IEEE Power & Energy Magazine - May/June 2018 - 50

300

400

200
100
0
50

100 150 200 250
Azimuth Angle (°)
(a)

300

400
Number of Sites

500
Number of Sites

Number of Sites

400

300
200
100
0
0

40
60
80
Tilt Angle (°)
(b)

<100 kW

≤500 kW

300
200
100

100

0
0

0.6 0.8 1 1.2 1.4 1.6 1.8
ac-dc Ratio
(c)

>500 kW

figure 7. Distributions of simulated BTM PV system design parameters for Connecticut: (a) azimuth angle, (b) tilt angle,
and (c) dc-to-ac ration.

file and input data describing the system's characteristics. the
sam simulation core sDK enables the user to access the sam
application programmatically. sam simulations used the most
recent version of pvWatts (version 5). matlaB was used for
all simulation work; however, the sDK also contains language
wrappers for c#, Java, and python.
for the New england test case, simulations include more
than 41,000 unique pv systems located throughout the region
to capture diversity in the influence of both weather and pv
system design. Designs include project sizes ranging from
4 KWdc to 2,500 KWdc and a variety of azimuth angles (also
known as array orientation), tilt angles, and ac-to-dc ratios.
since tracking pv systems are rare in New england, all simulated systems were assumed to be fixed axis. in the absence of
known values for these parameters, a few high-level assumptions that reflect general observations of Btm pv installations
guided parameter selection. first, the total population of each
parameter's values reflects a normal distribution centered on
approximated expected values. these expected values are 180°
for azimuth angle (i.e., solar south), 1.2 for the dc-to-ac ratio,
and 30° for system tilt. second, parameter distributions reflect
a tendency for their values associated with larger pv systems
to be closer to expected values (i.e., their values exhibit less
variance), and those of smaller systems reflect a tendency for
greater diversity (i.e., to exhibit more variance). these guiding
principles reflect the observation that smaller rooftop systems
are often installed in less-optimal conditions, whereas larger
table 3. The simulated BTM PV systems by size class.

Capacity (kW)

Total Simulations

Total Selected

<100

26,676

≥100 ≤500

8,208

554

>500 ≤2,500

6,156

126

Total

41,040

27,356

ieee power & energy magazine

greenfield pv developments are more likely to be optimally
designed to maximize revenue, which reflects performancebased incentives. in addition to their use for addressing grid
integration issues, it is worth noting that Btm pv simulation
capabilities are also valuable for policy, investment, and system design decision-making considerations. figure 7 shows
resulting distributions of simulated parameter values for azimuth angle, tilt angle, and ac-dc ratio for different system size
classes in connecticut. table 3 lists the total number of systems simulated and selected across the entire New england
region for the development of the finalized town-level profiles.

Benchmarking Simulations
to Measured BTM PV Production
Given the lack of both detailed Btm pv design and a complete set of measured Btm pv performance data, iso
New england uses an approach analogous to one used in
Germany, for which a subset of information concerning
installations and available production data from a sample of
representative Btm pv sites are the basis of accurate estimates of Btm pv production. in iso New england's case,
the two primary inputs to the estimation are Btm pv installation data and vendor-provided Btm pv performance data,
both at the town/municipal level of granularity. the performance data at any given time step represent relative Btm
pv performance on a per-unit nameplate capacity basis, as
illustrated in the heat maps in figure 8. for example, the colors depicted within each map represent the Btm pv power
output located in each town as a relative share of its total
nameplate capacity. for both the simulated and measured
Btm pv production data, there are different numbers of
individual installations in each town with differing design
characteristics that are the basis of the per-unit profile.
towns without any pv systems are colored gray or black in
the measured Btm pv data maps. With the data structured
in this manner, the user can simply scale up a town-level profile by the amount of capacity installed in the town to yield
may/june 2018

Table of Contents for the Digital Edition of IEEE Power & Energy Magazine - May/June 2018