Article

Statistical Steady-State Stability Analysis for

Transmission System Planning for Offshore Wind
Power Plant Integration
Amirhossein Sajadi,1 , Kara Clark 2 and Kenneth Loparo 1 *
1 Department of Electrical Engineering and Computer Science, Case Western Reserve University, Cleveland,
OH, USA, email: axs1026@case.edu, kal4@case.edu
2 Transmission Grid Integration, National Renewable Energy Laboratory (NREL), Golden, CO, USA; email:
Kara.Clark@nrel.gov
* Correspondence: kal4@case.edu

Version July 10, 2020 submitted to Journal Not Specified

1 Abstract: This paper presents a statistical steady-state stability analysis for transmission system
2 planning studies to identify operational issues for integration of offshore wind power plants. It
3 includes normal and contingency operation. This study considers the integration of a 1,000-MW
4 offshore wind power plant into the FirstEnergy/PJM service territory in the U.S. Great Lakes region
5 as a case study and uses a realistic computer model of the U.S. Eastern Interconnection, a 63,000-bus
6 test system. The results show the utility of the statistical analysis used here and effectively identify
7 the operational impacts in this region as a result of the integration the offshore power plant.

8 Keywords: offshore wind, power system planning, steady-state stability, statistical analysis, wind
9 integration.

10 1. Introduction
11 Integration of offshore wind power plants into an existing power grid can impose operational
12 challenges for the grid. This is because of the topological and technological changes that occur as a
13 result of the wind integration. During the wind integration planning process, the steady-state and
14 dynamic analyses are conducted for the power system of interest to ensure that the after the wind
15 integration, the system retains the measures of stability and is able to continue its normal operation.
16 Steady-state analysis of electric power systems refers to the study of the behavior of the system
17 while operating at any given equilibrium point [1]. The state-state analysis of power systems should
18 be assessed under two operational conditions of (1) Normal operation and (2) Contingency operation.
19 Given the uncertain and variable nature of wind, injecting electrical power generated from wind into
20 the transmission system can introduce additional complexities into transmission system operation and
21 stability. This paper introduces a statistical steady-state stability analysis tool that is practical for the
22 integration of offshore wind power plants.
23 The first objective of the current work is to provide a statistical steady-state analysis for large-scale
24 power systems under the normal operation conditions to identify the effects of offshore wind power
25 plants on transmission system operation. A number of studies have investigated the effects of wind
26 power plant integration on power system operation, and some of them have recommended different
27 analysis techniques. The authors of [2] introduced an error-based index to measure voltage regulation
28 performance with implications of how different types of wind generators in offshore wind power
29 plants impact voltage regulation in distribution feeders when integrating a wind power plant at the
30 transmission level. This study suggested that voltage regulation in distribution feeders is a function
31 of the geographical location of the point of interconnection (POI). The authors of [3] investigated

Submitted to Journal Not Specified, pages 1 – 24 www.mdpi.com/journal/notspecified

 Version July 10, 2020 submitted to Journal Not Specified 2 of 24

32 the steady-state stability of power systems with high penetrations of doubly-fed induction generator
33 (DFIG)-based wind generators using P − V curves. This study suggested that focusing on traditional
34 worst-case operating points does not necessarily capture the true worst-case scenario in systems that
35 have integrated wind power plants and that the interconnection of DFIGs at both the transmission
36 and distribution levels has a positive impact on the voltage control of the lines. This work used the
37 2013 Ireland power system as a case study and incorporated historical data, power flow, and economic
38 dispatch. The authors of [4] introduced a tool to assess stability impacts of land-based wind generators
39 on weak grids. They investigated the impact of land-based wind generation on a sub-transmission
40 network by using a combination of sensitivity analysis and continuous power flow to look into the
41 sensitivity of the level of power generation by wind machines and the operation of tap changers. They
42 also examined the effects of wind machines connected to the sub-transmission lines on transformers
43 that connect sub-transmission lines to transmission lines. This work used models of a regional power
44 system in Michigan and considered two wind power plants in the area. The authors of [5] carried out
45 a P − V analysis on the California Independent System Operator and proposed a method to handle
46 power systems including renewable energy sources; however, this method was not implemented or
47 verified. The majority of the above-mentioned efforts have investigated land-based wind generation
48 on relatively small test systems by using simplified models. In addition, interconnection points on
49 distribution and sub-transmission networks were a major focus of these studies.
50 The second objective of this paper is to introduce a statistical tools for the contingency operation
51 assessment of large-scale power systems that include offshore wind power plant integration. Various
52 contingency ranking and monitoring methods have been developed for traditional power systems.
53 In [6], five continuous power flow methods including the multiple load flow method, test function
54 method, Q − V curve fitting and generalized curve fitting methods were implemented on a 234-bus
55 power system and compared. The multiple load flow method, proposed in [7], approximates the
56 security margin of the system rather than computing the actual bifurcation point using the voltage
57 gradient on at any given point on the stable branch of the P − V curve. The test function method,
58 introduced in [8], uses a quadratic approximation of the load parameter to determine the voltage
59 margin following a small perturbation to the load. The Q − V curve fitting method [9] uses three
60 point on the stable branch of the Q − V curve to fit identify the bifurcation point. This method is not
61 capable of handling active and reactive power changes when the load in multiple branches change,
62 simultaneously. The generalized curve fitting method approximates the stable branch at the nose
63 of the V − λ curve by a second order polynomial. This allows computing the active and reactive
64 power change, simultaneously. In [10], a method for AC networks was proposed that includes the
65 effect of voltage regulators and generators using a voltage sensitivity index with application to the
66 IEEE 24-bus Reliability Test System [11]. In [12], a method to identify outages that lead to operating
67 point infeasability using a nonlinear optimization approach was proposed. Such a condition might
68 occur if minimum values of constraints are greater than operator-defined thresholds such as bus
69 voltage and load shedding thresholds. This method was implemented on the IEEE-30 bus test system
70 with 6 generators. In [13], several contingency ranking and monitoring techniques were surveyed
71 and discussed. In [14], the authors introduced a fast method for on-line voltage security assessment
72 following large numbers of line outages to identify the worst case using the relationship between a
73 node voltage and the availability of reactive power in the local neighborhood. Reliability indices of
74 Loss-of-Load Probability (LOLP) and Expected Energy Not Served (EENS) were used in contingency
75 ranking in IEEE 24-bus Reliability Test System, in [15]. This work suggests that the reliability indices
76 may not predict the real severity of the system contingencies. The study presented in [16], outlines
77 the computation of linear and quadratic estimates of the loading margin to changes in parameters
78 for a given power system such as reactive power support, generation redispatch, and load changes
79 using the IEEE 118-bus test system. The focus, however, was on changing a single parameter at a time
80 rather than multiple parameters in multiple nodes. In [17], the genetic algorithm was used to rank
81 contingency severity with a focuses on a small number of contingency events in relatively small test
 Version July 10, 2020 submitted to Journal Not Specified 3 of 24

82 systems, 30-bus and 57-bus. In [5], the performance of security assessment platforms in California
83 Independent System Operator (CAISO) was presented. This work very briefly mentions renewable
84 energy handling, but no actual performance was presented. A risk-based method to assess security
85 of power systems based on sensitivity of load margin to the bifurcation point using Poison model of
86 system uncertainties was presented in [18]. This work assumes the risk calculation is being done over
87 short term time horizon, and consequently, changes in operating conditions are small. However, in
88 power system planning studies, this method would not be reliable since long term changes should
89 be considered. The above-mentioned studies all used a small scale power system for verification of
90 methods, and all of them involved dispatchable power generation units. A relevant tool for voltage
91 stability assessment for power systems with stochastic renewable energy units was presented in [19].
92 This tool computed the classic P − V curves for every level of penetration of wind power with testing
93 on a 23-bus system in MATPOWER and does not provide a ranking approach for all levels of wind
94 power.
95 This paper, by considering the findings of above-mentioned works, first provides a statistical
96 approach to analyze the data from a large power system in which hundreds to thousands of nodes
97 must be monitored and analyzed to find meaningful results by quantifying the associated uncertainty.
98 Second, it introduces a risk-based approach for quantifying the impacts of wind power variability in a
99 single index that enables comparing and ranking the severity of contingency events to identify the most
100 secure integration scenario. This research is intended to develop insight into the impact of large-scale
101 offshore wind generation on steady-state operation assuming that the existing operating point is stable.
102 The contribution of this paper is to provide a statistical approach for different wind generation and
103 integration scenarios that might affect the large-scale power system’s performance under normal and
104 contingency operation conditions. This is a major area of concern in the wind integration projects. The
105 advantage of this methodology is its ease in implementation with existing practical power systems
106 models which will inherently include the effects of continuous and discontinuous system’s elements.
107 Additionally, it is easy and efficient to implement.
108 This work uses a simulation model of the U.S. Eastern Interconnection as the test system and
109 focuses on the integration of a 1,000-MW offshore wind power plant operating in Lake Erie into
110 the FirstEnergy/PJM service territory as a case study. After wind availability estimation, potential
111 geographical locations of the offshore plant POIs were identified and, accordingly, integration scenarios
112 were developed using using 63,000-bus system model, which relies on system models that are used by
113 industry. Databases were developed based on the previous work performed by GE Energy Consulting
114 and the National Renewable Energy Laboratory (NREL) for the 2013 Eastern Frequency Response
115 Study [20]; however, the previous databases were slightly modified to represent the current available
116 generation in the FirstEnergy system due to recent retirements. The wind turbines were modeled
117 based on the GE 3.6-MW commercial wind machines [21]. The export cables that transmit offshore
118 power to land-based substations were designed and modeled based on data available on the ABB
119 High-Density Crosslinked Polyethylene (XLPE) commercial sea cables [22].

120 2. Mathematical Formulation

121 Wind integration projects typically involve adding new wind generation units to an existing
122 power system. Thus, a wind integration study could be treated as an expansion study. However, due
123 to the highly variable nature of wind generation, it is necessary to conduct stability and reliability
124 studies for all levels of contribution of the wind power plant into the power system, from 0% to 100%
125 of its capability. If the appropriate levels of the stability measures are not reached, additional power
126 injected into the area may be necessary to improve reliability [23].
127 The steady-state analysis under the normal operational conditions refers to the system’s
128 performance when all components are intact and in operation. This analysis is associated with
129 two criteria: (1) line thermal rating and (2) voltage regulation at both ends of the lines [24]. Typically,
 Version July 10, 2020 submitted to Journal Not Specified 4 of 24

130 the voltage stability margin is much smaller than the thermal stability margin, and therefore the
131 voltage stability margin determines the overall stability margin for the system [24].
132 The steady-state analysis under the contingency operational conditions is the assessment of
133 the system’s performance to maintain normal operation and deliver power to consumers without
134 interruption following disturbances referred to as contingencies. A contingency represents a credible
135 event that can occur in the system such as an outage of a main component or multiple components
136 [1]. The results represent the response of the system to those contingencies, and the events are ranked
137 based on their severity to identify the worse case scenarios [25]. This process is so-called contingency
138 ranking and monitoring. The most common approach in power system planning studies considers N-1
139 security criteria in which a system should curtail outage of any single contingency event, transmission
140 line or generator outage, while remaining in the normal or alert operating state [12]. In practice, one
141 way of quickly computing any possible violation of operating limits during contingency events is
142 to start with a base case and then continue to apply contingency events and calculate their severity
143 indices in order to evaluate the system’s strength [25].
144 Bearing the above-mentioned issues in mind, this research introduces a statistical methodology
145 to assess the steady-state stability of large-scale power systems for the integration of offshore wind
146 power plants. The first step is to initiate the base case with no contingencies to ensure the system
147 with the given loading scenario is stable and power flow converges. This serves as the steady-state
148 normal operation scenario. The next step is to evaluate the impact of contingency events, one at a time,
149 and conduct analysis on voltage magnitudes of the busbars, power flow of the lines, and active and
150 reactive power of the generation units.

151 2.1. Power Flow Equations with the Wind Model

152 This section, introductions the power flow equations that includes the variability of the wind
153 power. The behavior of a power system, using differential-algebraic equations, is given by:

ẋ = f ( x, u) (1)

where x is a n×1 state vector and u is a k×1 representing inputs to the system. For every equilibrium
point of x ∗ , the steady-state solution of the system is given by:

f ( x∗ , u) = 0 (2)

To simulate the variability from wind power generation, a variability parameter λ could be
defined as:

0 ≤ λ ≤ λmax (3)

154 where λ = 0 represents maximum wind power generation and λ = λmax represents no wind power
155 generation.
This study relies on the LOLP reliability metric which uses only the peak values of daily load
profiles [23]. Therefore, constant peak load data for each operating scenario is considered. Hence,
the power equations for a power system including synchronous and wind power generator could be
written as:

PGi (λ) = PGi (1 + λK PG ) (4)

0 i

PWi (λ) = PWi (1 + λK PW ) (5)

0 i

156 where PGi and PWi are the active power generation dispatch by conventional and wind power plants
157 with wind power at the maximum active power generation, respectively, at the i-th bus, and K PG and
i
 Version July 10, 2020 submitted to Journal Not Specified 5 of 24

158 K PW are their active power participation factors. The participation factor for generators is positive and
i
159 for the wind generator is negative. In order to investigate the impact of variability of wind power
160 generation on voltage, the real power generation from wind should be shifted in certain steps with λ
161 changed at each step.
While wind power decreases, the other conventional generators attempt to compensate the deficit
of real power that occurred because of the decrease in wind power generation. Consequently, according
to the types of generators and the controllers, the reactive power generation at any given P will be
determined. Reactive power output of a generator is a function of various factors including real power
output of the generator, the technology used in the generator, the capability of the machine to provide
reactive power support, and the controller settings of the machine. Therefore, the reactive power
equations can be defined by:

QGi (λ) = Q( PGi (λ)) · (1 + λKQG ) (6)

0 i

QWi (λ) = Q( PWi (λ)) · (1 + λKQW ) (7)

0 i

where QGi and QWi are the reactive power generation by conventional and wind power plants,
respectively, at the i-th bus. The total reactive generation power in a system with W wind power plants
and G synchronous generators is given by:

G W
QG (λ) = ∑ QGi (λ) + ∑ QWi (λ) (8)
i =1 i =1

162 The available reactive power in the area is an accumulation of compensation devices and reactive
163 power output of individual generators in the system, including the wind machines.
The equations that represent the variation of reactive power corresponding to changes in real
power can be given by:

∂QGi ∂QGi ∂PGi

= · (9)
∂λ ∂PGi ∂λ
∂QWi ∂QWi ∂PWi
= · (10)
∂λ ∂PWi ∂λ
   
∂Q G ∂QW
164 where ∂PG
i
and ∂PW
i
are determined by the P − Q characteristics of the generators in the area.
i i
Rearranging these equations produces the sensitivity of reactive power in the system to variability
of the wind power plant:

G ∂Q W ∂Q
∂QG (λ)
=∑ PGi K PG + ∑
Gi Wi
PWi K PW (11)
∂λ ∂P 0 i ∂P 0 i
i =1 Gi i =1 Wi

The acive power and reactive power sensitivity vectors could be given by:
h iT
∆P(λ) = ∆PGi (λ) ∆PWi (λ) (12)

h iT
∆Q(λ) = ∆QGi (λ) ∆QWi (λ) (13)

165 The voltage vector for conventional generation busbars, subscripted by G, wind power plant
166 busbars, subscripted by W, and load busbars, subscripted by L, can be described by:
 Version July 10, 2020 submitted to Journal Not Specified 6 of 24

 ∂Q G1 ∂Q G1 ∂Q G1 ∂Q G1
 −1
∂VG1 ··· ∂VGG ∂VW1 ··· ∂VWW
∆VG1
   
 . .. .. .. .. ..
 .

 .   . . . . . .
  h ∂Q
G1
i 
 ..   ∂QG ∂Q GG ∂Q GG ∂Q GG 
 ∆ ∂PG1 · K PG · PG1
1 0
   G
··· ···  
..

 ∆VGG   ∂QG1
   ∂V ∂VGG ∂VW1 ∂VWW   
  W ∂QW1 ∂QW1 ∂QW1   . 
 ∆VW1   ∂VG 1
 ··· ···   h
∂Q GG
i
∂VGG ∂VW1 ∂VWW  ∆
 
 h ∂PGG · K PGG · PGG0 i 
1
 ..   .
    
 .  =  .. .. .. .. .. ..  (14)
. . . . .   ∆ ∂QW1 · K
 
  
∆V   ∂QWW ∂QWW ∂QWW ∂QWW 

 ∂PW1 PW1
· PW10


 WW  
  ∂V ··· ∂VGG ∂VW1 ··· ∂VWW 
 
..

 ∆VL1   ∂QGL1
  
∂Q L1 ∂Q L1 ∂Q L1   . 
 .  
 .   ∂VG
1
··· ···   h
∂QWW
i
∆ ∂P
∂VGG ∂VW1 ∂VWW 
 .   1 · K PW · PWW
 .. .. .. .. 

.. .. WW W 0
∆VLn  .

. . . . . 

∂Q Ln ∂Q Ln ∂Q Ln ∂Q Ln
∂VG1 ··· ∂VGG ∂VW1 ··· ∂VWW

167 Power flow across the transmission system is determined by angles difference between both end
168 of lines. The angles in respect to variation of wind power can be described by:

 ∂PG1 ∂PG1 ∂PG1 ∂PG1

 −1
∂θ G1 ··· ∂θ GG ∂θW1 ··· ∂θWW
∆θG1
   
 . .. .. .. .. ..
 .

 .   . . . . . .   h i 
 ..   ∂PG ∆ K PG · PG1

∂PGG ∂PGG ∂PGG 
   G
··· ···   1 0 
 ∆θGG   ∂PG1
   ∂θ ∂θ GG ∂θW1 ∂θWW   .. 
  W ∂PW1 ∂PW1 ∂PW1   . 
 ∆θW1   ∂θG 1
 ··· ∂θ GG ∂θW1 ··· ∂θWW 
  h i
 ∆ K P · PG
 
1
 ..   .
    
 .  =  .. .. .. .. .. ..   h GG G0 
(15)
   . . . . .   ∆ K P · PW 
 i
∆θ   ∂PWW ∂PWW ∂PWW ∂PWW   W1 10 
 WW  
  ∂θ ··· ∂θ GG ∂θW1 ··· ∂θWW 
  .. 
 ∆θ L1   ∂PGL1
  
∂PL1 ∂PL1 ∂PL1   . 
 .   1
··· ···   h i
 .   ∂θG
 .   1
∂θ GG ∂θW1 ∂θWW  ∆ K PW · PWW
 .. .. .. .. 
 W
.. .. 0
∆θ Ln  .
 . . . . . 

∂PLn ∂PLn ∂PLn ∂PLn
∂θ G1 ··· ∂θ GG ∂θW1 ··· ∂θWW

169 Finally, replacing the angle and voltage values into power flow equations produces active power
170 and reactive power flow in the system:

n
P(λ) = ∑ |Vi (λ)||Vj (λ)| Bij sin(θi (λ) − θ j (λ)) (16)
i

n
Q(λ) = − ∑ |Vi (λ)||Vj (λ)| Bij cos(θi (λ) − θ j (λ)) (17)
i

q
S(λ) = P2 ( λ ) + Q2 ( λ ) (18)

171 Having the network data and operational data of components under normal and contingency
172 operations described, the next step is to process the computed data for contingency ranking purposes.
173 Therefore, following metrics described below were used.
 Version July 10, 2020 submitted to Journal Not Specified 7 of 24

174 2.2. Normal Operation Analysis

175 To this end, the mathematical foundation to obtain the system’s variable following the variability
176 of the wind power has been established. In the previous section, it was discussed that the steady-state
177 normal analysis in transmission systems is associated with two criteria: (1) voltage stability and (2)
178 thermal stability [24]. The voltage regulation and active power flow across the system in respect to
179 variability of wind power correspond to these two criteria and are described in the previous section.
180 First criterion is the voltage stability. The voltage stability in transmission level is a function of
181 reactive power. In a transmission line that delivers a constant real power P and supplies the reactive
182 power Q with supply end voltage magnitude of E, delivery end voltage magnitude of V, and reactance
183 of X, the reactive power can be described by [1]:

r
h EV i2 V2
Q (V ) = − P2 − (19)
X X

184 This equation determines the reactive power-voltage characteristic and takes the form of an
185 inverted parabola as shown in Fig. 1.

Figure 1. Voltage stability assessment in the presence of a wind power plant

186 The plot shown in Fig. 1 visualizes the effects of wind variability on the system’s Q − V curve as
187 the wind has the effect of shifting the parabola downwards. The voltage stability criterion dictates that
188 the operating on the right hand side of the reactive power characteristic curve ensures the steady-state
189 voltage stability of the system [1]; thus, a preliminary indicator of steady-state voltage instability is
190 voltage drop in the transmission system.
191 Thermal stability analysis in the transmission system is intended to determine the power transfer
192 capabilities of the lines by assessing the temperature generated by the energized conductors as result
193 of power transfer, and it is also used to prevent line congestion. In this study, the active power loading
194 ratio is used to assess the thermal stability in transmission systems, where heavily loaded lines that
195 transfer active power over long distances can be an indication of thermal instability. Active power
196 loading level of transmission lines in respect to variability of wind power is described in Equation (16).

197 2.3. Contingency Operation Analysis

198 The severity indices are meant to identify the severity of changes between pre-contingency and
199 post-contingency operating states of the system. The rankings for contingency analysis share the
200 following criteria [13], [25], [26]:
 Version July 10, 2020 submitted to Journal Not Specified 8 of 24

201 1. A sole severity index cannot appropriately represent the entire system state and any individual
202 index does not provide a 100 percent accurate ranking.
203 2. A combination of indices should be investigated in any given system to reliably rank and select
204 contingencies.
205 3. Indices might be computed based on single or multiple methods.
206 4. If the system is currently operating at maximum generation or maximum scheduled generation,
207 and none of the pre-contingency or post-contingency analyses leads to instability, then the system
208 is assumed to be safe.
209 5. In power system operating centers, using sophisticated algorithms and procedures is
210 computationally burdening due to the complexity and size of the system and the time
211 requirements.

212 This research uses a security index that is capable of identifying the most severe contingency
213 events of the system in the presence of offshore wind power plants and to determine, from a large
214 number of contingencies, those that may lead the system to operational risk. This index considers
215 multiple facets of system’s operation and is computed as follows.

216 2.3.1. Voltage Regulation Index

Voltage regulation across the system in respect to variability of wind power was described in
Equation (14). To assess the effectiveness of voltage regulation across the system for any given level of
wind power generation, the Voltage Regulation Index (VRI) could be defined by:
v
u n 2
Vimeasured − Vischeduled
u1 
VRI = t
n ∑ Vischeduled
(20)
i =1

217 where Vscheduled and Vmeasured represent the scheduled and computed voltages in the transmission
218 lines and n is the number of transmission lines, respectively. This index measures the error in
219 voltage regulation across the power system and is ranged between 0 and 1. As the voltage regulation
220 performance improves, this index approaches zero. Higher values of this index indicate considerable
221 error in voltage regulation as a result of poor voltage performance in the system along with an
222 indication that system is approaching a voltage violation boundary. The values greater than 1 are the
223 indication of voltage violation that might require corrective action in the system.

224 2.3.2. Power Loss Index

Power loss across the system is associated with the level of active power flowing across the
conductors. The active power flow across the system in respect to variability of wind power was
described in Equation (16). To assess the cumulative power loss in the transmission lines in the system,
for any given level of wind power generation, the Power Loss Index (PLI) could be defined by:

n 2
Pimeasured

1
PLI =
n ∑ Pirated
(21)
i =1

225 where Pmeasured and Prated represent the computed active power flow and rated active power transfer
226 capability of the lines and n is the number of lines, respectively. The power loss in the transmission
227 lines and the heating rate are directly correlated. The increase of loading rate in the lines results in
228 higher power loss, and greater values of this index indicate higher levels of line loading rate. Values
229 between 0 and 1 indicate a safe and reliable operation of the system. Whereas values greater than 1
230 indicate one or multiple overloaded line(s) in the system. In this case, if a corrective action is not taken
231 appropriately, then a cascading trip might occur as a result of heating stability violation.
 Version July 10, 2020 submitted to Journal Not Specified 9 of 24

232 2.3.3. Transmission Line Loading Index

Loading of transmission lines including active and reactive power in respect to variability of wind
power was described in Equation (18). To quantify the loading level of the transmission lines for any
given level of wind power generation, the Transmission Line Loading Index (TLLI), could be given by:
n 
Simeasured

1
TLLI =
n ∑ Sirated
(22)
i =1

233 where Smeasured and Srated represent the computed loading level and rated total power transfer capability
234 of lines and n is the number of lines, respectively. This index measures total power, including active
235 and reactive powers. As the loading level in the system increases, this index also increases indicative
236 of an increased likeliness for line congestion to occur. Values between 0 and 1 are an indication of safe
237 operation of the system. However, values greater than 1 are an indication of an overloaded line in
238 the system that requires a corrective action to prevent line congestion and the interruption of power
239 delivery.

240 2.3.4. Reactive Reserve Support Index

Voltage security of the system depends on the availability of reactive support in the system.
Availability pf reactive power in the area in respect to variability of wind power was described in
Equation (11). Therefore, Reactive Capability Index (RCI) could be defined by:

Q available Area
RCI = (23)
Qload

241 where Q available Area and Qload represent total available reactive power and total reactive power
242 consumption in the power area, respectively. Greater values of RCI measure higher availability
243 of reactive power for voltage support and is an indicator of higher voltage security; the operating
244 point of the system is a further distance from the point of bifurcation in which voltage collapse occurs.
245 Conversely, smaller values are indicative of the potential for lack of available reactive power in the
246 area that might trigger a voltage collapse in the system. Fig. 2 illustrates the relationship between
247 available reactive power and voltage security.

Figure 2. Conceptual visualization of relationship between available reactive power and voltage
security

RCI could be rearranged by:

Qload
RRSI = (24)
Qreserve Area

248 This index is referred to as RRSI, the Reactive Reserve Support Index and its values ranges 0
249 and mathematically infinite. In fact, the smaller values of this index (closer to zero) indicate higher
 Version July 10, 2020 submitted to Journal Not Specified 10 of 24

250 available reactive power in the area which is an indication of a greater voltage security margin. As the
251 state of system begins to move towards the bifurcation point, this index tends to larger values and,
252 eventually, tends to infinite as the system runs out of available reactive power. As this trend continues
253 corrective action should be taken to avoid collapse of the system.

254 2.3.5. Security Index

In the process of contingency analysis, contingency events are applied to the system one at a time.
For each contingency event, to quantify variability of the wind power plant, power system variables
including power flows, voltage magnitudes and generation data are computed following incrementing
the wind power throughout the full operational range of the offshore wind power plant with credible
steps. Using the output variables of the syste, these four indices can be computer. As a results, for m
steps of wind power shift and u contingency events, a security matrix can be formed by:

Ψ11 · · · Ψ1k
 
 .. ..
Ψ= . (25)

. 
Ψ j1 · · · Ψ jk

255 for Ψ = VRI, PLI, TLLI, and RRSI, j = 1, ..., m and k = 1, ..., u. The VRIjk , PLIjk , TLLIjk and
256 RSSIjk are the calculated indices for j-th level of wind power throughout the full span of wind power
257 variability of m steps of power increment and the k-th operating condition among u contingency
258 events.
259 Then, for each index, there can be formed a vector that quantifies variability of the wind power.
260 This vector is defined by:

j
Ψk = ∑ Ψφk (26)
φ =1

261 for Ψ = VRI, PLI, TLLI, and RRSI, j = 1, ..., m and k = 1, ..., u. The VRIjk , PLIjk , TLLIjk and RSSIjk .
The final Security Index (SI), defined by the risk function for a k-th contingency event can be
defined by SIk :

SIk = s1 VRIk + s2 PLIk + s3 TLLIk + s4 RRSIk (27)

262 where s1 , · · · , s4 are weighting factors and are chosen based on planner’s preference and experience
263 and the grid architecture. The range of values for this index depends on the chosen values for the
264 weighting factors. Accordingly, greater values of this index represent higher severity of the given
265 operating condition or contingency event.

266 3. Case Study

267 This study considers integration of 1,000 MW of offshore wind power in Lake Erie into
268 the FirstEnergy/PJM transmission system as a case study to verify the utility of the introduced
269 methodology. This section describes the details of this case study and the computational
270 implementation of this methodology.

271 3.1. FirstEnergy/PJM Power System

272 In this study, an accurate computer model of the PJM system is used. The PJM is a regional
273 transmission operator in the Midwestern United States. It is part of the Eastern Interconnection and
274 operates an electric transmission system. The FirstEnergy is a regional utility company, based in
 Version July 10, 2020 submitted to Journal Not Specified 11 of 24

275 Akron, Ohio, within a geographical sub-region of the PJM and serves 6 million customers in Ohio,
276 Pennsylvania, West Virginia, Virginia, Maryland, New Jersey, and New York.

277 3.2. Wind Power Integration Scenario Development

278 The estimated geographic distribution of available wind power over the surface of Lake Erie
279 was provided by the NREL, and, based on these estimates, five candidate POIs, including Ashtabula,
280 Perry, Eastlake, Lake Shore, and Avon Lake, were identified in the FirstEnergy transmission system.
281 Accordingly, three scenarios for the integration of a total of 1,000 MW of offshore wind generation at a
282 variety of POIs situated within the FirstEnergy/PJM system, subject to compatibility with the extant
283 grid infrastructure, were developed:
284 1. Interconnecting a total of 1,000 MW of offshore wind generation at a single POI, referred to as
285 EC01
286 2. Interconnecting a total of 1,000 MW of offshore wind generation through five 200 MW POIs
287 across the lake, referred to as EC02
288 3. Interconnecting a total of 1,000 MW of offshore wind generation through two 500 MW POIs
289 across the lake, referred to as EC03
290 Fig. 3 shows the geographical location of the offshore wind power plant and the POIs.

Figure 3. Map of Interconnection Scenarios into the FirstEnergy System

291 3.3. Offshore Wind Power Plant Model Development and Computer Modeling
292 For computational purposes, the 1,000 MW offshore wind power plant is modeled by an
293 aggregated model that includes a conventional generator connected to a PV bus through a 34.5kV
294 collector and export cables for the offshore system. The wind turbine model parameters used in this
295 study are based on a Type 3 GE 3.6-MW offshore wind turbine with a variety of capability to provide
296 reactive power support and the control strategy. Accordingly, the R00 models correspond to a machine
297 with no reactive support capability, the R01 to machine with limited reactive support capability (-7.5% /
298 7.5%), and R02 to machine with full reactive support capability (-43.0% / +57.8%). The wind machines
299 were set to cover a complete range of power generation by offshore wind power plant from 1,000MW
300 to 0MW. The substation transformer is rated on the basis of the wind turbine generators and has a
301 typical impedance of approximately 8% [27].
302 A static VAR compensator (SVC) with significant compensation capacity was incorporated at each
303 POI to provide ancillary service (reactive power generated/consumed) as needed to the grid. During
304 steady-state analysis, the SVC was modeled as a controllable shunt device. This allowed for automatic
305 voltage control during load flow simulations, and the device was modeled to regulate voltage at the
306 POI [27].
 Version July 10, 2020 submitted to Journal Not Specified 12 of 24

307 3.4. Generation Dispatch Scenarios and Load Assumptions

308 In this investigation, summer 2015 load data from the Eastern Interconnection were modeled.
309 This model contains 63,000 system buses, 9,091 generators, with a total of 936,266 MW and 444,200
310 MVAr of installed generation capacity to serve the load of 685,469 MW and 201,005 MVar.
311 The main objective of this project is to study the integration of 1,000 MW of offshore wind
312 generation into the FirstEnergy/PJM system. The second objective is to consider the impacts of the
313 retirement or loss of one of the largest power plants in the area (Perry, with a capacity of 1,200 MW
314 and 625 MVAr) upon integration of the offshore wind power plant. Thus, the generation units in
315 the area were redispacthed by running power flow to maintain the balance between generation and
316 consumption in the area for each of the developed interconnection scenarios.

317 3.5. Contingency Events

318 The FirstEnergy provided a contingency list that included generation units, transformers, line
319 segments, complete lines, and synchronous condenser outages. These events were used to perform the
320 contingency analysis.

321 3.6. Computer Implementation

322 Computational implementation of the steady-state analysis was carried out using 36 cases that
323 were developed for each of the interconnection scenarios (EC01, EC02, and EC03), for a total of 108
324 cases. These cases included three degrees of freedom for the reactive capability of the offshore wind
325 machines (0MVAr, -7.5%/+7.5%MVAr and -43.0%/+57.8%MVAr), three degrees of freedom for the
326 point of voltage regulation (voltage regulation at the machine terminals, voltage regulation at the POI,
327 and voltage regulation at the collector system), two degrees of freedom for the operation of the SVC
328 (on and off), and two degrees of freedom for the operation of Perry (on and off). The description of
329 these cases is shown in Table 1.
330 First, accurate models of FirstEnergy/PJM were provided. Then cases for each of the scenarios
331 were developed and integrated into the base case models. After loading the system, generation
332 dispatch, and load data, the wind models were initiated.

333 3.6.1. Normal Operation

334 For each of the cases, the wind generation was shifted from its 100% capacity to 0% by steps of
335 1%. Through every step change in wind generation, other non-wind generator units in the FirstEnergy
336 area were redispatched to pick up missing generation of the wind power plant. Then a power flow
337 calculation was done.
338 The results of the power flow for each 1% of wind generation variability from all transmission
339 lines, generation units, and substations in the FirstEnergy operating territory were recorded and
340 merged. As a results, two datasets were produced for each step:
341 • The first dataset contains information of voltage magnitudes of the bus bars.
342 • The second dataset contains information of power flows of the transmission lines.
343 These two sets of data consist of all possible voltage magnitude and power flow data for the
344 system and cover the complete range of operating points for the system.
345 In an ideally stable power system, voltage magnitudes across the system are regulated at 1.0 p.u.
346 This determines voltage regulation data for this system with identical measures of mode, mean, and
347 median of 1, with an estimated density function that follows a normal distribution. As the voltage
348 magnitudes across the system begin to deviate from 1.0 p.u. and spread out with a wider range, these
349 measures start to vary. As a result, estimated density function begins to deviate from the normal
350 distribution.
351 Based on this notion, to analyze each dataset and process them to reveal meaningful information,
352 the following statistical approach was used:
Version July 10, 2020 submitted to Journal Not Specified 13 of 24

Table 1. List of Studied Cases in Steady-State Analysis

Case No. Voltage Control Reactive Capability SVC Perry
1 V00
2 V01 R00
3 V02
4 V00
5 V01 R01 OFF
6 V02
7 V00
8 V01 R02
9 V02
ON
10 V00
11 V01 R00
12 V02
13 V00
14 V01 R01 ON
15 V02
16 V00
17 V01 R02
18 V02
19 V00
20 V01 R00
21 V02
22 V00
23 V01 R01 OFF
24 V02
25 V00
26 V01 R02
27 V02
OFF
28 V00
29 V01 R00
30 V02
31 V00
32 V01 R01 ON
33 V02
34 V00
35 V01 R02
36 V02
R00: (0MVAr), R01: (-7.5%/+7.5%MVAr), R02: (-43.0%/+57.8%MVAr)
V00: Voltage regulation at the terminals of machines
V01: Voltage regulation at the POI
V02: Voltage regulation at the collector system
 Version July 10, 2020 submitted to Journal Not Specified 14 of 24

353 1. First, kernel density estimation (KDE) was applied to each of the datasets to estimate the density
354 function of each dataset with an equal weight for all data points. The outcome of KDE shows
355 how the variability of wind generation would impact voltage regulation and loading levels for
356 each of the cases studied.
357 2. The next step was to calculate values of maximum, minimum, mode, mean, and median of each
358 of the new datasets. These values provide a sufficient understanding of how the data points are
359 distributed and how close to a normal distribution they are distributed.

360 • Maximum and minimum measures indicate whether or not any voltage violation occurs,
361 according to acceptable grid operating criterion.
362 • Mode measure shows what voltage magnitude has been recorded the most frequently. The
363 closer the mode values are to 1 is an indication of greater performance of the system in
364 voltage regulation.
365 • The last metric is the difference between mean and median. The smaller values for this
366 measure indicate that the voltage regulation data have a symmetric distribution, a normal
367 distribution. Positive values for this difference show a trend to voltage rise across the
368 system, whereas negative values indicate a trend to voltage drop across the system. This
369 measure is similar to kurtosis and skewness in which the heaviness of the tail and shoulders
370 of the distribution is used for interpretation. but it is easier to compute for datasets from
371 power systems where the long tails of the probability distribution may not be a concern.
372 3. The last step was to compare the datasets to the normal distribution. The outcome of this
373 comparison provides information about the probability of any given operating condition, both
374 voltage regulation and line loading level, with respect to variability of the wind within the full
375 operating span of the offshore wind power plant.

376 3.6.2. Contingency Operation

377 For each of the cases studied, the wind power generation was shifted from 100% capacity to 0%
378 by steps of 10%. Then a power flow calculation was performed. For each step of power shift, the
379 results from all of transmission lines, generation units and substations in the FirstEnergy operating
380 territory were recorded.
381 If the power flow calculations converged and intact operation was stable, then a series of
382 contingency events were applied to the system, one outage event at a time. This list was provided by
383 FirstEnergy and consisted 64 condensed contingency events composed by generation, line segments,
384 entire lines, and synchronous condenser outages. During each outage event, the sensitivity analysis
385 for all levels of wind power was carried out the operational variables of the system recorded.
386 Then, by using the proposed contingency ranking method, for each level of wind power, the four
387 indices, VRI, PLI, TLLI and RRSI, were computed. The next step was to compute the overall security
388 index, SI for each selected operating condition/contingency. In this study, the weighing factors for
389 s1 , · · · , s4 were assigned, equally, 1. Then, the contingency events were ranked in a descending order
390 according to their security index.
391 To this end, the security index was computed for each step of wind, SIjk for j-th level of wind
392 power and kth contingency event. Then, SIk , the security index for k-th event, was computed by
393 summing up the computed SIjk from all levels of wind power, as described in Equation 27.
394 The final results of the security index computation were ranked in descending order to identify
395 the most severe events for the offshore wind power plant integration.

396 4. Results and Discussion

397 During normal operation, no contingency occurred, and all of the generation units and load are
398 operational. The goal of this study was to investigate how variability of the offshore wind power
399 plant affects voltage stability and loadability of the transmission system and assess whether or not the
 Version July 10, 2020 submitted to Journal Not Specified 15 of 24

400 stability of the system was threatened as a result of the operation of the offshore wind power plant
401 during its complete operational range. This section presents the results from the cases studied.

402 4.1. Normal Operation

403 4.1.1. Voltage Stability

404 The main purpose of this section is to investigate the impacts of offshore wind generation on
405 voltage regulation in transmission system and its steady-state voltage stability. The FirstEnergy/PJM
406 requires that 138kV to 345kV transmission and sub-transmission level apparatus operate within the
407 ranges of 0.92-1.05 pu [28].
408 Fig. 4a and Fig. 4b show the results from voltage regulation across the transmission system in
409 the base case (574 lines in total). These curves are obtained by using probability density estimation
410 of the computed voltage magnitudes data points and by comparing their distribution to the normal
411 distribution.

(b) Distribution probability

(a) Probability density function
Figure 4. Distribution measures for the voltage regulation in the FirstEnergy transmission system -
base case

412 These two curves show that no undervoltage condition has been recorded; however, two 138-kV
413 lines are operating in a slightly overvoltage condition. In addition, 99.6% of the lines operate within
414 the range of 0.977-1.035 p.u. The mean and mode values of the measured voltage magnitudes are
415 1.002 p.u. and 1.013 p.u., respectively. This is not surprising because reactive power generation is
416 significantly greater than required in this area.
417 Figs. 5 and 6 show how the density function of the voltage magnitude distribution changes with
418 respect to the variability of the offshore wind power plant.
419 As shown in Figs. 5 and 6, as the level of power generation by the wind power plant increases,
420 the distribution of the measured voltage magnitudes tend toward normal. It is notable that for higher
421 levels of power generation by the offshore wind power plant, the voltage magnitudes at a greater
422 number of bus bars tend toward 1.014 p.u. and the local peak at 0.99 p.u. vanishes. At 1,000 MW of
423 offshore wind generation, the mode, median, and mean values of the voltage data are 1.014 p.u., 1.003
424 p.u., and 1.004 p.u., respectively. This suggests that as the offshore wind power plant generates higher
425 power, it improves the voltage regulation across the transmission system and the recorded dataset has
426 a symmetric distribution, similar to that of a normal distribution.
427 However, the lower levels of offshore wind generation cause more asymmetry in the distribution
428 of the measured voltage data. For instance, at 0 MW of offshore wind generation, the mode value of
429 the voltage data remains at 1.014 p.u., whereas the median and mean values of the voltage data are
430 1.006 p.u. and 1.005 p.u., respectively. In fact, the level of offshore wind generation has a direct impact
431 on the median and mode values of the measured voltage magnitudes.
 Version July 10, 2020 submitted to Journal Not Specified 16 of 24

Figure 5. Wind power variability vs. probability density function of voltage regulation curve in
FirstEnergy/PJM transmission system

Figure 6. Impact of wind generation variability on voltage regulation in the FirstEnergy/PJM

transmission system

432 However, it is evident that no voltage violation takes place at all levels of offshore wind. In
433 addition, the bus that experienced an overvoltage condition in the base case is properly regulated after
434 adding wind generation.
435 Fig. 7 illustrates the impact of variability of the wind power plant on the distribution of voltage
436 magnitudes across this transmission system. The plot in this figure consists of a group of curves
437 wherein each represents the operational performance of voltage regulation in the system for a certain
438 level of offshore wind generation. Each curve compares the distribution of the voltage data points from
439 a given level of wind generation to the normal distribution. These results reveal that after installing the
440 offshore wind power plant, for all levels of offshore wind generation, this system is regulated within
441 the standard voltage range. In the base case, there were two data points within the overvoltage range
442 that were regulated properly upon operation of the offshore wind power plant. This is an indication of
443 improved voltage stability, and it is also shown that the impact of offshore wind generation variability
444 on voltage regulation is insignificant because the data distribution for all levels of the wind generation
445 is identical.
446 Figs. 5, 6, and 7 illustrate the results from the EC01 interconnection scenario. The results from
447 EC02 and EC03 are identical and are not shown.
448 From these results, it can be concluded that the integration of the offshore wind power plant into
449 this transmission system improves the voltage regulation in the area for all levels of offshore wind
450 generation. In fact, the buses that experience overvoltage in the base case (this system without a wind
 Version July 10, 2020 submitted to Journal Not Specified 17 of 24

Figure 7. Wind power variability vs. distribution probability of voltage regulation curve in the
FirstEnergy/PJM transmission system

451 power plant) were regulated properly upon the integration of the offshore wind power plant. As the
452 level of offshore wind generation increases, the improvement in voltage regulation also increases.
453 To take a closer look at voltage regulation in this transmission system and assess the distribution
454 of their dataset, the simulation results from the cases studied and integration scenarios were collected
455 in datasets and then the datasets corresponding to various levels of wind were merged and analyzed.
456 Figs. 6-8 show range, mode, and difference between median and mean of measured voltage for all of
457 the interconnections: EC01, EC02, and EC03.

Figure 8. Range of voltage magnitudes in the FirstEnergy transmission system

458 Fig. 8 shows the range of measured voltage magnitudes across the this transmission system
459 from all of the cases studied. The results here show that the integration of offshore wind into this
460 transmission system improves the voltage regulation in the area for all levels of wind generation. In
461 fact, the maximum recorded voltage in the base case (system without a wind power plant), shown by
462 the red solid line , were regulated properly and dropped closer to the standard range. This agrees with
463 the findings from Figs. 5, 6, and 7.
464 The second point to note is that there are no voltage violations with the integration of the offshore
465 wind power plant in all of the EC01 and EC03 cases and the majority of the EC02 cases. For EC02
466 interconnection scenario (5 POIs), the cases in which the wind machines have either no or limited
467 reactive support capability, voltage violations occurred (shown by the red dashed line). Fig. 9 shows
468 the distribution of the voltage magnitudes for the EC02 cases with a voltage violation.
469 The results shown in Fig. 9 outline that in those cases in which voltage violation occurred, the
470 distribution of the probability of the voltage violation is less than 4% during the full operational range
 Version July 10, 2020 submitted to Journal Not Specified 18 of 24

Figure 9. Distribution probability of voltage regulation curves in the FirstEnergy/PJM transmission

system for the EC02 cases 1-9 with a voltage violation

471 of the offshore wind power plant. This indicates that the location of the POI and the capability of the
472 wind machine to provide reactive power support are important factors in determining the voltage
473 regulation following the integration of the offshore wind power plant.

Figure 10. Mode of voltage magnitudes in FirstEnergy transmission system

474 Fig. 10 shows the mode value for the voltage data for all the cases studied. These results indicate
475 that the mode value for the voltage data in all cases studied with and without the offshore wind power
476 plant is identical (1.01 p.u.). This is an indication that this transmission system remained stiff and well
477 regulated upon the integration of the offshore wind power plant.
478 Fig. 11 shows the difference between mean and median values for the voltage data for all the
479 cases studied. These values are calculated using aggregated voltage data points from all levels of
480 offshore wind generation. It can be seen that the symmetry of the voltage magnitudes in the EC01 and
481 EC03 cases with Perry online remained approximately the same as the base case. The EC01 cases with
482 Perry offline showed a slight tendency toward voltage rise, whereas the EC03 cases with Perry offline
483 showed a slight tendency toward voltage drop. However, neither of these two integration scenarios
484 led the system to instability (no undervoltage violation). The EC02 cases with the SVC operated at
485 the POI showed a distribution identical to that of the base case, whereas the EC02 cases with no SVC
486 operating at the POI showed a tendency toward voltage rise. The contribution of the reactive capability
487 and the voltage control scheme of the offshore wind power plant are not significant.
488 It is noticeable that the EC01 cases deliver the most robust performance in voltage regulation
489 across this system for all levels of wind generation, with and without Perry. In all the cases studied,
490 the computed metrics were very small; thus, it is safe to conclude that the voltage magnitudes under
 Version July 10, 2020 submitted to Journal Not Specified 19 of 24

Figure 11. Difference between mean and median of voltage magnitudes in the FirstEnergy transmission
system

491 steady-state operation were nearly symmetrical and the distribution of data in the datasets is very
492 close to that of the base case (a normal distribution).
493 One of the primary signs for steady-state voltage instability is undervoltage violation at any given
494 bus across the system as the operating point begins to approach the voltage collapse point. The results
495 of the voltage analysis reveal that this transmission system does not experience this condition. In
496 addition, in all cases studied, including the base case, the difference between mean and median was
497 very small; therefore, for all levels of offshore wind generation, the system operates in a stable region.
498 It is safe to say that the integration of the offshore wind power plant in Lake Erie will not negatively
499 affect the steady-state voltage stability of this transmission system. Additionally, the results from
500 cases without Perry (cases 19 to 36) show that upon integration of 1,000 MW wind power plant using
501 all three studied interconnection scenarios, retirement or loss of Perry does not adversely influence
502 voltage stability of the system.

503 4.1.2. Thermal Stability

504 The aim of this section is to understand how offshore wind generation will impact the loading
505 level of this transmission system and its steady-state thermal stability.
506 Fig. 12 presents the results from the analysis of the transmission system loading in the base case.
507 The results here indicate that almost 99.5% of this transmission lines in the base case were loaded
508 below 80% of their nominal rating, whereas more than 50% of them are loaded below 25% of their
509 nominal rating. However, there is a 138-kV line operating in an overloaded condition, exceeding 3.1%
510 of its nominal loading capacity.
511 Figs. 13 and 14 visualize the effect of wind power generation variability on the loading level of
512 this transmission system. From these curves it can be noticed that for a higher level of wind generation,
513 the loading level mode moves to 23% from 21% because of power injected by the offshore wind power
514 plant into this system, which changes the flow of the power in the system.
515 As shown in Fig. 14, there is a very small difference in the distribution of line loading data for this
516 transmission system after the introduction of the wind power plant and for different levels of wind
517 generation. This is because this system is very lightly loaded and therefore can easily accommodate
518 the generated power by the wind power plant.
519 To investigate how the different integration scenarios could impact line loading, Fig. 15 compares
520 the results from the cases studied. This figure illustrates the distribution of loading level of this
521 transmission system for the cases studied. This figure shows that the distribution plots for all of
522 the cases studied follow two main patterns and overlap to form two main plots. The factor that
523 distinguishes the cases studied and their plots is the operational status of Perry. As shown in this
 Version July 10, 2020 submitted to Journal Not Specified 20 of 24

Figure 12. Distribution probability of line loading in the FirstEnergy transmission system - base case

Figure 13. Wind power variability vs. probability density function of loading level in FirstEnergy/PJM
transmission system

Figure 14. Impact of wind power variability on loading level in the FirstEnergy/PJM transmission
system

524 figure, in all of the cases, with and without Perry, there is a narrow chance of experiencing an overload
525 condition in a line or multiple lines within this system. The chance of this condition for lines in cases
526 with Perry online is 0.002% at 13% of their nominal rating and in cases with Perry offline is less than
527 0.001% at 3% of their nominal rating. The plots shown in Fig. 15 are from EC01. The results from EC02
528 and EC03 are nearly identical and are not shown. Thus, it can be concluded that the location of the
529 POIs does not really affect the overall loading of this system.
 Version July 10, 2020 submitted to Journal Not Specified 21 of 24

Figure 15. Distribution probability of loading levels in the FirstEnergy/PJM transmission system

530 In summary, the primary factor in the loading level of this transmission system is the operational
531 status of Perry. Lack of operation of this power plant, with 1,290 MW and 625 MVAr capacity, causes
532 the loading level to drop slightly. The second factor is the level of power production by the offshore
533 wind power plant. Other factors examined have no notable impact on the loading level of the system.
534 From the results shown in this section, it can be concluded that this transmission system is capable,
535 in terms of capacity, of accommodating 1,000 MW of offshore wind generation, as introduced in this
536 study. Additionally, retirement or loss of Perry does not threatens the transfer capability of this system
537 upon integration of the wind power plant, for all three integration scenarios. Overall, the thermal
538 stability is a concern when the transmission lines operate in a heavily loaded condition; however, in
539 this study, it was found that this transmission system does not experience this condition.

540 4.2. Contingency Operation

541 This section describes the results from the steady-state contingency analysis using the 64 most
542 credible events and outlines the findings. The main application of the proposed security index is to
543 conduct contingency ranking and monitoring in power systems with variable generation units and to
544 provide an assessment of how their variability effects the security of the system.
545 Figs. 16a through 16c shows the security index for the cases studied with and without operation
546 of Perry. The higher values of the security index indicate lower security for the system. Having this in
547 mind, in each of the figures, the upper plate represents the cases with Perry offline while the lower
548 plate represents the cases with Perry online. Accordingly, these figures reveal that the loss of Perry
549 adversely affects the security of the system. But it does not interrupt the power delivery in the system.
550 Obviously, loss of a 1,200MW power plant adversely influences the security of the system. Hence, this
551 shows the effectiveness of the contingency ranking method introduced in this research.
552 The SI for the cases studied holds the information necessary to identify the severe events. The
553 greater the index, the more severe the contingency event, for any given case. Table 2 provides a list of
554 the top 10 severe contingencies for six selected cases studied. The selection of these cases was done
555 on basis of the significance of Perry’s operation in the contingency operation of the system as it was
556 established in Figs. 16a through 16c). Hence, for comparison purposes, the cases with and without
557 Perry are compared here.
558 The results shown Table 2 yield a consistent ranking for the contingency of the cases studied for
559 all interconnection scenarios, EC01, EC02 and EC03. For instance, Events 24 and 40 are consistently
560 ranked 1st or 2nd regardless of the interconnection scenario and operational status of Perry. However,
561 rankings of event 17 are noticeably different for the cases with and without Perry. The SI of this event
562 for Case 1, EC01, is 1.240 placing it in the 8th position in the severity ranking while for Case 7, EC01,
563 the SI is 1.446, ranking this event at 34th. This is because of change of power flow in the system as a
564 result of this topology change. In fact, this contingency represents loss of a 345 kV transmission line
 Version July 10, 2020 submitted to Journal Not Specified 22 of 24

(b) Contingency Ranking of EC02 Cases

(a) Contingency Ranking of EC01 Cases

(c) Contingency Ranking of EC03 Cases

Figure 16. Contingency ranking in the FirstEnergy transmission system

Table 2. List of Top 10 Most Severe Contingency Events and Their Ranking for Selected Cases
Case 1 - EC01 Case 7 - EC01 Case 1 - EC02 Case 7 - EC02 Case 1 - EC03 Case 7 - EC03
Event Rank SI Rank SI Rank SI Rank SI Rank SI Rank SI
40 1 1.280 2 1.530 1 1.287 2 1.564 1 1.362 1 1.593
24 3 1.276 1 1.544 2 1.285 1 1.576 2 1.348 2 1.582
39 4 1.247 6 1.488 5 1.255 6 1.521 5 1.322 7 1.534
18 5 1.247 3 1.521 1 258 3 1.561 4 1.323 3 1.574
41 6 1.244 7 1.486 7 1.248 9 1.505 6 1.319 8 1.528
30 7 1.243 8 1.484 6 1.251 7 1.516 7 1.311 9 1.522
17 8 1.240 34 1.446 12 1.241 43 1.470 10 1.300 40 1.480
9 9 1.236 25 1.451 13 1.237 42 1.471 17 1.296 44 1.479
22 10 1.234 9 1.479 9 1.237 8 1.512 8 1.304 10 1.522
62 11 1.232 14 1.460 8 1.243 13 1.492 15 1.297 23 1.490

565 that is transmitting power from Perry. Hence, the offline status of Perry affects its contingency severity.
566 This shows the effectiveness of the contingency ranking method introduced in this research.
567 The results also indicate that the EC01 interconnection provides the highest security during the
568 contingency operation. The SI values for the EC01 interconnection is the smallest amongst the three
569 interconnection considered here.
570 It is noticeable that the SI values for cases in which Perry is offline, are significantly higher than
571 the same cases in which Perry is online, demonstrating that the loss of Perry causes a drop in security
 Version July 10, 2020 submitted to Journal Not Specified 23 of 24

572 of the system. This agrees with the finding evidenced in Figs. 16a through 16c and indicates the utility
573 of the present contingency ranking approach.

574 5. Conclusion
575 This study present a statistical approach for steady-state stability analysis of large-scale power
576 systems including normal and contingency operation. This method is practical for system planning
577 studies while integrating offshore wind power plants and is easy and efficient to implement. In
578 addition, it provides explicit information about the steady-state operating conditions of a large-scale
579 transmission system throughout the complete range of operation of the variable generation power
580 plant. The normal operation relies on statistical measures within the system’s recorded data and their
581 distribution. The contingency analysis relies on a risk-based security index to identify the most severe
582 contingency events and most secure interconnection scenario by considering variability of the offshore
583 wind power plant.
584 This study focused on investigation of steady-state operational impacts of integration of 1,000 MW
585 of offshore wind in the U.S. Great Lakes region. Specifically, this study focused on FirstEnergy/PJM
586 transmission system by considering its current configuration and operational status. In particular,
587 operational issues including voltage regulation and power transfer capability were investigated.
588 Feasible offshore wind integration scenarios were examined to identify vulnerability of the system.
589 The 63k-bus U.S. Eastern Interconnection was used.
590 The results showed that the integration of offshore wind power plants could improve voltage
591 stability across the system. The results also showed that in a lightly loaded power system, such as
592 FirstEnergy, line congestion is not a concern when adding wind generation capacity. Additionally, in
593 this particular case study, the retirement or loss of Perry does not deteriorates the steady-state stability
594 of the system upon integration of 1,000 MW wind power plant for all three studied interconnection
595 scenarios.
596 Although this study focused on integration of an offshore wind power plant, this methodology is
597 applicable for integration of all types of variable generation units.

598 Funding: This work has been supported by US Department of Energy under grant No. DE–EE0005367: Great
599 Lakes Offshore Wind: Utility and Regional Integration Study. The NREL’s contribution to this work was supported
600 by the U.S. Department of Energy under Contract No. DE-AC36-08GO28308 with the National Renewable Energy
601 Laboratory.
602 Acknowledgments: The authors thank S. Barnes and R. D’Aquila from GE and R. Kolacinski from Case Western
603 Reserve University for their assistance.
604 The U.S. Government retains and the publisher, by accepting the article for publication, acknowledges that the
605 U.S. Government retains a nonexclusive, paid-up, irrevocable, worldwide license to publish or reproduce the
606 published form of this work, or allow others to do so, for U.S. Government purposes.
607 Conflicts of Interest: The authors declare no conflict of interest.

608 References
609 1. J. Machowski, J. Bialek, and J. Bumby, Power System Dynamics: Stability and Control (John Wiley & Sons, 2008).
610 2. A. Sajadi, R. Kolacinski, and K. Loparo, in Innovative Smart Grid Technologies Conference (ISGT), 2015 IEEE
611 Power & Energy Society (IEEE, 2015), pp. 1–5.
612 3. E. Vittal, M. O’Malley, and A. Keane, Power Systems, IEEE Transactions on 25, 433 (2010).
613 4. S. S. Baghsorkhi and I. A. Hiskens, in Power and Energy Society General Meeting, 2012 IEEE (IEEE, 2012), pp.
614 1–7.
615 5. H. Li, E. Haq, K. Abdul-Rahman, J. Wu, and P. Causgrove, International Transactions on Electrical Energy
616 Systems 24, 1618 (2014).
617 6. G. Ejebe, G. Irisarri, S. Mokhtari, O. Obadina, P. Ristanovic, and J. Tong, in Power Industry Computer Application
618 Conference, 1995. Conference Proceedings., 1995 IEEE (IEEE, 1995), pp. 249–255.
619 7. A. Yokoyama and Y. Sekine, Proceedings of Bulk Power System Voltage Phenomena-Voltage Stability and
620 Security, Potosi, Missouri (1989).
 Version July 10, 2020 submitted to Journal Not Specified 24 of 24

621 8. H.-D. Chiang and R. Jumeau, Power Systems, IEEE Transactions on 10, 584 (1995).
622 9. O. B. Foss, N. Flatabo, B. Hahavik, and A. T. Holen, in Athens Power Tech, 1993. APT 93. Proceedings. Joint
623 International Power Conference (IEEE, 1993), vol. 1, pp. 216–221.
624 10. A. S. Meliopoulos and C. Cheng, Proceedings of the SoutheastconâTM 89 pp. 837–842 (1990).
625 11. P. M. Subcommittee, Power Apparatus and Systems, IEEE Transactions on pp. 2047–2054 (1979).
626 12. V. Donde, V. Lopez, B. Lesieutre, A. Pinar, C. Yang, and J. Meza, in Power Symposium, 2005. Proceedings of the
627 37th Annual North American (IEEE, 2005), pp. 59–66.
628 13. V. Brandwajn, A. Kumar, A. Ipakchi, A. Bose, and S. D. Kuo, Power Systems, IEEE Transactions on 12, 1136
629 (1997).
630 14. H. Liu, A. Bose, and V. Venkatasubramanian, Power Systems, IEEE Transactions on 15, 1137 (2000).
631 15. T. R. Limbu, T. K. Saha, and J. D. McDonald, in Australasian Universities Power Engineering Conference 2005
632 (AUPEC2005, 2005), pp. 568–573.
633 16. S. Greene, I. Dobson, and F. L. Alvarado, Power Systems, IEEE Transactions on 12, 262 (1997).
634 17. J. W. Nims, A. El-Keibb, and R. Smith, Electric power systems research 43, 69 (1997).
635 18. H. Wan, J. D. McCalley, and V. Vittal, Power Systems, IEEE Transactions on 15, 1247 (2000).
636 19. P. Vijayan, S. Sarkar, and V. Ajjarapu, in Power & Energy Society General Meeting, 2009. PES’09. IEEE (IEEE,
637 2009), pp. 1–8.
638 20. N. Miller, M. Shao, S. Pajic, and R. DâTM Aquila, Contract 303, 275 (2013).
639 21. N. W. Miller, W. W. Price, and J. J. Sanchez-Gasca, GE-Power systems energy consulting (2003).
640 22. ABB, Submarine Cable Systems: Attachment to XLPE Land Cable Systems-Users Guide (2010).
641 23. J. Bebic, Tech. Rep., National Renewable Energy Laboratory, Golden, CO (United States) (2008).
642 24. R. Gutman, Power Systems, IEEE Transactions on 3, 1426 (1988).
643 25. S. C. Savulescu, Real-time stability assessment in modern power system control centers, vol. 42 (John Wiley & Sons,
644 2009).
645 26. K. Schäfer and J. Verstege, Power Systems, IEEE Transactions on 5, 539 (1990).
646 27. S. Barnes, Tech. Rep., General Electric (GE), Pittsburgh, Pennsylvania (United States) (2014).
647 28. J. Mackauer, J. Syner, P. Bowers, and J. Detweiler, Tech. Rep., FirstEnergy, Akron, OH (United States) (2013).

648 c 2020 by the authors. Submitted to Journal Not Specified for possible open access
649 publication under the terms and conditions of the Creative Commons Attribution (CC BY) license
650 (http://creativecommons.org/licenses/by/4.0/).