1

Solar Power Plant Design and Interconnection

E.H. Camm, Member, IEEE S. E. Williams, Member, IEEE

inverter design to meet interconnection requirements. While

Abstract—The development of newer technologies in
concentrating solar power (CSP) plants, particularly plants using
dish Stirling systems, as well as changes in the design of
photovoltaic (PV) inverters is creating new challenges in the
design of low- and medium-voltage collector systems for large
solar power plants. Furthermore, interconnect requirements for
reactive power, voltage, and ramp rate control and the
characteristics of solar power require unique solutions for
optimal plant design. To ensure large solar plants can be
connected successfully to the grid without impacting grid
stability or reliability, the design process must include the
development of suitable models of these plants for transient and
dynamic simulation. Simulation tools and models can then be
used to determine special requirements to deal with issues such
as daily plant energization, low voltage ride-through, temporary
overvoltage and feeder grounding, etc. The provision of dynamic
and static reactive power and the optimization thereof for
application at either low, medium, or high voltage and the
control issues associated with plant-wide reactive power and
voltage control are also key issues in the design.
The presentation will focus on the key technical issues and
design optimization of large solar power plants.
Index Terms—Concentrating solar power, photovoltaic
systems, reactive power, solar energy.
I. INTRODUCTION
small distributed generation applications have traditionally
required no reactive power capability for voltage control, fast
disconnection from the utility system in the event of a
disturbance, and no intentional islanding [3], the rules for
large transmission-connect plants require this performance
and are leading to changes in PV plant design practices.
Large PV power plants are now being designed to meet
reactive power, voltage, ramp rate control, and voltage and
frequency ride-through requirements.
II. UTILITY-SCALE PV POWER PLANT DESIGN
Utility-scale PV power plants are composed of arrays of
PV panels, which typically connect to a common DC bus and
through one or more inverters to a step-up transformer to the
collector system. See Figure 1. Designs using lower power
rated inverters also include an isolation transformer with each
inverter to step the voltage up from 208 Vac to 480 Vac. The
design of the low voltage DC and AC system is typically in
accordance with National Electrical Code (NEC)
requirements, particularly Article 690 [4]. Fully integrated
modular systems with power conversion system (i.e. inverters
and inverter controls), step-up transformer, and medium-
voltage disconnect switches are commonly used. The
collector system is similar to that of a wind farm, with

S OLAR power installations are continuing to grow at a

rapid rate in North America and worldwide. In 2010 solar
installations in the United States were up more than a 100%
multiple step-up transformers connected via medium-voltage
cable in a feeder arrangement and connected to the collector
or interconnect substation.
over that of 2009 [1]. Many of the new installations in North
America are utility-scale solar power plants connecting to
utility distribution and transmission systems. Photovoltaic
(PV) solar installations are seeing the most significant growth,
with applications for several hundred megawatts currently in
the interconnection queues. Although the US pioneered large-
EMC Filter

EMC Filter

scale use of concentrating solar power (CSP) with the nine

Filter

Solar Energy Generating Systems totaling 354 MW,

constructed in California's Mojave Desert from 1984 to 1990,
the US did not put another full-scale CSP plant online until
the 60 MW Nevada Solar One plant in 2007. Three relatively
small CSP plants have been constructed since then [2]. The
EMC Filter

EMC Filter
Filter

development of newer technologies in CSP plants, particularly

plants using dish Stirling systems, sparked new interest in
CSP in 2010.
Large utility-scale PV plants are driving changes in

Fig. 1. Typical utility-scale PV plant low-voltage system architecture.

E. H. Camm is with S&C Electric Company, Chicago, IL 60626 USA (e-
mail: ernst.camm@sandc.com).
S. E. Williams is with S&C Electric Company, Chicago, IL 60626 USA
(e-mail: stephen.williams@sandc.com).

978-1-4577-1002-5/11/$26.00 ©2011 IEEE


A. Grounding
Grounding in the low-voltage system will be in accordance
with the applicable articles of the NEC. This includes PV
array grounding at a single point, as well as equipment
grounding. Grounding on the medium-voltage system will be
in accordance with applicable NEC and NESC [5]
requirements.
Requirements for grounding transformers on the medium-
voltage feeders will be dependent on the inverter design.
Current inverter designs that are IEEE Std 1547/UL 1741 [6]
compliant will not allow generation to continue once the
feeder circuit breaker is tripped open. Thus, there will not be
a need for a grounding transformer or other means of feeder
grounding once the feeder is isolated. However, expected
changes in inverter designs to meet reactive power and low
voltage ride-through requirements will require grounding
transformer or other means of feeder grounding to limit
temporary overvoltages on the feeder to within acceptable
levels.
B. Reactive Power and Voltage Control
PV plants using inverters that are IEEE Std 1547/UL 1741
compliant will typically not have reactive power capability
and thus operate with a unity power factor. To meet
transmission interconnection requirements for reactive power
and voltage control, substation-based reactive power
compensation systems will typically be applied.
Compensation systems would typically consist of a
STATCOM with dynamic reactive power capability and
switched capacitor and reactor banks.
In the case of distribution system interconnections the
variation in plant power output can cause frequent voltage
variations and excessive operations in feeder voltage
regulators if IEEE Std 1547/UL 1741-compliant inverters
operating at unity power factor are used. Inverters with
reactive power capability and voltage control can eliminate
these problems. For existing installations, small STATCOMs
installed at the point of interconnection (POI) would be a
good alternate solution. If step voltage regulators are present
on the feeder, additional analysis will be necessary to properly
coordinate these various voltage regulation devices to avoid
negative interactions.
C. Low Voltage Ride-Through and System Stability
PV plants using inverters that are IEEE Std 1547/UL 1741
compliant do not have low voltage ride-through (LVRT)
capability since plants are required to disconnect in the event
of a disturbance on the utility system. Large penetration
levels of PV plants using these types of inverters at
distribution voltage level can affect the grid stability. New
inverter designs with reactive power and LVRT capability and
other “grid friendly” features are expected to be required in
new plant designs.
D. Short-Circuit Contributions
PV plants using inverters that are IEEE Std 1547/UL 1741
compliant will not provide short-circuit current contribution
2
for more than a few cycles. A fault on the line will typically
trigger an “instantaneous AC undervoltage” trip of the
inverter. Inverter gating will stop immediately, followed by
the opening of the AC contactor after a few cycles. The output
filter capacitor will ring with the effective grid inductance for
a short time [7].
If inverters are designed to have LVRT capability, the
short-circuit contribution will be limited by the inverter
apparent power (kVA) rating and will essentially be similar to
that of a Type 4 (full converter) type wind turbine generator.
III. UTILITY-SCALE CSP POWER PLANT DESIGN
Utility-scale CSP power plants have been predominantly
solar thermal plants using parabolic trough technology. Since
these plants use conventional synchronous generators with or
without thermal energy storage the electrical characteristics of
the plant does not differ appreciable from that of a
conventional power plant. However, the development of
newer technologies in CSP plants, particularly plants using
dish Stirling systems, is creating new challenges in the design
of the low- and medium-voltage collector systems for large
solar power plants.
New dish Stirling systems use a 10 to 25 kW solar
concentrator in a dish structure that supports an array of
curved glass mirror facets. The mirrors collect and
concentrate solar energy into a power conversion unit, which
includes a reciprocating Stirling engine. The Sterling engine
uses an internal working fluid, typically hydrogen or helium,
which is recycled through the engine [8]. The Stirling engine
drives a squirrel-cage induction generator. Multiple 10 to
25 kW units are connected at low voltage into groups and
connected to a step-up transformer. As in large PV plants,
multiple step-up transformers are connected via medium-
voltage cable in a feeder arrangement and connected to the
collector or interconnect substation.
The thermal, electrical, and control systems of the dish-
Stirling system, along with a method for simulation are
presented in [8].
A. Reactive Power and Voltage Control
CSP plants using dish Stirling systems are designed for
transmission system interconnection and must meet typical
reactive power and voltage control requirements. The
induction generators operate at an inductive power factor
without compensation. Substation-based reactive power
compensation systems are included in the design to
compensate for the generator reactive power and to meet the
required power factor range at the POI. Dynamic reactive
power provided by a STATCOM is required for voltage
control at the POI and to provide for in-plant voltage
regulation during starting conditions. Switched capacitor
banks can be applied at medium and high voltage to meet the
reactive power requirements for large plants.
B. Low Voltage Ride-Through and System Stability
The induction generators and Stirling engines in dish
Stirling systems have minimal rotational inertia that results in
 3

the need for additional internal speed control measures to V. BIOGRAPHIES

avoid overspeeding during a grid fault. However, the side- E. H. Camm is a Principal Engineer at S&C Electric Company. He is
benefit of these speed-control measures allow the units to ride responsible for consulting and engineering analysis in areas including voltage
stability, wind plant modeling and interconnection, transients and power
through up to a few seconds of grid or in-plant faults prior to
quality, and application of large-scale power quality equipment solutions. He
eventual disconnection. The characteristics of the system, in is the chairman of the IEEE Power & Energy Society’s Subcommittee on
combination with dynamic and switched reactive power Integration of Renewables into the T&D Grids and secretary of the Working
equipment can be used to meet typical LVRT requirements. Group on Dynamic Performance of Wind Power Generation. He holds a
MSEE from the Ohio State University.
Dynamic simulation models of the plant, based on lumped
equivalents of power generating groups and the reactive S. E. Williams is a Senior Engineer at S&C Electric Company. He is
power compensation system, can be used to determine the primarily involved in modeling, analysis, and design of grid interconnected
large wind and solar facilities. Mr. Williams has also been involved with
LVRT capability of the plant. protective relay and power system sensor development, as well as analysis and
design of utility, industrial, and ship power systems. He graduated from
C. Short-Circuit Contributions Texas A&M University with BSEE and MSEE degrees in 1991 and 1995.
The induction generators in dish Stirling systems will
respond similarly to small induction machines to contribute to
faults. The peak fault current contribution will be
approximately 5 to 6 per unit, decaying to nominal current in
less than 6 cycles. Specific fault current characteristics will
also be affected by the response of the dynamic compensator
associated with the plant’s reactive power compensation
system.
D. Ramp rate control
Rapid power fluctuations in dish Stirling system plants will
be mitigated by the thermal inertia of the Stirling engine. The
composite effect of a large number of units will also mitigate
power ramp rates. However, during plant start-up, a large
number of units must be brought on-line in a relatively short
time period to maximize plant productivity. Start-up
sequences must be controlled to limit the power ramp rates to
within the specified limits. While each Stirling generator has
minimal output regulation, the net plant output can be readily
regulated on a per-machine curtailment basis.

IV. REFERENCES
[1] RenewableEnergyWorld.com Editors, "US Solar Industry Grows Despite
Economy," available at
http://www.renewableenergyworld.com/rea/news/article/2010/10/us-
solar-industry-grows-despite-economy.
[2] "Spain surpasses United States as world leader in installed CSP,”
available at
http://www.solarserver.com/solar-magazine/solar-
news/current/kw29/spain-surpasses-united-states-as-world-leader-in-
installed-csp.html .
[3] IEEE Standard for Interconnecting Distributed Resources With Electric
Power Systems, IEEE Std 1547-2003.
[4] NFPA 70: National Electrical Code®.
[5] National Electrical Safety Code (NESC®).
[6] UL Standard for Safety Inverters, Converters, Controllers and
Interconnection System Equipment for Use With Distributed Energy
Resources, UL 1741.
[7] Ben Kroposki,, "Distribution System ModelsPower System Studies and
Modeling PV Inverters," Utility/Lab Workshop on PV Technology and
Systems, Tempe, Arizona, November 8-9, 2010..
[8] Dustin Howard and Ronald G. Harley, "Modeling of Dish-Stirling Solar
Thermal Power Generation," in Proc. 2010 IEEE Power & Energy
Society General Meeting,, Minneapols, Minnesota, USA, July 25-29,
2010.