River Basin stream flows for three
separate historic years representing low (2004), average (2009), and high (2006) water years
Exhibit No. 1
Case No. IPC-E-I3-22
P. DeVoI, IPC
Page 6
Page 8 of 48
ldaho Power Company Executive Summary
Wind lntegration Costs
The integration costs in Table 2 were calculated from the system simulations.
Table 2 Wind integration costs ($/MWh)
Nameplate Wind
Water Condition 800 Mw 1,000 Mw 1,200 MW
Average (2009)
Low (2004)
High (2006)
Average
$7.18
$7.26
$9.73
$8.06
$1 1.94
$12.44
$14.79
$13.06
$18.15
$18.15
$20.73
$r9.0r
Simulations with the proposed Boardman to Hemingway transmission line were also performed,
yielding the results in Table 3.
Table 3 Wind integration costs with the Boardman to Hemingway transmission line ($/TlWtrl
Nameplate Wind
Water Condition 800 Mw 1,000 MW 1,200 MW
Average (2009)
Low (2004)
High (2006)
Average
$6.51
$6.66
$9.72
$7.63
$11.03
$11.04
$13.78
$r 1.95
$16.38
$16.67
$19.53
sr7.53
Curtailment
The study results indicate customer demand is a strong determinant of Idaho Power's ability to integrate
wind. During low demand periods, the system of dispatchable resources often cannot provide the
incremental balancing reserves paramount to successful wind integration without creating an imbalance
between generation and demand. Under these circumstances, curtailment of wind generation is often
necessary to maintain balance. Modeling demonstrates that the frequency of curtailment is expected to
accelerate greatly beyond the 800 MW installed capacity level. While the maximum penetration level
cannot be precisely identified, study results indicate wind development beyond 800 MW is subject to
considerable curtailment risk. Importantly, curtailed wind generation was removed from the production
cost analysis for the wind study modeling, and consequently had no effect on integration cost
calculations. The curtailed wind generation simply could not be integrated, and the cost-causing
modifications to system operations designed to allow its integration were assumed to not be made.
The curtailment of wind generation observed in the wind study modeling is shown in Figure 2.
Exhibit No. 1
Case No. IPC-E-I3-22
P. DeVol, IPC
Page I of48
PageT
Executive Summary
50,m
45,(m
4,0,(m
35,(m
Pc
E *,*o
Ei zs,om(,
.C3 2o,om
=15,(m
10,(m
5,mo
o
Figure 2 Curtailment of wind generation (average annua! MWh)
lncremental Cost of Wind lntegration
The integration costs previously provided in Tables 2 and 3 represent the cost per MWh to integrate the
full installed wind at the respective penetration levels studied. For example, the results of Table 2
indicate that the full fleet of wind generators making up the 800 MW penetration level bring about costs
of $8.06 for each MWh integrated. However, wind generators comprising the 678 MW of current
installed capacity on the Idaho Power system are assessed an integration cost of only $6.50ArIWh'.
In order to fully cover the $8.0644Wh integration costs associated with 800 MW of installed wind
capacity, wind generators in the increment between the current penetration level (678 MW) and the
800 MW penetration level will need greater assessed integration costs. Study analysis indicates that
these generators will need to recognize integration costs of $16.704{Wh to allow full recovery of
integration costs associated with 800 MW of installed wind capacity. Similarly, generators between the
800 MW and 1000 MW penetration levels introduce incremental system operating costs requiring the
assessment of integration costs of $33.42lIvIWh, and generators between 1000 MW and 1,200 MW
require incremental integration costs of $49.464,IWh. A graph showing both integration costs and
incremental integration costs is provided in Figure 3 below. The incremental integration costs are
summarized in Table 4.
Exhibit No. 'l
Case No. IPC-E-13-22
P. DeVol, IPC
lntegration cost stipulated by ldaho Public Utilities Commission Case No. IPC-E-07-03, Order No. 30488.
Page 8
'10 of 48
ldaho Power Company Executive Summary
z S4o
BE
fi;{lg$o
=o{Et,,E, $zo
6m &x)
rAuEHArEW6aDlr$rirl
Figure 3 lntegration costs with incremental integration costs ($/ttwh)
Table 4 lncremental wind integration costs ($/MWh)
-* TNCREMENTAL CO5T {s/Mwh}
..r-wr{D THTEGRATTON COSI ($lUWtrl
Nameplate Wind
678 - 800 MW 800 - 1,000 Mw 1,000 - 1,200 Mw
lncrcmental cost per MWh $16.70 $49.46
Exhibit No. 1
Case No. IPC-E-13-22
P. DCVOI, IPC
Page 11 of48
Page 9
Executive Summary ldaho Power Company
This page left blank intentionally.
Exhibit No. 'l
Case No. IPC-E;13-22
P. DeVol, IPC
Page 10
12 ol 18
ldaho Power Company lntroduction
lrurnooucnoN
Electrical power generated from wind turbines is commonly known to exhibit greater variability and
uncertainty than that from conventional generators. Because of the incremental variability and
uncertainty, it is widely recognized that electric utilities incur increased costs when their systems are
called on to integrate wind power. These costs occur because power systems are operated less optimally
to successfully integrate wind generation without compromising the reliable delivery of electrical power
to customers. Idaho Power has studied the unique modifications it must make to power system
operations to integrate the rapidly expanding amount of wind generation connecting to its system.
The purpose of this report is to describe the operational modifications taken to integrate wind and the
associated costs. The study of these costs is viewed by Idaho Power as an important part of efforts to
ensure prices paid for wind power are fair and equitable to customers and generators alike.
Idaho Power first reported on wind integration in2007. While there was a sizable amount of wind
generation under contract in2007, the amount of wind actually connected to the ldaho Power system at
the time of the first study report was just under 20 MW nameplate. Over recent years, the amount of
wind generation connected to the Idaho Power system has sharply risen. As of January 2013, Idaho
Power has reached on-line wind generation totaling 678 MW nameplate. The rapid growth in wind
generation is illustrated in Figure 4.
^nd *d ^nd *C "C "-*." ^nn*' "on*" ^nuo" "on*" ^nd' "od^nd ,-d "d "dFigure 4 Installed wind capacity connected to the Idaho Power system (MW)
The steep upturn in wind generation has driven Idaho Power to expand its area of concern beyond the
operational costs associated with wind integration to the consideration of the maximum wind penetratton
.rr,,on,uo. .,
Case No. IPC-E-13-22
P. Devol, IPC
Page 13 of48
Page 1 1
lntroduction ldaho Power Company
level its system can reliably integrate. Thus, the objective of the Idaho Power wind integration study is
to answer the following two questions:
o What are the costs of integrating wind generation on the Idaho Power system?
o How much wind generation can the Idaho Power system accommodate without
impacting reliability?
A critical principle in the operation of a bulk power system is that a balance between generation and
demand must generally be maintained. Power system operators have long studied the variability and
uncertainty present on the demand side of this balance, and as a matter of standard practice carry
operating reserves on dispatchable generators designed to accommodate potential changes in demand.
The introduction of significant wind power causes the variability and uncertainty on the generation side
of the balance to markedly increase, requiring power system operators to plan for carrying incremental
amounts of operating reserves, in this case necessary to accommodate potential changes in
wind generation.
For the purposes ofthis study report, the term balancing reserves is used to denote the operating
reserves necessary for integrating wind. A document review on wind integration indicates a variety of
terms for this quantity. Regardless of term, the property being described is generally the flexibility a
balancing authority must carry to reliably respond to variability and uncertainty in wind generation
and demand.
A key component in the study of wind integration, as well as the successful in-practice operation of a
power system integrating wind, involves the estimation of the additional balancing reserves dispatchable
generators must carry to allow the balance between generation and demand to be maintained.
Thus, three essential objectives of this report are to describe the analysis performed by Idaho Power to
estimate the incremental balancing reserves requirements attributable to wind generation, describe the
power system simulations conducted to model the scheduling of the reserves, and estimate associated
costs. The study also evaluates situations where the incremental wind-caused balancing reserves exceed
the capabilities of Idaho Power's dispatchable generators, putting the system in a position where it
cannot accept additional output from wind generators without compromising reliability.
Technical Review Committee
Idaho Power held a public workshop on April 6,2012, to discuss its work on wind integration.
This workshop included a discussion of methodology and preliminary results, as well as a question and
answer session. Following the workshop, the company began working with a technical review
committee comprised of individuals selected by Idaho Power based on their knowledge of regional
issues surrounding wind generation and the operation of electric power systems.
The following members agreed to serve on the committee:
o Ken Dragoon (EcofysA.iorthwest Power and Conservation Council)
o Kurt Myers (Idaho National Laboratory [INL])
o Frank Puyleart (Bonneville Power Administration [BPA])
o Rick Sterling (Idaho Public Utilities Commission [IPUC])
The purpose of the work with the technical review committee was to describe in greater detail the study
methodology, including an in-depth review of the model used for system simulations for the study.
Given this information, the company asked the members of the committee for their specific comments
Exhibit No. 1
Case No. IPC-E-13-22
P. DeVol, IPC
Page 12
'14 of 18
ldaho Power Company lntroduction
upon release of this wind integration study report. These comments will be specially noted as having
been provided by the technical review committee on the basis of its in-depth review of study methods.
Energy Exemplar Contribution .
Idaho Power contracted with Energy Exemplar (formerly PLEXOS Solutions) for assistance with the
wind integration study. Energy Exemplar's involvement was critical in the development of the wind
generation data used for the study, particularly in the development of representative wind generation
forecasts used in the analysis to estimate appropriate balancing reserves requirements. Energy Exemplar
was also instrumental in the design of the study methodology, providing key counsel in the formulation
of the two-scenario study design detailed later in this report.
With respect to system simulations for the wind study, Idaho Power has developed considerable
expertise modeling the power system over recent years. In parallel with the Energy Exemplar efforts,
Idaho Power developed a model that optimizes the wind, hydro, and thermal generation production.
This internally-developed model was used for system simulations included in the wind study.
Exhibit No. 1
Case No. IPC-E-13-22
P. DeVol, IPC
Page 15 of48
Page 13
lntroduction ldaho Power Company
This page left blank intentionally.
Exhibit No. 1
Case No. IPC-E-13-22
P. DeVol, IPC
Page 14
16 of48
ldaho Power Company ldaho Power System Overview
lolno Powen Svsrem OveRvrew
Idaho Power serves approximately 500,000 customers in southern Idaho and eastem Oregon through the
operation of a diversified power system composed of supply- and demand-side resources, as well as
significant transmission and distribution infrastructure. From the supply-side perspective, ldaho Power
relies heavily on generation from l7 hydroelectric plants on the Snake River and its tributaries.
These resources provide the system with electrical power that is low-cost, dependable, and renewable.
Idaho Power also shares joint ownership of three coal-fired generating plants and is the sole owner of
three natural gas-fired generating plants, including the recently commissioned Langley Gulch Power
Plant. With respect to demand-side resources, Idaho Power has received recognition for its demand
response programs, particularly the part these dispatchable programs have played in meeting critical
summertime capacity needs. Finally, Idaho Power maintains an extensive system of transmission and
distribution resources, allowing it to connect to regional power markets, as well as distribute power
reliably at the customer level.
Hyd roelectric Generati ng Projects
Idaho Power operates l7 hydroelectric projects located on the Snake River and its tributaries.
Together, these hydroelectric facilities provide a total nameplate capacity of 1,709 MW and annual
generation equal to approximately 970 average megawatts (aMW), or 8.5 million megawatt hours
(MWh), under median water conditions. The backbone of Idaho Power's hydroelectric system is the
Hells Canyon Complex (HCC) in the Hells Canyon reach of the Snake River. The HCC consists of
Brownlee, Oxbow, and Hells Canyon dams and the associated generation facilities. [n a normal water
year, the three plants provide approximately 68 percent of ldaho Power's annual hydroelectric
generation. Water storage in Brownlee Reservoir also enables the HCC projects to provide the major
portion of Idaho Power's peaking and load-following capability. The capability to respond to varying
load is increasingly being called on to regulate the variable and uncertain delivery of wind generation.
Hydro is Idaho Power's wind integration resource of choice because of its quick response capability as
well as large response capacity. However, the capacity of the hydro system to respond to wind
variability is recognized as finite; power-system operation, in practice and as simulated for this study,
indicates the hydro system is not always able to sufficiently provide the balancing reserves needed for
responding to wind. Using the hydro system for wind integration also limits its availability for other
opportunities. The costs of these lost opportunities are a significant part of wind integration costs.
For the wind integration study, the hydroelectric generators at the Brownlee and Oxbow dams were
designated in the modeling as available for providing wind-caused balancing reserves. This is consistent
with system operation in practice, where the generators at these projects are dispatched to provide the
overwhelming majority of operating reserves. Under standard operating practice, the remaining
hydroelectric generators of the Idaho Power system are not called on for providing operating reseryes.
Generators at the Lower Salmon, Bliss, and C. J. Strike plants are capable of some ramping for
responding to intra-day variation in load. However, under certain flow conditions, the flexibility of the
smaller reservoirs to follow even load trends is greatly diminished, and the facilities are operated strictly
as run-of-river (ROR) projects.
Exhibit No. 1
Case No. IPC-E-13-22
P. DeVoI, IPC
Page '17 of 48
Page 15
ldaho Power System Overview ldaho Power Company
Coal-Fired Generating Projects
Idaho Power co-owns three coal-fired power plants having a total nameplate capacity of l,l 18 MW.
With relatively low operating costs, these plants have historically been a reliable source of stable
baseload energy for the system. The output from these plants over recent years is somewhat diminished
because of a variety of conditions, including relatively high Snake River and Columbia River stream
flows, lagging regional demand for electricity associated with slow economic growth, and an oversupply
of energy in the region. Idaho Power is currently studying the economics of operating its coal-fired
plants, specifically the cost effectiveness of plant upgrades needed for environmental compliance at the
Jim Bridger and North Valmy coal plants. The Boardman coal plant in northeastem Oregon will not
operate beyond 2020 and ldaho Power's 64 MW share of the plant will no longer be available to serve
customer load.
Coal is one of the thermal resources Idaho Power uses to integrate wind generation. Unlike hydro,
the fuel for the coal plants comes at a cost. These fuel costs, as well as the lost opportunities created by
using the coal capacity to integrate wind, make up another part of the wind integration costs. The coal
generators do not have the large range and rapid response provided by the hydro units.
Natural Gas-Fired Generating Projects
Idaho Power owns and operates four simple-cycle combustion turbines totaling 416 MW of nameplate
capacity, and recently commissioned a 300 MW combined-cycle combustion turbine. The simple-cycle
combustion turbines (located at Danskin and Bennett Mountain project sites) have relatively low capital
costs and high variable operating costs. As a consequence ofthe high operating costs, the simple-cycle
turbines have been historically operated primarily in response to peak demand events and have seldom
been dispatched to provide operating reserves. Expansion of their operation to provide balancing
reserves for integrating wind is projected to lead to a substantial increase in power supply costs.
Idaho Power commissioned in July 2012 the 300 MW Langley Gulch Power Plant. As a combined-cycle
combustion turbine, this generating facility has markedly lower operating costs than the simple-cycle
units and is consequently expected to be a critical part of the fleet of generators dispatched to provide
balancing reseryes for responding to variable wind generation.
Transmission and Wholesale Market
Idaho Power has significant transmission connections to regional electric utilities and regional energy
markets. The company uses these connections considerably as part of standard operating practice to
import and export electrical power. Utilization of these paths on a day-to-day basis is typically driven by
economic opportunities; energy is generally imported when prices are low and exported when prices are
high. Transmission capacity across the connections does not reduce system balancing reserves
requirements. Thus, balancing reserves necessary for reliable power system operation in practice are
provided by dispatchable generators. The wholesale power market, as accessed through regional
transmission connections, is not able to provide balancing reserves.
Idaho Power's existing transmission system spans southern Idaho from eastern Oregon to westem
Wyoming and is composed of transmission facilities having voltages ranging from I l5 kilovolts (kV)
to 500 kV. The sets of lines transmitting power from one geographic area to another are known as Exhibit No. 1
transmission paths. There are defined transmission paths to other states and between southern Idah61ffidf;i1""i;li;t
Page 16
18 of48
ldaho Power Company ldaho Power System Overview
centers such as Boise, Twin Falls, and Pocatello. Idaho Power's transmission system and paths are
shown in Figure 5.
The critical paths from the perspective of providing access to the regional wholesale electricity market
are the Idaho-Northwest, Idaho-Utah (Path C), and ldahe-Montana paths. The Boardman to
Hemingway transmission line identified by ldaho Power in the preferred portfolio of its 201I IRP will
be an upgrade to the tdaho-Northwest path. The combination of these paths provides Idaho Power
effective access to the regional market for the economic exchange of energy.
While ldaho Power does not consider the regional market part of its day-to-day solution for integrating
wind generation, it may be necessary during extreme events to use the regionaltransmission connections
and rely on the regional energy market to accommodate wind. The company expects that at times even
the regional market will be insufficient to integrate wind. During these times when ldaho Power and the
regional market have insufficient balancing reserves to successfully integrate wind generation, it may be
necessary to curtail wind, or even curtail customer load, to maintain electrical system stability
and integrity.
Exhibit No. 'l
Case No. IPC-E-13-22
P. DeVol, IPC
Page 19 of48
\.. rsrncPtrR
mrrescrr ,l\iii\^lirEal \t
ttl P.rtC ... S(V
o12525$x-
ROCKiruualrptrR
lo S &rtE
,
Figure 5 ldaho Power transmission paths
Page 17
ldaho Power System Overview ldaho Power Company
Power Purchase Ag reements
In addition to power purchases in the wholesale market, Idaho Power purchases power pursuant to
long-term power purchase agreements (PPA). The company has the following notable firm wholesale
PPAs and energy exchange agreements:
o Raft River Energy I, LLC-For up to l3 MW (nameplate generation) from its Raft River
Geothermal Power Plant Unit #l located in southern Idaho. The contract term is through April 2033.
o Telocaset Wind Power Partners, LLC-For l0l MW (nameplate generation) from the Elkhorn
Valley wind project located in eastern Oregon. The contract term is through 2027.
o USG Oregon LLC-For 22MW (estimated average annual output) from the Neal Hot Springs
geothermal power plant located near Vale, Oregon. The contract term is through 2037 with an option
to extend.
r Clatskanie People's Utility District-For the exchange of up to l8 MW of energy from the
Arrowrock project in southern Idaho for energy from Idaho Power's system or power purchased at
the Mid-Columbia trading hub. The initial term of the agreement is January l, 2010 through
December 3l,2015.Idaho Power has the right to renew the agreement for two additional
five-year terms.
System Demand
Idaho Power's all-time system peak demand is 3,245 MW, set on July 12,2012, and the all-time winter
peak demandis2,527 MW, set on December 10, 2009. An important characteristic of the ldaho Power
system is the intra-day range from minimum to maximum customer demand, which during the summer
commonly reaches 1,000 MW and occasionally exceeds 1,200 MW. Thus, generating resources that can
follow this demand as it systematically grows during the day are criticalto maintaining reliable system
operation. Hydro generators, particularly those of the HCC, provide much of the demand following
capability. Recent natural gas-fired resource additions are also instrumental in allowing the system to
reliably meet system demand. An additional resource available to the system is the targeted dispatch of
demand response programs. These demand-side programs have proven to dependably reduce system
demand during extreme summer load events. From the perspective of system reliability, the nature of
Idaho Power's customer demand places a premium on the value associated with capacity-providing
resources; energy resources, such as wind, contribute markedly less towards promoting
system reliability.
It is recognizedthat production from wind projects does not dependably occur in concert with peak
customer demand. In fact, there is a tendency to experience periods during which production from wind
and hydro facilities is high and customer demand is low. The coincidence of these circumstances leads
to an excess generation condition, where the capability of system generators to reduce their output in
response to wind is severely diminished. Such excess generation events have been observed in recent
years by Idaho Power and other balancing authorities in the Pacific Northwest. System stability for the
balancing authority is maintained during these events through the curtailment of generation, including
that from wind-powered facilities.
Exhibit No. 1
Case No. IPC-E-13-22
P. DeVol, IPC
Page 18
20 oI 48
ldaho Power Company ldaho Power System Overview
System Scheduling
Idaho Power schedules its system with the primary objective of ensuring the reliable delivery of
electricity to customers at the lowest possible cost. System planning is conducted for multiple time
frames ranging from years/months in advance for long-term planning to hour-ahead for real-time
operations planning. A fundamental principle in system planning is that each time frame should be
driven by the objective of readying the system for more granular time frames. Long-term resource
planning (i.e., the tRP) should ensure the system has adequate resources for managing customer demand
over the 18-month long-term operations planning window. Long-term operations planning should
position the system such that customer demand can be managed over the balance-of-month perspective.
Balance-of-month planning should result in a system that can manage demand when scheduling
generation day-ahead. Day-ahead scheduling should enable operators to meet demand from a real-time
perspective. Finally, real-time energy schedulers should ensure the system is positioned hour-ahead such
that reliable service is maintained within the hour.
With the possible exception of the IRP, the scheduling horizons considered by Idaho Power involve
transacting with the regional wholesale market. Where the economic scheduling of system generation is
insufficient to meet demand, Idaho Power enters into contracts to purchase power off-system through its
transmission connections. Conversely, where economically scheduled generation exceeds customer
demand, surplus power is sold into the market. Importantly, Federal Energy Regulatory Commission
(FERC) rules (FERC order nos. 888/890) stipulate that surplus power sales are sourced by generating
resources that have been undesignated from network load service. Undesignation of a variable
generating resource, such as wind, for sourcing a third-party sales transaction results in the transacted
energy being given a dynamic tag, where tag is the North American Electricity Reliability Corporation
(NERC) term representing an energy transaction in the wholesale electricity market.
Balancing authorities experience considerable difficulty attracting a purchaser of dynamically tagged
energy. Therefore, as a standard operating practice, Idaho Power sources off-system power sale
contracts from its fleet of hydro and thermal generators. With their recognized level of dependability,
hydro and thermal generators can be undesignated for sourcing surplus power sales while allowing
conventional tagging procedures to be followed.
Exhibit No. 1
Case No. IPC-E-13-22
P. DeVol, IPC
Page21 ol 48
Page'19
ldaho Power System Overview
This page left blank intentionally.
Exhibit No. 1
Case No. IPC-E-1}22
P. Devol, IPC
Page 20
22 oI 48
ldaho Power Company Study Design
Sruoy Desrctt
Idaho Power designed its wind integration study with the objective of isolating in its operations
modeling the effects directly related to integrating wind generation. A common study design used
towards meeting this objective, and employed by Idaho Power for this study, is to simulate system
operations of a future year with projected wind build-outs under the following two scenarios:
o Base scenario for which the system is not burdened with the incremental balancing reserves
necessary for integrating wind
o Test scenario for which the system is burdened with the incremental balancing reserves
necessary for integrating wind
A critical feature of this design is to hold equivalent model parameters and inputs between these two
scenarios except for balancing reserves. The incremental balancing reserves built into the test scenario
simulation necessarily result in higher production costs for the system, a cost difference that can be
attributed to wind integration.
The test year selected by Idaho Power for its study is 2017. While in-service for the 500-kV Boardman
to Hemingway transmission line is not anticipated before 2018, the study still considered scenarios to
investigate the effects of the expanded transmission on wind integration costs. The study assumed
customer demand and Mid-Columbia trading hub wholesale prices as projected for 2017 in the
20r1rRP.
As noted previously, as of January 2013 Idaho Power has 678 MW of nameplate wind capacity.
Future wind penetrations considered in the study are 800 MW, 1,000 MW, and 1,200 MW of nameplate
capacity. The synthetic wind data at these penetration levels, as well as representative day-ahead
forecasts, were provided by 3TIER and Energy Exemplar. The synthetic wind data were provided for
43 wind project locations requested by Idaho Power corresponding to project sites having a current
purchase agreement with the company, as well as sites proposed to the company for future projects.
Further discussion of the study wind data and assobiated day-ahead forecasts is provided in a May 9,
2012 explanation released by the company (Appendix A).
To investigate the effect of water conditions on wind integration, the study considered Snake River
Basin stream flows for three separate historic scenarios representing low (2004), average (2009),
and high (2006) water years. Because of their importance in providing balancing reserves to integrate
wind, the HCC projects were simulated using the study model to determine their hydroelectric
generation under the selected water years. Generation for the remaining hydroelectric projects,
which are not in practice called on to provide balancing reserves for integrating wind, was entered for
the study as recorded in actual operations for the water years selected.
Exhibit No. 'l
Case No. IPC-E-13-22
P. DeVol, IPC
Page 23 of 48
Page21
Study Design ldaho Power Company
This page left blank intentionally.
Extribit No. 1
Case No. IPC-E-1*22
P. Devol, IPC
Page22
24 ol 48
ldaho Power Company Balancing Reserves Calculations and Operating Reserves
BeIRUcING RESERVES CnIcuIATIoNS AND
OpennnNc RESERVES
Critical to the two-case study design is the calculation of the incremental balancing reserves necessary
for successfully integrating the future wind penetration build-outs considered. The premise behind these
calculations is that Idaho Power's dispatchable generators must have capacity in reserve, allowing them
to respond at an acceptable confidence level to the variable and uncertain delivery of wind. Estimates of
the appropriate amount of balancing reserves were based on an analysis of errors in day-ahead forecasts
of system wind for the wind build-outs considered in the study. In addition to the synthetic time series of
hourly wind-generation data, 3TIER provided a representative day-ahead forecast of hourly wind
generation. To provide a larger sampling, Energy Exemplar created 100 additional day-ahead forecasts
having similar accuracy as the 3TIER forecast. Summaries of the synthetic wind data and day-ahead
forecasts are included in Appendix B. An illustration of this design is given in Figure 6.
1mt**'*'l_
ffi
ffi
ffi
ffi
WI'{I' GENTArIOX IfruAt
ffi
Figure 6 Wind-forecasting and generation data
In recognition of the seasonality of wind, the data were grouped by month, yielding balancing reserves
estimates specific to each month. The sample size for each month was extremely large. As an example,
for July there were 74,400 deviations between the day-ahead forecast and actual wind generation
(100 forecasts x 3l days x 24 hours). The balancing reserves requirements were calculated as the
bi-directional capacity covering 90 percent ofthe deviations. The use ofthe 90 percent confidence level
for the wind integration analysis is consistent with the criterion used for hydro conditions in assessing
peak-hour resource adequacy in integrated resource planning.
Figure 7 is an illustration of a full year of deviations for a single forecast iteration at the 1,200 MW
penetration level. In this figure, the deviations on the positive side correspond to deviations where actual
wind was lower than day-ahead forecast wind, while deviations on the negative side reflect instances
where actual wind exceeded the forecast. Importantly, the balancing reserves requirements did not cover
the full extent ofthe deviations, leaving extreme tail events in both directions uncovered.
Exhibit No. 1
Case No. IPC-E-13-22
P. DeVol, IPC
Page 25 of 48
Balancing Reserves Calculations and Operating Reserves ldaho Power Company
BE
1fi)O
800
600
400
200
o
-200
-4{m
-500
-800
-lfl)O s8 oo,z. cl
Date
Figure 7 Deviations between forecast and actua! wind generation with monthly balancing resen es
requirements (MllU)
The requirements are dynamic in that the forecast wind was taken into account in imposing the amount
of balancing reserves. For example, the requirements suggest that for the 1,200 MW wind penetration
level,295 MW of unloaded generating capacity should be held as balancing reserves in January to guard
against a drop in wind relative to the forecast. However, ifthe forecast wind generation is only 250 MW,
then the most wind can drop relative to forecast is 250 MW, which is then the amount of balancing
reserves built into the generation schedule. As a second example, if the forecast wind generation is
350 MW, the analysis of wind data indicates that balancing reserves should be held to guard against
wind dropping to 55 MW. The likelihood of wind dropping below 55 MW is small (5 percent),
and balancing reserves are not scheduled on dispatchable generators for covering a drop in wind to less
than 55 MW.
The monthly requirements for balancing reserves are given in Table 5 for the wind penetration levels
studied. The term ReS Up is used for generating capacity that can be brought online in response to a
drop in wind relative to the forecast. Reg Down is used for online generating capacity that can be turned
down in response to a wind up-ramp.
Exhibit No. 1
Case No. IPC-E-13-22
P. DeVol, IPC
Page24
26 ol 18
ldaho Power Company Balancing Reserves Calculations and Operating Reserves
Table 5 Balancing reserve requirements (MW)
Wind Gen 800 MW 1,000 MW 1,200 MW
Reg Up Reg Down Reg Up Reg Down Reg Up Reg Down
January 199 -262 246 -325 295 -390
February 252 -246 319 -297 379 -351
March 226 -295 281 -368 339 444
April 255 -353 331 450 395 -540
May 258 -290 328 -366 392 -439
June 266 -285 339 -363 409 -436
July 274 -256 355 -322 423 -384
August 172 -179 215 -224 257 -267
September 242 -219 309 -280 371 -337
October 217 -248 275 -308 329 -367
November 226 -336 277 421 333 -507
December 267 -338 326 424 394 -510
Balancing Reserves for Variability and Uncertainty in
System Demand
As described previously, power system operation has long needed to hold bidirectional capacity for
responding to variability and uncertainty in system demand. For the wind study modeling, Idaho Power
imposed a balancing reserves requirement equal to 3 percent of the system demand as capacity reserved
to allow for variability and uncertainty in load. This capacity was carried in equal amounts in the two
scenarios modeled: the base scenario where the system was not burdened with wind-caused balancing
reseryes, and the test scenario where a wind-caused balancing reserves requirement was assumed
necessary. For the test scenario modeling, the separate load- and wind-caused reserves components were
added to yield the total bidirectional balancing reserves requirement. This approach for combining the
reserves components is consistent with Idaho Power operations in practice for which system operators
receive separate forecasts for wind and demand and combine the estimated uncertainty about these
proj ections through straight addition.
Contingency Reserve Obl igation
The variability and uncertainty in demand and wind are routine factors in power system operation and
require a system to carry the bidirectional balancing reserves described in this section for maintaining
compliance with reliability standards. However, balancing authorities, such as [daho Power, are also
required to carry unloaded capacity for responding to system contingency events, which have
traditionally been viewed as large and relatively infrequent system disturbances affecting the production
or transmission of power (e.g., loss of a major generating unit or major transmission line).
System modeling for the wind study imposed a contingency reserve intended to reflect this obligation
equal to 3 percent ofload and 3 percent ofgeneration, setting aside this capacity for both scenarios
(i.e., base and test). The requirement to carry at least half of the contingency reserve obligation on
generators that are spinning and grid-synchronized was also captured in the modeling.
case No. HBI=ili:;l
P. Devol, IPC
Page27 of48
Page 25
Balancing Reserves Calculations and Operating Reserves ldaho Power Company
This page left blank intentionally.
Exhibit No. 1
Case No. IPC-E-13-22
P. DeVol, IPC
Page 26
28 ol 18
ldaho Power Company System Modeling
Svsrem Mooeuruc
Idaho Power used an internally developed system operations model for this study. The model determines
optimal hourly scheduling of dispatchable hydro and thermal generators with the objective of
minimizing production costs while honoring constraints imposed on the system. System constraints used
in the model capture numerous restrictions goveming the operation of the power system,
including the following:
r Reservoir headwater constraints
o Minimum reservoir outflow constraints
o Reservoir outflow ramping rate constraints
. Wholesale market activity constraints
o Generator minimum/maximum output levels
o Transfer capacity constraints over transmission paths
o Generator ramping rates
The model also stipulated that demand and resources were exactly in balance, and importantly that
hourly balancing reserves requirements for variability and uncertainty in load and wind were satisfied.
The incremental balancing reserves required for wind variability and uncertainty drove the production
cost differences between the study's two cases.
Day-Ahead Scheduling
The hourly scheduling determined by the model was intended to represent the optimal day-ahead system
dispatch. This dispatch schedule included generation scheduling for thermal and hydro generators,
as well as market transactions. Key inputs to the generation scheduling were the forecasts for wind
production and customer demand. These two elements of the generation/load balance commonly carry
the greatest uncertainty for power system operation in practice. A fundamental premise of reliable
operations for a balancing authority is the need to cary reasonable and prudent flexibility in the
day-ahead generation schedule, allowing the system to respond to errors in demand and wind generation
forecasts. This principle was built into the wind study modeling in the form of balancing reserves
constraints the model must honor. In the two-case study design, the system modeling for the base case
included constraints only for demand uncertainty, whereas constraints for the test case included the need
to carry additional balancing reserves for wind uncertainty. The derivation of the balancing reserves
constraints is described previously in this report.
The critical decision day-ahead generation schedulers must make involves how to schedule dispatchable
generation units taking into account the following factors:
o Forecasts for demand and wind production
o Production from other non-dispatchable resources (e.g., PPAs)
o Production from ROR hydro resources
o Operating costs ofthermal resources
o water supply for dispatchable hydro resources
.r* *". =,J8r=llrr;r
P. DeVol, IPC
Page 29 of 48
Page27
System Modeling ldaho Power Company
o Operating reserves for contingency events
o Flexibility in the schedule for dispatchable generation units allowing them to respond if
necessary to deviations between forecast and actual conditions in load and wind
The essence of wind integration and the associated costs is that the amount of balancing reserves that
must be carried is greater because of the uncertainty and variability of wind generation.
Demand and Wind Forecasts
The demand forecast used for the modeling was based on the projected hourly load used in the 201t IRP
for the calendar year 2017. The wind production forecast used for the modeling was based on the
average of the 100 forecasts provided by 3TIER and Energy Exemplar.
The forecasts for both elements were identical between the study scenarios; the test scenario simply
imposed greater balancing reseryes constraints to allow for variability and uncertainty in the wind
production forecast.
Transmission System Modeling
As noted in the Idaho Power System Overview section, the critical interconnections to the regional
market are over the Idaho-Northwest, Idaho-Utah (Path C), and Idaho-Montana paths. For the
wind-study modeling, the separate paths were combined to an aggregate path for off-system access.
Every October, Idaho Power submits a request to secure firm transmission across its network based on
its expected monthly import needs for the next l8 months. The maximum levels used in the modeling
for firm import capacity were based on the October 2010 request. The modeling assumed additional
import capacity using non-firm transmission. Non-firm imports were assessed a $50/NIWh penalty
designed to represent the less favorable economics associated with non-firm transmission and typical
hourly pricing. The export limits were based on typical levels of outbound capacity observed in practice.
The transmission constraints in Table 6 were used in the wind study modeling.
Table 6 Modeled transmission constraints (MW)
Maximum Firm lmport (MW)
Maximum Nonfirm
lmport(MW)Maximum Export (MW)
January
February
March
April
May
June
July
August
September
October
November
December
179
35
0
0
320
262
149
230
217
0
113
325
500
500
500
500
500
500
500
s00
500
500
500
500
300
300
300
300
300
300
300
300
300
300
300
300
Exhibit No. 'l
Case No. IPC-E-I3-22
P. Devol, IPC
Page 28
30 of 48
ldaho Power Company System Modeling
Idaho Power's transmission network is a fundamental part of the vertically integrated power system,
and allows the company to participate in the regional wholesale market to serve load or for economic
benefit. However, Idaho Power does not view its transmission network with associated regional
interconnections as a resource for providing balancing reserves allowing it to respond to variability and
uncertainty in wind generation and customer demand. In the region, each balancing authority provides
its own balancing reserves. Idaho Power provides its balancing reserves from company-owned
dispatchable generation units (thermal and hydro).
Idaho Power also investigated scenarios with the 500-kV Boardman to Hemingway transmission line.
For these scenarios, the maximum firm import constraint was increased by 500 MW during April
through September and by 200 MW for the remainder of the year. The maximum export constraint was
increased by 150 MW throughout the year. The following transmission constraints were used in the
wind study modeling for the system with the proposed Boardman to Hemingway transmission line.
Table 7 Modeled transmission constraints-simulations with 500-kV Boardman to Hemingway
transmission Iine (MW)
Marimum Firm lmport (MW)Maximum Non-firm lmport (MW) Maximum Export (MW)
January
February
March
April
May
June
July
August
September
Oclober
November
December
379
23s
200
500
820
762
649
730
717
200
313
525
300
300
300
300
300
300
300
300
300
300
300
300
650
650
650
650
650
650
650
6s0
650
650
6s0
6s0
Overgeneration in System Modeling
At a fundamental level, the reliable scheduling of the power system is based on the following
simple equation:
Fo re c ast lo ad: Fo re cast gene rot ion
An expanded form ofthis equation is as follows:
Forecast retail sales * Forecast wholesale sales
Forecast dispatchable generation * Forecast wind generation -l Forecast other generation
Exhibit No. 1
Case No. IPC-E-13-22
P. DeVol, IPC
Page 31 of 48
System Modeling ldaho Power Company
In the expanded equation, dispatchable generation includes scheduled production from resources the
balancing authority (i.e., Idaho Power) can vary at its discretion to achieve reliable and economic system
operation. Built into this term of the equation is the bidirectional balancing reserves intended for use in
case the forecasts for demand or wind generation are incorrect. The other generation in the expanded
equation is the amount of energy that cannot be varied. This term includes minimum generation levels at
baseload thermal plants, ROR hydro generation, and non-wind power purchased under contract.
At times, the left side of the equation can become very low; Idaho Power customer use is low and
wholesale exports are capped by transmission capacity. During these times, providing the balancing
reserves necessary for responding to wind, specifically for responding to wind up-ramps, is not possible
without upsetting the balance between the two sides of this equation. In effect, the terms of the right
side of the equation cannot be reduced enough to match the left. For these times, the wind study
modeling assumed the wind, or potential wind, was excessive and could not be accepted; curtailment of
wind energy was necessary to maintain balance. Further discussion of overgeneration and curtailment is
provided in the following section.
Exhibit No. 1
Case No. IPC-E-13-22
P. Devol, IPC
Page 30
32 of 48
ldaho Power Company
Resulrs
As noted previously, the objective of this study is to answer two fundamental questions:
l. What are the costs of integrating wind generation for the Idaho Power system?
2. How much wind generation can the Idaho Power system accommodate without
impacting reliability?
Thus, the results produced by the study's system modeling were designed to address these
two questions.
Wind lntegration Costs
From a cost perspective, a comparison of annual production costs between two scenarios having
different balancing reserves requirements-where the difference in balancing reserves is related to
wind's variability and uncertainty-was used to estimate the costs of integrating wind. The production
cost difference between scenarios was divided by the annual MWh of wind generation to yield an
estimated integration cost expressed on a per MWh basis. The integration cost calculation is summarized
as follows:
o Base scenario for which the system was not burdened with incremental balancing reserves
necessary for integrating wind (wind integration is "not our problem", a theoretical case used as
a benchmark for comparing costs)
o Test scenario for which the system was burdened with incremental balancing reserves necessary
for integrating wind
The wind integration cost is the net-cost difference of the two scenarios divided by the MWh of wind
generation (the amount of wind generation was the same in both scenarios):
Wind integration cost = Iesf scenario nef cosf - Base scenario nef cosf
Wind generation in MWh
As noted earlier, the study included three water years and three wind penetration levels.
These conditions are shown in Table 8.
Table 8 Wind penetration levels and water conditions
Wind Penetration Leve! (MW Gapacity)Water Year
Low (2004)
Average (2009)
High (2006)
A matrix of the wind integration costs on a per MWh basis is given in Table 9. These costs are based on
a system without the proposed Boardman to Hemingway transmission line.
800
1,000
1,200
Exhibit No. 1
Case No. IPC-E-13-22
P. DeVol, IPC
Page 33 of 48
Page 31
ldaho Power Company
Table 9 lntegration costs ($/MWh)
Nameplate Wind
Water Condition 800 Mw 1,000 Mw 1,200 MW
Average (2009)
Low (2004)
Hish (2006)
Average
$7.18
$7.26
$9.73
$8.06
$11.94
$12.44
$14.79
$13.06
$18.1s
$18.15
$20.73
$19.01
The addition of the Boardman to Hemingway transmission line reduced integration costs slightly.
Table l0 provides the wind integration costs for a system having the proposed Boardman to Hemingway
transmission line.
Table 10 lntegration costs with the Boardman to Hemingway transmission line ($/MWh)
Nameplate Wind
Water Condition 8OO MW 1,000 i'w 1,200 MW
Average (2009)
Low (2004)
High (2006)
Average
$6.51
$6.66
$9.72
$7.63
$11.03
$11.04
$13.78
$1r.95
$16.38
$16.67
$19.53
$r7.53
lncremental Cost of Wind lntegration
The integration costs previously provided in Tables 9 and l0 represent the cost per MWh to integrate the
full installed wind at the respective penetration levels studied. For example, the results of Table 9
indicate that the full fleet of wind generators making up the 800 MW penetration level bring about costs
of $8.06 for each MWh integrated. However, wind generators comprising the 678 MW of current
installed capacity on the Idaho Power system are assessed an integration cost of only $6.5044Wh2.
In order to fully cover the $8.064{Wh integration costs associated with 800 MW of installed wind
capacity, wind generators in the increment between the current penetration level (678 MW) and the 800
MW penetration level will need greater assessed integration costs. Study analysis indicates that these
generators will need to recognize integration costs of $16.7044Wh to allow full recovery of integration
costs associated with 800 MW of installed wind capacity. Similarly, generators between the 800 MW
and 1000 MW penetration levels introduce incremental system operating costs requiring the assessment
of integration costs of $33.42lI\4Wh, and generators between 1000 MW and 1,200 MW require
incremental integration costs of $49.4644Wh. A graph showing both integration costs and incremental
integration costs is provided in Figure 8 below. The incremental integration costs are summarized in
Table 11.
Exhibit No. 1
Case No. IPC-E-13-22
P. Devol, IPC
2 Integration cost stipulated by Idaho Public Utilities Commission Case No. IPC-E-07-03, Order No. 30488.
Page 32
34 of 48
ldaho Power Company Results
S4o.E
=Eo;iI ssozotrf,(,r2 s2o
Figure 8
Table 11
..D rNcR EM EiffAr COSr (s/MWh)
+wlilD INTEGRAnON COST (s/tvtWhl
Integration costs with incremental integration costs ($/MWh)
lncrementa! wind integration costs ($/MWh)
Nameplate Wind
578 - 800 MW 800 - 1,000 Mw 1,000 - 1,200 illw
lncrcmental cost per MWh $16.70 $33.42 $49.46
Spilling Water
The modeling suggests that providing balancing reserves to integrate wind leads to increased spill at the
HCC hydroelectric projects. Spill is observed in actual operations during periods of high Brownlee
Reservoir inflow coupled with minimal capacity to store water in the reservoir. Minimal storage
capacity at Brownlee occurs when the reservoir is nearly full or when the reservoir level is dictated by
some other constraint, such as a flood control restriction. Flow through the HCC cannot be significantly
reduced during these periods; the three-dam complex is essentially operated as a ROR project during
these high-flow periods. As a consequence, holding generating capacity in reserve for balancing case No. HllI]Yil
P. DeVol, IPC
Page 35 of 48
Page 33
Results ldaho Power Company
purposes is frequently achieved only through increasing project spill, rather than reducing turbine flow.
Table 12 provides the total incremental HCC spill in thousands of acre-feet (kaf) associated with
integrating wind.
Table 12 lncremental Hells Canyon Complex spill (thousands of acre-feet)
Nameplate Wind
Water Condition 800 Mw 1,000 tuw 1,200 MW
Average (2009)
Low (2004)
High (2006)
534 kaf
33 kaf
2,101kat
949 kaf
93 kaf
2,698 kaf
1,446kat
255kaf
2,9't6 kaf
Simulations for the high water condition (2006) with 800 MW of wind capacity provide a good
illustration of the effect of wind integration on spill. Under the base scenario, the theoretical "not our
problem" case, wind study system simulation shows spill totaling 3,590 kaf at Brownlee alone.
For reference, this simulated spill is within 5 percent of the actual total Brownlee spill in 2006,
which was about 3,800 kaf. By comparison, the total Brownlee spill under the test scenario,
where integrating wind is ldaho Power's problem, is 4,475 kaf. The excess spill under the test scenario
translates to about 185 gigawatt hours (GWh) of lost power production at Brownlee----energy that is no
longer available for serving load or off-system sales.
Maximum ldaho Power System Wind Penetration
The capability ofthe Idaho Power system to integrate wind is finite. The rapid growth in wind capacity
connecting to the system over recent years has heightened concern that the limits of this integration
capability are being neared, and that development beyond these limits will severely jeopardize system
reliability. The quantity of wind generation Idaho Power can integrate varies throughout the year as a
function of customer load. During times of high load, Idaho Power can integrate more wind than during
times of low load.
Modeling performed for the wind study has demonstrated the occurrence during low load periods where
the balancing reserves necessary for responding to a wind up-ramp (i.e., generation that can be
dispatched down in response to an increase in wind) cannot be provided without pushing the system to
an overgeneration condition. Customer load for these periods, where load consists of sales to retail
customers and to wholesale customers by way of regional transmission connections, is too low to allow
for the integration of a significant quantity of wind. This situation requires curtailment of wind
generation to maintain system balance. For the wind study modeling, the curtailed wind generation was
removed from the production cost analysis and consequently did not affect the calculated integration
cost. Curtailed wind was not integrated in the modeling and had no influence on the calculated
integration costs. Not surprisingly, curtailment was found in the wind study modeling to have a strong
correlation with customer load, water condition, and wind penetration levels. A summary of the amount
of curtailment in the study is provided in Table 13.
Exhibit No. 1
Case No. IPC-E-13-22
P. Devol, IPC
Page 34
36 of 48
ldaho Power Company Results
Table 13 Curtailment of wind generation (annual MWh)
Nameplate Wind
Water Gondition 800 itlw {,000 Mw 1,200 iiw
Average (2009)
Low (2004)
High (2006)
Average
738l'n /h
204]rrwh
890 t'il\ rh
6fl MWh
8,755lr^ /h
3,494 t'A rh
12,519 MWh
8,256 MWh
48,942IrIWh
29,574 trA /h
61,557 IrA /h
46,691 MWh
c
E *,0-tt zs,m(,
.E
=
2o,mo
E
Figure 9 illustrates the projected exponential increase in curtailment as a function of the wind
penetration level.
50,m
4t(m
40,(m
35,ff)o
Figure 9 Gurtailment of wind generation (average annual MWh)
A key feature of Figure 9 is the rapid acceleration of projected curtailment as installed wind capacity
increases beyond the 800 MW level. The addition of 200 MW of installed wind capacity from 800 MW
to 1,000 MW is projected to result in about 7,600 MWh of additional curtailment. Increasing the
installed wind capacity 200 MW further to 1,200 MW is projected to result in another 38,000 MWh of
curtailment. It is important to note the effect of a procedure for curtailment. Spreading the curtailed
MWh over the full installed wind capacity of 1,200 MW results in a projected curtailment of about
1.5 percent of produced wind energy. However, if wind generators comprising the expansion from
1,000 MW to 1,200 MW are required under an established policy to shoulder the curtailment burden
arising from their addition to the system, curtailment of their energy production is projected to reach
Exhibit No. 1nearly 8.5 percent. case r.ro.
-r6i'_Eii-l
Page 37 of 48
Page 35
Results ldaho Power Company
The study results suggest that the occurrence of low load periods for which curtailment is necessary is
likely to remain relatively infrequent for wind penetration levels of 800 MW or less.
However, the results indicate that operational challenges are likely to grow markedly more severe with
expanding wind penetration beyond 800 MW of installed nameplate capacity. The occurrence of low
load periods for which balancing reserves cannot be provided without causing overgeneration is
expected to become more frequent and require deeper curtailment of wind production. This is
particularly true in that it is often necessary to maintain the operation of thermal (i.e., gas- and
coal-fired) generators during periods of low load and high wind, in order to have the dispatchable
generation from these resources available should customer loads increase or winds decrease.
Effect of Wind lntegration on Thermal Generation
Idaho Power operates its coal resources to provide low-cost, dependable baseload energy.
However, the study results suggest that the operation of the company's coal resources is likely to
decrease on an annual basis with expanding wind penetration. The reduction in coal output is principally
the result of displacement of coal generation by wind generation, as well as the displacement by flexible
gas-fired plants required to help balance the variable and uncertain delivery of wind.
The operation ofcoal-fired generators has been affected by energy oversupply conditions over recent
years in the Pacific Northwest. Coal plants have historically been operated less during periods of high
hydro production, and maintenance is typically scheduled to coincide with spring runoff when customer
demand is relatively low. However, the expansion of wind capacity over recent years in the region has
caused overgeneration conditions to become more severe and longer lasting, leading to extended periods
during which prices in the wholesale market have been very low or negative. The effect on coal plants
has been a decline in annual energy production. However, during periods when customer load is high,
such as during summer 2012,ldaho Power's coal fleet is consistently relied upon for energy to meet the
high customer demand.
While the operation of baseload coal-fired power plants is expected to decline as a consequence of
adding wind to a power system, this decline is offset by a marked increase in generation from gas-fired
plants. The rapidly dispatched capacity from the gas-fired plants is widely recognized as critical to the
successful integration of variable generation. Wind study modeling suggests that the need to dispatch
gas-fired generators for balancing reserves is likely to displace the economic operation of coal-fired
generators, particularly during times of acute transmission congestion.
This situation where relatively low-cost baseload resources are displaced by flexible cycling plants
(i.e., gas-fired) is described in a2010 NREL report (Denholm et al. 2010). Table l4 lists the annual
generation from the wind study modeling for thermal resources for the case when Idaho Power is
responsible for providing the balancing reserves and integrating the wind energy.
Table 14 Annual generation for thermal generating resources for the test case (GWh)
Nameplate Wind
Thermal Resource 8OO MW 1,000 Mw 1,200 mw
Coal-fired
Gas-fired
7,568 GWh
963 GWh
7,291 GWh
1,238 GWh
6,851 GWh
1,918 GWh
Exhibit No. 1
Case No. IPC-E]13-22
P. DeVol, IPC
Page 36
38 of 48
ldaho Power Company Recommendations and Conclusions
RecoUIMENDATIONS AN D COI.ICI-USIONS
Idaho Power has 678 MW of nameplate wind generation on its system. This is a growth in wind capacity
of about 290 MW over the last two years, and 490 MW over the last three. The explosive growth in
wind generation has heightened concems that the finite capability of Idaho Power's system to integrate
wind is being rapidly depleted. Because ofthese concerns, the objective of this investigation is to
address not only the costs to modifr operations to integrate wind, but also the wind penetration level at
which system reliability becomes jeopardized. The questions that drove the investigation are the
following:
l. What are the costs of integrating wind generation for the Idaho Power system?
2. How much wind generation can the Idaho Power system accommodate without
impacting reliability?
The study utilized a two-scenario design, with a base scenario simulation of operations for a system that
was not burdened with incremental balancing reserves for integrating wind and a test scenario
simulation for a system burdened with incremental wind-caused balancing reserves. Averaged over the
three water conditions considered, the estimated integration costs are $8.06/MWh at 800 MW of
installed wind, $l3.06A,lwh at 1,000 MW of installed wind, and $19.01/IrrIWh at 1,200 MW of installed
wind. A summary of the estimated costs is given in Table 15.
Table 15 lntegration costs ($/MWh)
Nameplate Wind
Water Gondition 800luw 1,000 Mw 1,200 MW
Average (2009)
Low (2004)
Hish (2006)
Average
$7.18
$7.26
$9.73
$8.06
$11.94
$12.44
$14.79
$13.06
$18.15
$18.1s
$20.73
$r9.01
Importantly, the system modeling conducted for the study indicates a major determinant of ability to
integrate is customer demand. This finding is not to be confused with the pricing of wind contracts and
the wide recognition that wind occurring during low load periods is of little value. Instead, the study
indicates that during periods of low load, the system of dispatchable resources often cannot provide the
incremental balancing reseryes paramount to successful wind integration without creating an imbalance
between generation and demand. Modeling demonstrates that the frequency of these conditions is
expected to accelerate greatly beyond the 800 MW installed capacity level, likely requiring a sharp
increase in wind curtailment events. Even at current wind penetration levels, these conditions have been
observed in actual system operations during periods of high stream flow and low customer demand.
While the maximum penetration level cannot be precisely identified, study results indicate that wind
development beyond 800 MW is subject to considerable curtailment risk. It is important to remember
that curtailed wind generation was removed from the production cost analysis for the wind study
modeling, and consequently had no effect on integration cost calculations. The curtailed wind generation
simply could not be integrated, and the cost-causing modifications to system operations designed to
allow its integration were not made. The curtailment of wind generation observed in the wind study
modeling is shown in Figure 10.
",," *".''J8-E1lll;l
P. DeVol, IPC
Page 39 of 48
Page 37
Recommendations and Conclusions ldaho Power Company
50,mo -
1 45.000 ,
40,ooo .
35,0O0
c9 Eo,omE=ots zs.mo
=Ur
=
lo,ffi
E
15,(nO
10.m0
5,000
O i-
o 200 400
Wind Capacity (MW)
Figure 10 Curtailment of wind generation (average annual MWh)
Conversely, during periods of high customer demand, the dispatchable resources providing the
balancing reserves for integrating wind are needed and thus are positioned at levels where they are ready
to respond to changes in wind. While the costs to integrate wind still exist during these higher customer
demand periods, the system can much more easily accommodate high levels of wind without impacting
system reliability.
lssues Not Addressed by the Study
The focus of this study was the variability and uncertainty of wind generation. The study then
established that these attributes of wind bring about the need to have balancing reserves at the ready on
system dispatchable resources, and finally that having balancing reserves for integrating wind brings
about greater costs of production for the system. A consideration not addressed by the study is the
increased maintenance costs expected to occur for thermal generating units called on to frequently adjust
their output level in response to changes in wind production or that are switched on and off on a more
frequent basis. The effect of wind integration on these costs is likely to become evident and better
understood with the expanded cycling of these thermal generators accompanying the growth in wind
generation over recent years.
The controlof system voltage and frequency is receiving considerable attention in the wind integration
community. It is widely recognized that the addition of wind generation to a power system has an impact
on grid stability. On some transmission systems, controlling system voltage and frequency during large
ramps in generation within acceptable limits can be challenging. Idaho Power's system has not yet
exhibited this problem at current wind penetration levels. However, growth in wind penetration beyond
the current level will lead to greater challenges in maintaining system voltage and frequency within
control specifications of the electric system, and likely increase the incidence of excursions where Exhibit No. 1
Case No. IPC-E-13-22
P. DeVoI, IPC
Page 38
Page 40 of 48
ldaho Power Company Recommendations and Conclusions
system frequency deviates from normal bands. The effects of frequency excursions may extend to
customer equipment and operations.
Measures Facilitating Wind !ntegration
Idaho Power recognizes the importance of staying current as operating practices evolve and innovations
enabling wind integration are introduced. Some changes in operating parameters include mechanisms
such as Dynamic Scheduling System (DSS), ACE Diversity Interchange (ADI), and intra-hour markets.
Further development of these measures will, to varying degrees, make it easier for balancing authorities
to integrate the variable and uncertain delivery of wind generation. At this time, it is Idaho Power's
judgment that the effect of these measures is not substantial enough to warrant their inclusion in the
modeling performed for this study.
An additional measure that has been studied over recent years as a Western Electricity Coordinating
Council (WECC) field trial is reliability-based control (RBC).The essential effect of RBC on operations
is that a balancing authority is permitted to carry an imbalance between generation and demand if the
imbalance helps achieve wider system stability across the aggregated balancing area of the participating
entities. In effect, the balancing authority area is expanded, and the diversity of the expanded area allows
an aggregate balance to be more readily maintained. Idaho Power has participated in the RBC field trial
since the program's inception, and has recognized a resulting decrease in the amount of cycling required
of generating units for balancing purposes. However, the effect of RBC was not included in the
modeling for this study. This omission is in part related to the status of the program as a field trial,
and related uncertainty regarding the structure of RBC in the future, or whether RBC will exist at all.
Moreover, while RBC may allow balancing reserves-carrying generators to not respond to changes in
load or wind in real-time operations, the scheduling of these generators must still include appropriate
amounts of balancing reserves because it is not known at the time of scheduling to what extent an
imbalance between generation and load will be permitted.
Future Study of Wind Integration
Idaho Power continues to grapple with new challenges associated with wind integration. The expansion
in installed wind capacity over recent years has made the establishment of a best management plan for
integrating wind problematic; the amount of installed wind simply keeps growing. It is commonly
understood that wind does not always blow, leading to the legitimate concern about having backup
capacity in place for when wind generators are not producing. Somewhat ironically,
integration experience over recent years throughout the Pacific Northwest has led to heightened
concerns about what to do when wind generators are producing and that production is not needed and
unable to be stored in regional reservoirs because of minimal storage capacity, and the balancing
reserves carried on dispatchable generators only add to the amount of unneeded generation. While it has
been recognized that balancing reserves need to be carried for responding to wind up-ramps
(i.e., balancing reserves need to be bidirectional), it has only recently become apparent that the tdaho
Power system, and even the larger regional system, at times cannot provide these balancing reserves.
This experience has shown that it is diflicult to predict the integration challenges of tomorrow, but it is
safe to say that there will be a need for continued analysis as additional tools, methods, and practices for
integrating wind become available.
Idaho Power has experienced success in wind-production forecasting. The company has developed an
intemal forecast model which system operators are using with increasing confidence. It is likely that the
future study of wind integration will make use of this forecast model, specifically in that its relative
accuracy will ultimately lead to a reduction in the balancing reserves requirement for wind integration. t
",* *" ''J3-Ellllil
P. Devol, IPC
Page 41 of 48
Page 39
Recommendations and Conclusions ldaho Power Company
However, even accurate wind forecasting cannot eliminate the need for curtailment when wind
generation creates a significant imbalance between load and generation.
Finally, the wider region beyond Idaho has added considerable wind capacity over recent years, much of
the growth driven by requirements associated with state-legislated renewable portfolio standards.
Most of the wind generation has been added outside of local or regional integrated resource planning
efforts. The addition ofthis generating capacity has resulted in recurring energy oversupply issues for
the region, a situation that has led the BPA to propose a protocol for managing oversupply (BPA 2013).
Regional market prices during these oversupply periods have experienced pronounced declines to very
low or even negative levels. Sometimes even the larger regional system and larger regional market
cannot successfully integrate all of the wind energy that is produced. It is critical that future modeling
for studying wind integration continues to capture the regional expansion of wind generation and its
effect on the wholesale market.
Exhibit No. 1
Case No. IPC-E-13-22
P. DeVol, IPC
Page 40
42 ol 48
ldaho Power Company Literature Cited
LrreRlruRE Grreo
Bonneville Power Administration (BPA). February 2013. Oversupply management protocol, version 2.
http://transmission.bpa.eov/ts_businessJractices/Content/9_Redispatch_and_Curtailment/Oversuop
ly met protocol.htm. Accessed on: February 2,2013.
Denholm, P., E. El4 B. Kirby, and M. Milligan 2010. The role of energy storage with renewable
electricity generation. U.S. Department of Energy, Office of Energy Efficiency and Renewable
Energy, National Renewable Energy Laboratory. Technical Report NREL/TP-6A2-47187,
January 2010.
GE Energy 2010. Western wind and solar integration study. Prepared for: U.S. Department of Energy,
Office of Energy Efficiency and Renewable Energy, National Renewable Energy Laboratory.
Subcontract Report NREL/S R-5 5 0 - 47 43 4, May 20 I 0.
Exhibit No. 1
Case No. IPC-E-13-22
P. DeVol, IPC
Page 43 of 48
Page 41
Literature Cited ldaho Power Company
This page left blank intentionally.
Exhibit No. 1
Case No. IPC-E-13-22
P. DeVol, IPC
Page42
14 ot 48
ldaho Power Company Appendix A
Appendix A. May 9,2012, Explanation on wind data
Wrno IrurecmnoN WoRKSHoP
SruoY Wrtrto DnrR ExplaNlnoru
May 9,2012
Idaho Power received questions during the April 6 wind integration workshop related to the synthetic
wind data used for its study of wind integration. The company recognizes the importance of using
high-quality wind data, and consequently indicated at the workshop that it would thoughtfully review
the wind data in an effort to address the questions raised. As stated at the workshop, the wind data used
for the study were provided by 3TIER. 3TIER provided these data for 43 wind project locations
requested by ldaho Power corresponding to project sites having a current purchase agreement with the
company, as well as sites proposed to the company for purchase agreement. The 43 wind project
locations are given as Attachment No. 3 to comments filed by Idaho Power with the IPUC on
December 22,20103. It is important to note that 3TIER did not select from the more than 32,000
existing or hypothetical wind project sites used for the Western Wind and Solar Integration Study
(WWSIS), but instead pulled new time series directly from the WWSIS gridded model data set precisely
at the 43 locations requested by Idaho Power. Thus, the geographic diversity of the synthetic wind
data provided by 3TIER is representative of the geographic diversity for projects proposed to
Idaho Power.
3TIER also provided a synthetic day-ahead forecast for the wind generation time series. In providing
this forecast, 3TIER notes that a bias found in the forecast during completion of the WWSIS was
corrected on a site-by-site basis for the ldaho Power wind study, as opposed to the regional bias
correction used for the WWSIS. The site specific correction is preferable to the regional correction
because it mimics real forecasting practice, where project data at each site would be used to eliminate
long-term bias from the forecast. With respect to accuracy of the synthetic day-ahead forecast,
3TIER reports that hourly wind speed forecast errors for ten operational sites in Idaho or neighboring
states were compared to similarly calculated errors for the synthetic day-ahead forecast. 3TIER reports
that this comparison yielded values for mean absolute error and root mean squared error for the synthetic
day-ahead forecast only about l5% higher than equivalent statistics for the real errors at the ten
operational sites in the Idaho vicinity. This result suggests that the error characteristics of the
synthetic forecasts are very similar to those of actual wind forecasts.
To validate the synthetic actual wind time series, 3TIER has completed validation reports describing the
results of comparisons between the synthetic wind data and public tower data. The complete set of
validation reports for the WWSIS can be found through the NREL website4. Five of the validation
towers are located in ldaho. Review of these reports indicates that the synthetic actual wind time series
capture the seasonal and diumal wind cycles fairly well; however, the synthetic time series are
consistently low biased, at a 3TlER-reported average level of about -1.2 m/s at the five validation sites.
There is basis in suggesting that the low bias, while reducing the total production of modeled wind
projects, would have minimal impact on the overall variability of the synthetic actual wind time series,
and would consequently have little effect on the estimated integration cost.
3 Iduho Power Comments, Idaho Public Utilities Commission Case GNR-E-10-04, Attachment No. 3.
a http://wind.nrel.gov/public/WWIS/ValidationReports/wwis-vrpts.html#vmap Exhibit No. 1
Case No. IPC-E-13-22
P. DeVoI, IPC
Page 45 of 48
Page 43
Appendix A
However, Idaho Power recognizes the critical nature of the synthetic wind data used for the study,
and will discuss this low bias further with the technical review committee it has formed.
Finally, the synthetic actual wind time series created for the WWSIS have been found to exhibit
excessive ramping as described in the WWSIS final report and as reported by NRELS. The excessive
ramping in the WWSIS wind data occurs because the mesoscale model used to generate the synthetic
wind data was run in 3-day sections. Smoothing techniques were used to reduce the ramping across the
seam at the end of each third day; however, 3TIER reports that excessive variability remains in the
WWSIS wind data. 3TIER also reports that review of the synthetic actual wind time series data pulled
for the ldaho Power study indicates similar excessive ramping, with ramps tending to be 1.5 to 2.0 times
larger from two hours before to eight hours after the start of every third day. While Idaho Power intends
to discuss this condition with its technical review committee, the company believes that only a small
fraction of hours are affected, and that consequently the impacts on integration cost are likely small.
Idaho Power hopes that this follow-up helps to address questions on the wind data raised at the April 6
workshop. We value the questions and feedback received from workshop participants, and welcome
remaining questions related to the wind data or other features of the wind study. We are planning a
meeting with our technical review committee in early May, and are looking forward to the added value
this group will bring to our effort.
Idaho Power,l22l W Idaho Street,
Boise, Idaho 83702
email: I P C _Wind _$tudy@Idaho P ow e r. com
Exhibit No. 1
Case No. IPC-E-13-22
P. DeVol, IPC
s http://www.nrel.gov/wind/integrationdatasets/pdfs/westem/2009/westem-dataset-irregularity.pdf
Page 44
46 of 48
ldaho Power Company Appendix B
Appendix B. Wind data summaries
Table Bl Monthly and annual capacity factors (percent of installed nameplate capacity)
Nameplate Wind
Month 800 Mw 1,000 Mw 1,200 MW
January
February
March
April
May
June
July
August
September
October
November
December
Annual
30Yo
20Yo
31Yo
38Yo
24Yo
29Yo
20Yo
17o/o
18o/o
23o/o
360/0
38%
27%
30%
20Yo
32%
38%
24Yo
29o/o
19Yo
17o/o
18o/o
23o/o
35o/o
38o/o
27o/o
30%
19o/o
32%
37o/o
24o/o
29o/o
19o/o
17o/o
18%
23%
35%
38o/o
27%
Note: Wind generation data for study provided by 3T|ER.
Exhibit No. 1
Case No. IPC-E:13-22
P. Devol, IPC
Page 47 of 48
Page 45
Appendix B ldaho Power Company
This page left blank intentionally.
Exhibit No. 1
Case No. IPC-E-13-22
P. DeVol, IPC
Page 46
48 of48