he Value of HVDC for Continental Power Systems:
The Interconnection
Seams Study
A. Bloom, Member, IEEE, J. Novacheck, Member, IEEE, J. D. McCalley, Fellow, IEEE, Y. Makarov, N. Samaan,
1

Senior Member, IEEE, A. L. Figueroa-Acevedo, Member, IEEE, A. Jahanbani-Ardakani, H. Nosair, A.
Venkatraman, J. Caspary, D. Osborn, J. Lawhorn, R. Johnson, and J. Lau, Member, IEEE
II. APPROACH
Abstract—The Interconnections Seam Study examines the
potential economic value of increasing electricity transfer
between the Eastern and Western Interconnections using highvoltage direct current (HVDC) transmission and optimizing both
generation and transmission resources across the country. The
study conducted a holistic multi-model analysis which used cooptimized generation and transmission expansion planning,
production cost modeling, and AC power flow analysis. Four
future scenarios were developed and studied to quantify and
observe potential benefits. The results show benefit-to-cost ratios
that reach as high as 3.3 and annual operational savings
exceeding $2 billion US, indicating significant value to increasing
the transmission capacity between the interconnections and
sharing of generation resources.
Index Terms—Power generation planning, Power generation
dispatch, HVDC transmission, Wind power generation, Solar
power generation, Power system economics, Power system
reliability, Resource adequacy.

A

I. INTRODUCTION

t the western edge of the American prairie, just east of
the Rocky Mountains, lies a collection of electrical
transmission resources that string together the U.S.
and Canadian Eastern and Western Interconnections (EI and
WI). Seven back-to-back (B2B) high voltage direct current
(HVDC) facilities enable 1,320 megawatts (MW) of electricity
to flow between the U.S. EI and WI.1 The transfer capability
between the interconnections isn’t much more than a rounding
error compared to the networks they connect—the larger
Eastern Interconnection, is home to 700,000 MW of
generating capacity and the capacity of the WI has roughly
250,000 MW of capacity. But these B2B facilities, located
strategically at the “seam” where the East meets the West, are
aging, and thus present a timely and impactful opportunity for
utilities, developers, regulators, and policy makers to
modernize the U.S. electric grid.
Over the last 95 years, there have been a number of studies
that investigated joining together the North American
electrical systems. The earliest study (Tribune) , in 1923,
imagined the potential opportunity to grow industry by joining
together the continent’s hydro and thermal resources using the
relatively new idea of ultra high voltage transmission.
Subsequent studies ( WAPA,Bor, BPA,DoE ) investigated
joining the existing systems and limited consideration of new
resources and the potential evolution of the power system.
More recent analysis (REF &MacDonald/Clack) considered
……. The work presented below is the first practical
investigation of the potential to join together the Eastern and
Western Interconnections and the first to use multiple power
system models to evaluate resource adequacy and security
using the most detailed models ever made of the US electricity
systems.
1 An additional 150 MW of B2B transmission capacity is in Alberta,
Canada, and was not studied in this work.

The Interconnections Seam Study is a coordinated
transmission planning analysis of two major U.S.
interconnections using multiple models and methods. The goal
of the study is to identify the potential economic value of more
tightly integrating the EI and WI using HVDC transmission
technology. A technical review committee (TRC) comprised
of over 50 industry representatives met on six occasions to
discuss the approach, methods, scenarios, data, assumptions,
and results of the study. This study is the initial valuation of
increasing grid interconnection and should not be referenced
as final ready-to-build designs.
First, detailed capacity expansion analysis built four future
transmission scenarios that met long term reliability (resource
adequacy) requirements under two different policy
assumptions. Then, detailed nodal transmission models were
created to evaluate the various resource and transmission
scenarios needed to reliably schedule and dispatch to meet
8,760 hours of demand. Finally, we took initial steps to create
an alternating current (AC) power flow framework of the
interconnections to enable future analysis of the security of
various designs.
This study uses an integrated engineering approach that
includes both research-grade and commercial modeling tools.
For co-optimized generation and transmission expansion, Iowa
State University’s co-optimized generation and transmissionplan (CGT-Plan) was used for generation and transmission
expansion, Energy Exemplar’s PLEXOS was used for
production cost modeling (PCM), and Siemens’ PSS®E was
used for AC power flow analysis.
The results of the study indicate benefit-to-cost ratios as
high as 3.3 and annual operational savings exceeding $2
billion. Based on these results, the researchers and industry
partners believe that additional and more granular evaluation
is merited.
The capacity expansion model minimizes the net-present
value (NPV) of four HVDC transmission options along with
new AC transmission; new generation; retired generation; the
provision cost of regulation up; regulation down and
contingency reserves; and the fixed and variable operation and
maintenance cost (FOM and VOM) of new and existing
generation resources. Through this optimization, the capacity
expansion model produces generation and transmission
expansion in the technology type and MW amounts.
The first scenario, Design 1 (D1), maintains existing B2B
HVDC capability. In Design 2a (D2a), existing B2B HVDC
facilities are allowed to expand. Design 2b (D2b) allows
existing B2B HVDC facilities to expand (like D2a) and adds
three prescribed HVDC lines that cross the EI and WI. In
Design 3 (D3), a macrogrid overlay is created that includes a
variety of HVDC lines within and between the
interconnections.

DRAFT-for review purposes only. Send comments to aaron.bloom@nrel.gov

 2

III. INPUT DATA AND ASSUMPTIONS
The Interconnections Seam Study utilized a variety of input
data and assumptions to build a power system representation
of the EI and WI. The expected generation and transmission
for the EI and WI was obtained from North American Electric
Reliability Corporation (NERC) regional entities. The Eastern
Interconnection Reliability Assessment Group’s (ERAG)
Multiregional Modeling Working Group (MMWG) 2026
summer case and the Western Electricity Coordinating Council
(WECC) Transmission Expansion Planning Policy Committee
(TEPPC) 2024 common case were chosen as the starting point
for creating an updated nodal representation of the 2024 EI
and WI. Additional information on the 2024 data can be found
in [1].
The capacity expansion model optimizes with initial
conditions established in 2024. The capacity expansion model
has 169 buses that represent the full 100,000 nodal
transmission network of 2024. A full description of the model
is available at [2]. A variety of resources were available for
expansion according to their capital cost. Optimization
parameters included load growth, carbon pricing, wind
maximum expansion restrictions, and distributed solar
photovoltaic (PV) growth.
The capacity expansion planning program CGT-Plan was
run for each of the four designs D1, D2a, D2b, and D3
spanning the time-frame 2024-2038. CGT-Plan identified
investments in two year increments to minimize NPV of
investments plus operational costs occurring during the 15year decision-horizon. Operations were simulated for every
year using 19 operating conditions. These include five “energy
blocks” each for summer, winter, and shoulder to capture fivetime periods in each season: 1–7 a.m., 8 a.m.–12 p.m., 1–4
p.m., 5–6 p.m., and 7 p.m.–12 a.m. The remaining four
conditions were “peak blocks” to capture one-hour annual
peak conditions in each of four regions: West, Northwest,
Midwest, and East. The peak blocks were used to model
annual planning reserve requirements, and because different
regions peak at different times of the year, this enabled study
of inter-regional reserve-sharing subject to transmissionrelated deliverability constraints.
Distributed solar-PV was modeled to increase at rate of 3%
per year. Decision variables included investment options
among generation resources and transmission technologies.
The four scenarios were studied under two sets of economic
and policy conditions: current U.S. policy; and a carbon
policy. The current policies case enforces existing renewable
portfolio standards, while the other case uses a carbon price of
$3/tonne/year that grew to $45/tonne in 2038 as a proxy for
anticipated growth in renewables like wind and solar.
Generation resources were selected from among all typical
resource technologies and were modeled with appropriate
maturation rates at all buses, and in any amounts. Natural gas
price assumptions were adopted from the U.S. Energy
Information Agency’s 2017 Annual Energy Outlook, the
nominal price was $_____. Battery energy storage was not an
investment option. At each bus, the wind resources available
for selection included three 100-meter wind technologies, each

having different costs and the ability to be optimized for
unique wind resource characteristics by geography. This
included three different capacity factor categories that
identified the gigawatt (GW) investment potential at a
particular range of capacity factor. Solar investments were
limited to utility-scale and were split evenly between singleaxis tracking and fixed-tilt.
Investment options among transmission technologies
included additional AC capacity on any existing branch at the
voltage of that branch, at a cost per mile appropriate for that
voltage. In D2a, HVDC expansion was limited to only the
seven existing back-to-back ties in the U.S., and these ties
could expand independent of one another. This was also true
for D2b, but in this case, there were three additional HVDC
lines connecting EI and WI that could expand under the
constraint that all three lines maintained equal capacity. In D3,
the B2B ties were not allowed to expand, and all segments of
the macrogrid were able to expand only under the constraint
that they maintained equal capacity. Although the N-1
reliability criterion was not explicitly imposed, the “equal
capacity” constraints for the HVDC lines in D2b and D3 were
employed as rough proxies to avoid significant violation of
this criterion. “Base” costs for all generation resources and AC
and HVDC transmission were adjusted for location using
regional multipliers. Based on analysis of discount rates
recommended by the White House Office of Management and
Budget and other studies, we chose a nominal discount rate of
7.7% and an inflation rate of 2%, resulting in a real discount
rate of 5.7%. Demand growth was set within each region
consistent with recent studies [3, 4]; technology costs and
regional multipliers were based on [5, 6]. Other data and
associated sources are identified in [7, 2].
The capacity expansion model developed aggregated zonal
transmission and generation for the EI and WI. In order to
study the operation of these systems and determine operational
savings (in perpetuity) of the HVDC facilities, a full nodal
model of the interconnections was necessary. CGT-Plan
utilizes a 169-bus zonal model, while the PCM has over
98,000-buses. This requires a translation between the CGTPlan results (generation and transmission investments and
generation retirements) and the PCM network. The translation
is a two-step process, beginning with a 2024 nodal model.
Step 1 distributes generation investments and retirements
identified by CGT-Plan according to the following criteria:
 Generation is retired in the PCM model based on heat
rate until the CGT-Plan retirement amount is
satisfied;
 CGT-Plan natural gas expansion is added at locations
in the PCM model where coal plants were retired, up
to capacity of the coal plant, and then to buses with
existing natural gas, in proportion to the natural gas
plant’s existing capacity.
 Wind and PV investments identified by CGT-Plan
were added to the high voltage node (500, 345, or
230 kV) in the PCM model that is geographically
closest to the investment location.

DRAFT-for review purposes only. Send comments to aaron.bloom@nrel.gov

 3

Step 1 resulted in a PCM nodal model that contained 2038
load and generation but 2024 transmission. Step 2 expanded
the transmission in the PCM model. Because the CGT-Plan
zonal model represented the PCM transmission network in a
highly aggregated form, there was no appropriate way to map
the CGT-Plan transmission investments directly to the PCM
model. Instead, we developed a transmission expansion
planning (TEP) optimization program and applied it to the
PCM nodal model obtained from Step 1. Although the PCM
network was very large, our TEP application was fast because
(a) we applied it as a single period optimization, i.e., to just
one year (2038); (b) it was run separately for the EI and for the
Western Interconnection. However, because each transmission
investment changes both the circuit capacity and the circuit
reactance, the problem is nonlinear. To address this, we
developed the TEP as a sequence of linear programs (LPs),
where each LP minimized the total transmission investment
cost (subject to DC power flow equations) where only circuit
capacity was treated as a decision variable while circuit
reactance was held constant. Following the LP solution, the
reactance of each invested circuit was updated to reflect the
change in capacity, after which the LP was rerun. The
iterations were terminated when the circuit with the largest
change in capacity relative to the previous iteration was within
a specified tolerance.
The nodal PCM that resulted from the capacity expansion
scenarios was used to simulate 8,760 hours of continuous
operation in 2024 and 2038. The PCM conducted a unit
commitment and economic dispatch for both interconnections,
simultaneously. This created significant computational
challenges for the study. To address solve time constraints we
adopted a new decomposition method developed by [8]. This
method enables the unit commitment and dispatch to be
distributed across each simulation region (independent system
operator [ISO]/regional transmission organization [RTO]
equivalent). This enables a more realistic representation of the
operations and economics of multiple operating regions and
significantly reduces simulation solve time. This method
enabled the simulations to be completed in less than two days,
previous methods would have taken upwards of two months in
a high-performance computing environment.
The 2038 PCM includes approximately 13,000 generating
units, 98,000 transmission nodes, and 96,000 transmission
lines and transformers. The in the carbon policy case, the PCM
includes a $45/tonne cost for carbon to maintain consistency
with the capacity expansion model. Time series data for wind,
solar, and load were a critical component of the PCM exercise.
Wind data is from the Wind Integration National Dataset
(WIND) Toolkit. Solar data is from the National Solar
Radiation Database (NSRDB). Load data is from multiple
sources including the various regional transmission
organizations and independent system operators and FERC
[1]. Weather conditions for the years 2007-2013 were
evaluated for inclusion in the PCM. A geospatial analysis of
the wind and solar resource availability was conducted and
identified 2012 as the most average across the seven year data
set.

Thermal plant assumptions were adopted from [9] and [10]
and enabled detailed modeling of every thermal generator in
the EI and WI. In nearly all cases, existing thermal plants that
are still in operation in 2038 have unit specific plant flexibility
characteristics that were extracted by analyzing the
Environmental Protection Agency’s Continuous Emissions
Monitoring System. When unit specific data was unavailable,
generic assumptions were made based on the vintage and type
of generator.
The initial AC power flow study complements the previous
capacity expansion and PCMs by introducing steady-state and
dynamic analysis. Utilizing the same base case network
assumptions, a solved PSS®E case was created to model the
combined EI and WI case. The B2B HVDC connections were
studied and modeled in detail to enable the transfer of power
across the interconnections and assess any violations of flow,
voltage and contingencies.
A variety of model conflicts, e.g. renumber the buses, some
areas, zones, were corrected to combine the EI and WI. The
next step was to combine the EI and WI 2024/2026 models
using the PSS®E model modification functionality. The third
step was to identify the points of interconnection and
schedules of the B2B HVDC systems in the WI 2024/2026
models. Finally, we coordinated the power transfer schedule
megawatts and direction of B2B HVDC systems.
A new combined power flow case was developed by
importing PCM simulation results (generation, dispatch, load,
and HVDC schedules) for a certain time stamp (July 17 at 2
a.m.) into the original power flow case. This process is
explained in Figure 1. Future work aims to extend the results
from the 2038 PCMs to the combined AC power flow model.

Figure 1: Steps for Importing and Validation of PCM data into AC
power flow model

IV. RESULTS
In this section we detail the results of the capacity
expansion and PCM. We also describe the initial steps taken to
conduct future AC power flow analysis.
A. Capacity Expansion Modeling
All transmission designs, were limited to 600 GW capacity
investments in order to avoid unrealistic installation rates. The
results achieved a 2038 renewable (wind, solar and hydro)
penetration of 50% by energy, with a reduction of CO 2
production to 30% of its 2024 level. The presence of a carbon
tax, lower investment and operational costs of wind and solar

DRAFT-for review purposes only. Send comments to aaron.bloom@nrel.gov

 4

compared to other technologies resulted in these comprising
about 93% of the invested generation in all designs, with the
remainder comprised of gas-fueled combined cycle (~21%),
combustion turbines (~2%) and distributed generation (yy%).
Wind generation investments were mainly located in the
Midwest, and solar generation investments were primarily
located in the East and the South. Because D1 was the only
design that did not allow cross-seam transmission investment,
it was the benchmark, and results of the other three designs are
given relative to it. Table 1 summarizes the main metrics for
the three designs relative to D1. In addition to the base
designs, which include a carbon tax and do not include any
state RPS in effect for 2024-2038, three sensitivities
The most important observation from Table 1 is that the
benefit-to-cost (B/C) ratio, calculated as the change (relative
to D1) in the generation investment and operational cost
divided by the change in the transmission investment cost, is
well above the industry threshold of 1.25 considered necessary
to justify transmission investments. Although some benefit is
obtained through reduction in generation investment costs,
most of it occurs as a result of reduction in generation
operational costs. The values shown may be considered as
lower bounds on B/C ratios for two reasons. First, they reflect
economics from only 2024 to 2038; benefits continue to
accrue on an annual basis as indicated by the perpetuity costs.
Second, they do not reflect non-quantified benefits (NQBs)
such as the ability of the electric system to continue supplying
low-cost energy while adjusting and adapting to policy
changes and/or catastrophes such as large hurricanes and
widespread wildfires.

that results from the initial generation fleet and subsequent 15
years of investments and retirements. This reflects that crossseam DC transmission investments made in D2a, D2b, and D3
enable increased capacity sharing between regions. Second,
we observe that the B2B investments of D2a require more AC
transmission than in D1 (to move power to the coasts) and that
the DC line investments of D2b and D3 require less AC
transmission.
Figure 2 illustrates generation, line, and B2B investments
for D1 and D2b where, for generation, the area within the
bubbles is proportional to the generation capacity invested. All
black lines in both designs represent AC transmission
investments. The red circles for D2b represent DC B2B
transmission investments, with amounts given in the table
such that the top-to-bottom ordering for the first seven rows
correspond to the north-to-south circles. The next-to-last row
in the table for D2b indicates the capacity of each of the three
DC lines, represented in red in the figure. The last row of the
table gives the total cross-seam transmission capacity for D2b.
Comparison of these two figures shows that cross-seam
transmission tends to shift wind capacity build out from west
to east and solar capacity build out from east to west. This is
also observable in D2a and D3.

D1

Table 1: Summary of D2a, D2b, D3 metrics relative to D1
Capacity or cost item
AC transmission
invested, GW
HVDC transmission
invested, GW
2038 creditable
capacity, GW
Transmission
investment cost, $B
Gen investment cost,
$B

D1
228.9

ΔD2a
22.4

ΔD2b
-5.9

ΔD3
-33.8

0

25.7

28.4

125.8

838.5

-29.0

-46.5

-44.4

61.2

12.7

13.67

18.9

704.0

-0.7

-47.8
-3.5

Carbon Cost, $B
Operational cost, $B*
15-yr B/C ratio
Perpetuity cost
(annualized 20 yrs),
$B

171.1
1475.9
N/A
N/A

-1.66
-30.9
2.48
-1.37

45.16
-7.0
-3.33
-38.2
3.30
-2.51

D2b

Figure 2: Generation/transmission investments for D1, D2b

-6.66
-44.3
2.52
-4.19

Table 1 reveals two additional insights. First, D2a, D2b, and
D3 show a decrease in creditable capacity (the summation
over all available generation of each unit’s capacity multiplied
by its capacity credit, where capacity credit is percent capacity
applied towards satisfying the annual peak [11, 12], an effect

B. PCM
The zonal results from the final year of CGT-Plan were
translated into the nodal PCM model using the methods
described earlier. The translation process adds or retires
specific generators and the TEP model (see TEP description in
Section III) translates the AC transmission added to the CGTPlan designs to the nodal network. Figure 3 shows the results
of the TEP AC transmission translation and the nodal
connections of the HVDC network.

DRAFT-for review purposes only. Send comments to aaron.bloom@nrel.gov

 6

Figure 3: Additions to the existing transmission topology
represented in the 2038 PCM modeling

We ran the PCM at an hourly resolution for a full year
representing 2038. Figure 4 shows the similarity of all designs
on an aggregate basis. D1 has more thermal generation due to
its larger installed capacity share in D1. In these four futures,
38-39% of all load is met by wind and PV. and 50% is meet by
all forms of renewable energy (i.e. wind, solar, hydro,
biomass, and geothermal).

Figure 4: Total annual generation by type for all four designs

Figure 5 shows the amount of reserves each generator
type provided. Curtailed wind and natural gas combined
cycles provided the majority of reserves throughout the year.

Figure 5: Reserves provided by each generator type

CGT-Plan suggested annualized operational savings of the
system designs in 2038 to range from $1.3 to $4.2 billion
(Table 1). The PCM generally agrees with these operational
savings, finding that D2a, D2b, and D3 save between $2.0 and
$3.9 billion when compared to the production costs of D1.
However, the operational savings do not directly match
between scenarios. This is because of differences observed in
wind and solar curtailment between the CGT-Plan and PCM.
The production cost savings seen in PCM results validate the
estimated savings from CGT-Plan results, even though CGTPlan relied on a much less detailed representation of system
operations showing robustness of the approach.
CGT-Plan also assessed the value of peak load diversity
between different regions of the country and found that cross

DRAFT-for review purposes only. Send comments to aaron.bloom@nrel.gov

 7

seam transmission can enable greater capacity sharing
between regions. However, CGT-Plan does not have a full
picture of annual operations and requires the PCM to ensure
the reliable balancing of load and generation at all hours of the
year with reduced creditable capacity. This is especially
important given uncertainty and variability of operating
systems near 40% wind and solar penetrations. The PCM
gives confidence that the designs created by CGT-Plan can
handle the uncertainty and variability of these high variable
generation futures. The PCM found that for all hours of the
year, all designs balanced load and generation and served
between 99.69% and 99.98% of all spinning reserve
requirements. D1, the design with the least cross-seam
transmission, had the largest total reserve shortage.
Not only do the PCM results validate the balancing
reliability of the designs created by CGT-Plan, but they also
show how cross seam transmission can take advantage of load
diversity throughout the year, which allow D2a, D2b, and D3
operate with less creditable capacity. Figure 6 shows
operation of the aggregate cross seam transmission during the
three-day period around coincident peak load. All designs
show a diurnal pattern to the operation of the cross seam
HVDC and B2Bs. At noon2, load is nearing peak in the EI, but
load remains lower in the WI while PV in the WI is nearing its
peak output for the day. The cross-seam transmission is
heavily loaded, delivering power from west to east. As the day
progresses and the sun sets on the west coast, the cross-seam
transmission switches direction, operating at near 100%
loading from east to west as the WI reaches peak net-load, or
the “head of the duck”.

in the EI. An increase in thermal or hydro generation generally
indicates greater utilization of an existing resource in D1,
while an increase in wind or PV generation indicates the
utilization of a resource that did not exist in D1. A decrease in
thermal generation represents generation from a unit that was
retired in designs 2a, 2b, and 3, but existed in D1. A decrease
in wind or PV generation relative to D1 indicates that the
resource was moved to higher quality locations in designs 2a,
2b, and 3 (i.e. PV moved to the WI, wind moved to the EI).
Figure 7 demonstrates how cross seam transmission
delivers low cost PV and thermal resources from the WI to the
EI as the load in the EI nears peak at noon. Local expensive
thermal resources in the EI are no longer needed to meet peak
load, as transmission enables more efficient use of lower cost
resources and reduced creditable capacity overall.

Figure 7: Change in generation relative to D1 in SPP/MISO and the
WI on August 7 12:00 p.m.

As the day progresses, flow on the cross seam transmission
changes direction and begins exporting power from the EI to
the WI. Figure 8 shows the change in generaiton on August 7th
at 11:00 pm, when cross-seam tranmsision has changed to be
loaded at nearly 100% from east to west. Units that were
already online to help meet peak in the EI, such as combustion
turbines, remain online to meet the net-load peak in the WI,
avoiding the need to startup less efficient units in the WI. Low
cost EI wind is also able to be more fully utilizted, rather than
building lower capacity factor wind in the WI.

Figure 6: Cross seam power flow during peak period. A positive flow
is an export from the EI to the WI.

Figure 7 and 8 show the change in generation by type
between D1 and the designs with more cross seam
transmission at noon and 10:00 pm respectively on August 7th.
In each figure, one panel shows the change in generation in
the WI along with the net cross seam flow from the WI. The
other panel shows the same for just Southwest Power Pool
(SPP) and Midcontinent Independent System Operator
(MISO), where cross seam transmission has the largest impact
2 All times are Eastern Standard Time

DRAFT-for review purposes only. Send comments to aaron.bloom@nrel.gov

 8

Figure 8: Change in generation relative to D1 in SPP/MISO and the
WI on August 7 11:00 p.m.

Finally, the PCM results allow us to understand the
utilization of cross seam transmission flexibility during
extreme events, such as wind/PV forecast errors or large
ramping events. Figures 9–11 show the change in cross seam
flow and the change in generation in SPP/MISO and the WI
during a period of large instantaneous wind and PV
penetrations, followed by a large ramp down in both resources
at sunset on April 16th.

Figure 10: Change in generation relative to D1 in SPP/MISO and the
WI on April 16 6:00 a.m.

As the day progresses, there is a large ramp down in wind
throughout SPP and MISO. The increased cross-seam
transmission provides flexibility to mitigate the lost wind
output in SPP and MISO. The direction of the flow changes,
reaching peak exports from the WI at 7:00 pm, which
coincides with the April 16th net-load peak in the Eastern
Interconnection. Figure 11 shows how cross seam
transmission enables greater utilization of WI thermal and
hydro resources. This flexibility allows the Eastern
Interconnection to balance the system reliably with less
creditable capacity even during periods of large ramp downs
in wind and solar.

Figure 9: Cross seam power flow during a period of high variable
generation. A positive flow is an export from the EI to the WI.

The beginning of the day on April 16th has very high wind
penetrations, particularly in SPP and MISO. The cross-seam
transmission is heavily loaded, delivering the wind to the WI.
Figure 9 shows the large exports from the EI to the WI in the
early morning hours of April 16th. Figure 10 shows the change
in generation dispatch on both sides of the seam at 6 am.
Utilization of the high quality SPP/MISO wind increases due
to the presence of cross-seam transmission.

Figure 11: Change in generation relative to D1 in SPP/MISO and the
WI on April 16 at 7:00 p.m.

C. AC Power Flow Modeling
The combined 2024/26 power flow model was developed
using PCM data for a single instant in time. To validate the
data import process we compared the power flow results from
the DC power flow based PCM and the AC power flow based
model for flows of 230 kV lines and above. In general, the
Eastern Interconnection matches between PCM flows and AC
power flow model calculated flows. A comparison between
flows in PLEXOS and PSS®E for 500kV lines is shown in
Figure 12. For the WI, the line flows matches were not as

DRAFT-for review purposes only. Send comments to aaron.bloom@nrel.gov

 9

consistent. That could indicate that the PCM case is based on a
power flow case that is slightly different than the case used in
the power flow analysis.

Figure 12: 500kV lines PLEXOS vs PSS®E comparison
In this study, N-1 includes a single branch/transformer
connecting between buses of 230 kV and above, and single-tie
lines between different areas including DC tie lines and single
generators. For contingency impact violations, Python scripts
were prepared to extract the following violations: (1) voltage
violations of buses of 230 kV and above and (2) flow
violations for single branch/transformer connections between
buses of 230 kV and above. The number of contingences with
unsolvable power flow is recorded. The criteria used to
identify violations are: (1) voltage outside the boundaries of
0.9 pu to 1.1 pu (230kV to 416kV lines) and 0.95 pu to 1.15
pu (500kV to 765kV lines) and (2) flow limits of 130% of
branch Rate A. The number of voltage, flow violations, and
diverged cases in the N-1 contingency analysis in the WI
system and EI systems are shown in Tables 2 and 3.
Table 2. WI N-1 contingency analysis results in the combined
2025/26 case with PCM data

The significant value was demonstrated through the multimodel approach. This study demonstrated that the system was
able to balance generation and load at lower creditable
capacity, enable load diversity, and increase operating
flexibility.
This study examined the benefits of increasing transmission
pathways over four distinct scenarios across the country.
While fundamental elements of transmission and generation
were represented throughout the study, additional modeling
and analysis is required to further examine the best grid
designs to bring the anticipated technical and economic
benefits to the system. Industry review and input will remain
vital to plans to expand transmission capability across the
interconnections, as studies often present the most optimal
solution and sometimes lack consideration for market adoption
feasibility.
Full exploration of the potential benefits of cross-seam
transmission to the continent will require continued multimodel study, where only a suite of tools can examine multidimensional future planning needs. These analyses include
expanded AC power flow studies of steady state and stability
modeling; system and local-level benefits (electric production,
environment, jobs, etc.); grid reliability needs; gas-electric
coordination, and resilience. Additional capacity expansion
and PCM scenarios would be required to further understand
the uncertainty and impacts on resilience caused by weather,
fuel costs, and capital costs.
VI. ACKNOWLEDGMENT
The authors gratefully acknowledge the contributions of
Charlton Clark, Jian Fu, and Kevin Lynn from the Office of
Energy Efficiency and Renewable Energy, and Kerry Cheung
from the Office of Electricity Delivery and Energy Reliability.

VII. REFERENCES
[1]

Table 3. N-0 and N-1 contingency analysis results in Eastern
Interconnection combined 2025/26 case with PCM
data
Total number of contingencies
Number of solvable contingencies
Number of contingencies with islands
Number of unsolvable contingencies
Number of N-0 flow violations
Number of N-1 flow violations
Number of N-0 voltage violations
Number of N-1 voltage violations

1076
731
75
270
41
52
641
153

V. CONCLUSIONS/NEXT STEPS
Combining CGT-Plan, PCM, and AC power flow analysis
allows a thorough assessment and validation of the value of
increasing cross seam transmission. Benefit cost ratios reached
as high as 3.3 and annual operational savings exceeding $2
billion.

[2]

[3]
[4]
[5]
[6]

[7]

Novacheck, J., Bloom, A., Lau, J. “Production Cost Modeling for
the Interconnections Seam Study” NREL Technical Report,
Forthcoming.
Armando Luis Figueroa-Acevedo, "Opportunities and benefits for
increasing transmission capacity between the US eastern and
western interconnections" (2017). Graduate Theses and
Dissertations. 16128. https://lib.dr.iastate.edu/etd/16128
Eastern Interconnection Planning Collaborative (EIPC) (2011),
“Phase 1 Report: Formation of Stakeholder Process, Regional Plan
Integration and Macroeconomic Analysis.”
E3 (2011), “CA LSE and WECC load and energy forecasts.”,
GHG/11 Load_Growth_Forecasts v2.doc. [Accessed November
2016].
The National Renewable Energy Laboratory (NREL), 2016. “2016
Annual Technology Baseline.” National Renewable Energy
Laboratory. http://www.nrel.gov/analysis/data_tech_baseline.html.
R. Pletka, J. Khangura, A. Rawlins, E. Waldren, and D. Wilson,
“Capital Costs for Transmission and Substations - Updated
Recommendations for WECC Transmission Expansion Planning,”
B&V Project No. 181374, prepared for the Western Electric
Coordinating Council (WECC), February, 2014, available at
www.wecc.biz/Reliability/2014_TEPPC_Transmission_CapCost_
Report_B+V.pdf.
A. Figueroa-Acevedo, A. Jahanbani Arkadani, H. Nosair, A.

DRAFT-for review purposes only. Send comments to aaron.bloom@nrel.gov

 10
Venkatraman, J. Novacheck, and J. McCalley, “Benefits of
increasing transmission capacity between the US Eastern and
Western Interconnections,” under review by IEEE Transactions on
Power Systems.
[8] Clayton Barrows, Brendan McBennett, Josh Novacheck, Devon
Sigler, Jessica Lau, Aaron Bloom. Under review. 2018.
Decomposing Electricity System Models to Represent Multiple
System Operators.
[9] Aaron Bloom, Aaron Townsend, David Palchak, Joshua
Novacheck, Jack King, Clayton Barrows, Eduardo Ibanez,
Matthew O'Connell, Gary Jordan, Billy Roberts, Caroline Draxl,
Kenny Gruchalla. 2016. Eastern Renewable Generation Integration
Study.
NREL/TP-6A20-64472.
https://www.nrel.gov/docs/fy16osti/64472.pdf.
[10] Michael Rossol, Gregory Brinkman, Joshua Novacheck, Paul
Denholm, Aaron Bloom. Under Review. 2018. A National Analysis
of Thermal Plant Flexibility.
[11] IEEE PES Task Force on the Capacity Value of Wind Power,
“Capacity Value of Wind Power,” IEEE Trans. On Power Systems,
vol. 26, pp. 564-572, 2011.
[12] IEEE PES Task Force on Capacity Value of Solar Power, C. Dent
(chair), “Capacity value of Solar Power,” International Conference
on Probabilistic Methods Applied to Power Systems, 2016.

VIII. BIOGRAPHIES

DRAFT-for review purposes only. Send comments to aaron.bloom@nrel.gov